Book3 2015

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/284182791
Modern Geodetic Techniques in Spatial Measurement
Book · September 2015
CITATIONS READS
2 3,503
10 authors, including:
Pelc-Mieczkowska Renata Karol Dawidowicz

University of Warmia and Mazury in Olsztyn University of Warmia and Mazury in Olsztyn
19 PUBLICATIONS 105 CITATIONS 30 PUBLICATIONS 184 CITATIONS
SEE PROFILE SEE PROFILE
Michał Bednarczyk Monika Birylo

University of Warmia and Mazury in Olsztyn University of Warmia and Mazury in Olsztyn
22 PUBLICATIONS 66 CITATIONS 27 PUBLICATIONS 108 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Baltic Geodetic Congress (Geomatics) View project
Water Tower in Ciechanów View project
All content following this page was uploaded by Pelc-Mieczkowska Renata on 20 November 2015.
The user has requested enhancement of the downloaded file.

Michał Bednarczyk, Monika Biryło, Karol Dawidowicz, Volodymyr Hlotov, Kamil
Kowalczyk, Anita Kwartnik-Pruc, Renata Pelc-Mieczkowska, Zbigniew Siejka,
Dariusz Tomaszewski, Andrzej Wróbel
MODERN GEODETIC TECHNIQUES IN

SPATIAL MEASUREMENT
Zagreb, Croatia, 2015

Reviewers
Piotr Parzych
Katarzyna Pająk
Scientific Editor
Karol Dawidowicz
Published by:
Croatian Information Technology Society, GIS Forum
10 000 Zagreb, Ilica 191e, Croatia
Copyright ©
Croatian Information Technology Society, GIS Forum, Croatia
All rights reserved
Number of copies: 100
ISBN 978-953-6129-45-4 Nacionalna knjižnica, Zagreb, Croatia
2
CONTEXT
INTRODUCTION.............................................................................................................................................5
1. THE IDEA OF ADVANCED GNSS MISSION PLANNING BASED ON TERRAIN
OBSTACLES MODELS OBTAINED FROM HEMISPHERICAL IMAGES......................................6
1.1. Object and methods ............................................................................................................................. 9
1.2. Camera set calibration .................................................................................................................... 10
1.3. Hemispherical image........................................................................................................................ 12
1.4. Laser scanning data processing.................................................................................................. 13
1.5. Computer application for mission planning ........................................................................ 15
1.6. Results ..................................................................................................................................................... 20
1.7. Conclusion ............................................................................................................................................. 23
1.8. Additional conclusions .................................................................................................................... 24
2. POSSIBILITIES TO VERIFY SPATIAL DATA OF REAL ESTATE USING ARCHIVAL
AERIAL PHOTOGRAMMETRIC IMAGES............................................................................................26
2.1. A historical outline of creating the land cadastre in Poland since 1945 ............... 26
2.2. Current regulations on determining the cadastral data ................................................ 29
2.3. Procedures for determining the course of real estate boundaries .......................... 30
2.4. Aerial images - interpretation ..................................................................................................... 32
2.5. Capture of spatial data from aerial photographs .............................................................. 36
2.6. Conclusions ........................................................................................................................................... 39
3. DETERMINATION OF COORDINATES OF OBJECTS USING KINEMATIC
PHOTOGRAMMETRY FORMULAS WITH DIGITAL CAMERA AND GPS-RECEIVER........41
3.1. Determination of coordinates of digital photogrammetric survey points ........... 45
3.2. Topographical plans of the Vernadsky Antarctic station.............................................. 49
3.3. Device Connecting EXO with GPS .............................................................................................. 50
3.4. Laboratory facilities for research .............................................................................................. 51
3.5. Peculiarities of digital photogrammetric survey ............................................................... 56
3.6. Characteristics of orienting device ........................................................................................... 57
3.7. Conclusions ........................................................................................................................................... 60
4. OPTIMIZED TECHNOLOGY FOR GNSS HEIGHT DETERMINATION USING ASG-
EUPOS..............................................................................................................................................................62
4.1. European Position Determination System ........................................................................... 66
4.2. Height measurement using ASG-EUPOS ................................................................................ 68
4.3. Typical levelling situations ........................................................................................................... 72
4.4. Theoretical analysis of optimal technique ............................................................................ 73
3
4.5. Conclusions ........................................................................................................................................... 74
5. A RELATION BETWEEN A VERTICAL CRUSTAL MOVEMENT FROM SPIRIT
LEVELLING AND GLACIAL ISOSTATIC ADJUSTMENT (GIA) FROM GRACE DATA IN
THE AREA OF POLAND............................................................................................................................76
5.1. GRACE data and its usage for GIA determination ............................................................. 77
5.2. Vertical crustal movement in Poland ...................................................................................... 81
5.3. A relation between vertical crustal movement in Poland and GIA in Poland .... 84
5.4. Conclusions ........................................................................................................................................... 86
REFERENCES................................................................................................................................................88
LIST OF FIGURES........................................................................................................................................96
LIST OF TABLES..........................................................................................................................................98
NOTES ON THE AUTHORS......................................................................................................................99
4
INTRODUCTION
Friedrich Robert Helmert defined geodesy as the science “of measurements
and mappings of the Earth’s surface”. With the passing of time, this definition of
geodesy has been extended, generally as a consequence of technological
developments which allows geodesy to observe the Earth on global scales with high
accuracy.
Nowadays angles can be measured with single-second accuracy using optical and
electronic theodolites. The electromagnetic distance measuring equipment can obtain
distances up to several kilometers with millimeter precision. Global Navigation
Satellite Systems (GNSS) are used to set three-dimensional position with high
accuracy. The modern instruments and applications give the possibility to increase
the mass and the variety of achievable information. Currently, modern geodesy based
on robotic total stations and GNSS, digital photogrammetric techniques, high
resolution satellite imagery, airborne and terrestrial laser scanners devices,
gravimetric missions are able to provide a set of powerful tools for the geometric
surveying and modeling of Earth’s surface.
This book presents current research on the development of modern geodetic
techniques in spatial measurements.
In the first chapter the authors present advanced GPS mission planning
method. The distinguishing feature of this method is to enhance the well-known
process of GNSS mission planning for the introduction of terrain obstacles models.
The obstacles models were derived from hemispherical photographs and terrestrial
laser scanning. The next chapter presents an analysis of the possibilities to use
archival aerial photographs to determine the location of a real estate boundary, based
on specific cases from the conducted research studies. The third chapter discusses the
actual problem of photogrammetry: not only determination the position of object in
space but also in time. The solution of this problem became possible owing to modern
means used in geodetic and photogrammetric production. First of all this is stipulated
to application of GPS systems, combined with digital camera. The fourth chapter
examines in detail the limitations of GNSS heighting including those factors that affect
the GNSS height measurement itself and the associated issues of geoid modeling and
compatibility with the local vertical datum. The last chapter contains the studies of
the crustal vertical movement in Poland computed on the basis of spirit levelling data,
and its relation to the Glacial Isostatic Adjustment (GIA) from GRACE mission data.
This book reviews and interrelates a large number of theoretical and
experimental contributions to the research on GNSS technology, photogrammetry,
gravimetry and is suitable for engineers who are actively engaged in research as well
as for geophysists, geodesisists and students.
Scientific Editor
Karol Dawidowicz
5
1. THE IDEA OF ADVANCED GNSS MISSION PLANNING
BASED ON TERRAIN OBSTACLES MODELS OBTAINED
FROM HEMISPHERICAL IMAGES
The widespread availability of GNSS Permanent Station Networks as well as dynamic
development of GNSS systems constellation simplified satellite positioning and
surveys and made them quick, cheap and available for a wide group of users.
Moreover the GNSS positioning methods are weather independent and such
measurements can be conducted regardless of the time of the day. The major
limitation of GNSS measurement techniques is a need to provide direct, uninterrupted
reception of signal from a minimum four (three in the case of positioning 2D)
satellites. Nowadays the GNSS constellation consisting of 31 GPS
(http://www.gps.gov), 24 GLONASS (http://glonass-iac.ru/en/), 4 Galileo
(http://gpsworld.com/the-almanac), 16 Beidou (http://en.beidou.gov.cn) and ABAS
satellites (Satellite-Based Augmentation Systems), which ensures compliance with
this requirement always, throughout the of Earth, but the visibility of satellites may
be limited by the presence of terrestrial obstacles. Huge terrain denivelations, high
buildings, bridges or trees branches and canopies prevent GNSS satellite signal
reception what affects proper ambiguity of phase measurements solving and
significantly decreases reliable coordinates determination (BAKUŁA et. al., 2012;
BAKUŁA et. al., 2011; BAKUŁA et al., 2009). For those difficult measurement
conditions careful mission planning is essential because the limited visibility of
satellites prevents successful GNSS positioning.
For the proper mission planning of GNSS observation sessions one needs to know:
 the expected positions of the particular satellites relative to the GNSS
receiver,
 the shape terrestrial obstacles,
 the deployment of terrestrial obstacles.
Currently, users of GNSS have the possibility to use a variety of mission planning
applications in which the position of the satellites is determined on the basis of the
GNSS almanac. When it comes to terrain obstacles, most of those applications allow to
create just hand-drawn polar diagrams by determining each obstruction point by
point. This way of determining the neighbourhood of the point is inconvenient and
extremely inaccurate. This task can be performed more accurately by applying some
semi-automated terrain obstacles modelling methods.
If one want to consider modern data acquisition methods suitable for terrain
obstacles shape and position determination, among others Digital Terrain Model
(DTM) or Digital Surface Model (DSM), tacheometry, laser scanning and
hemispherical photography should be mentioned. The main advantage of using the
DTM or DMS is that it provides us with information about the terrain obstacles
occurring not only at individual measurement point but at a wider area (FEDERICI et
al., 2013; D’INCA et al., 2004; TAYLOR et al., 2007; GROVES, 2011; ACKERMANN et al.,
2013). This approach allows to identify areas convenient for conducting GNSS
positioning at any given time and also makes it possible to plan the kinematic
positioning, for instance in case of navigation. On the other hand, it is not possible to
determine observation window from DMS or DTM without knowledge about accurate
6
coordinates of measurement point, which coordinates are just to be measured. This is
because terrain obstacles are relatively close to the measuring point, so the shape of
the open sky over each point is changing rapidly with the small changes in the
position of the receiver. For that reason some methods of the direct measurement of
the shape and deployment of terrain obstacles for particular points should be applied.
In this study terrestrial laser scanning and hemispherical photography were
mentioned.
Terrestrial laser scanning allows for fast acquisition of 3D information needed
to represent objects that are complex in shape and whose measurement with use of
conventional methods would be very difficult or impossible. The basis of laser
scanner measurement is simple and the data acquisition process is relatively fast.
Thanks to laser scanning one can get sets of points presenting the environment
around the point that will be measured with satellite technique. Elaboration of these
data can, however, cause some problems arising from the size of the dataset and the
need to transform it so as to allow for the determination of satellites visibility above
the measured point. All points registered by the scanner are in topocentric coordinate
system defined within measuring device. Determination of the GNSS signal obstacles
with the use of such coordinates can be approached in two ways. First approach is to
calculate the position of GNSS satellites in scanners topocentric coordinate system
and determine whether the desired signals will face obstacles. However, this task can
be problematic because it requires the knowledge of the exact position of the scanner
in WGS84 coordinate system. The second approach is to determine elevation and
azimuth of all scanned points with respect to the horizontal plane determined by y
and x axes of the scanner and true north. Subsequently the elevation and azimuth of
satellites which can be located above the receiver's antenna are determined on the
basis of measurement point approximate coordinates and satellite orbit parameters
which can be found in the almanac.
Hemispherical photography, despite its large amount of distortions, can be
successfully used in many scientific fields, such as forest research, meteorology and
surveying. This kind of image offers the advantage of providing spatially resolved
information about all the obstacles above the measured point. Moreover, azimuth and
elevation of terrain obstacles, required in GNSS mission planning process could be
derived almost directly from the hemispherical image, without any sophisticated
transformations, therefore it can be a great tool for GNSS site satellite window
diagram direct determining (Colcord, 1989; Pelc-Mieczkowska, 2014). However, one
should bear in mind that proper projection model should be applied and used camera
set have to be calibrated. Fish-eye cameras are designed to map the whole
hemispherical field in front of the camera, what is impossible in perspective
projection. As indicated in the literature, the most common projections obey by the
fish-eye lense are: stereographic projection, equidistance projection, equisolid angle
projection and orthogonal projection (KANNALA et. al., 2006). Because the real
lenses do not exactly follow the designed projection model hemispherical images
suffer from radial distortion which is performed along the radial direction from the
centre of image and tangential distortion, directed perpendicularly. The value of
distortion at a given image point is a function of incoming ray angle of incidence.. The
radial distortion causes an inward or outward displacement of a given image point
from its ideal location. The negative radial displacement of the image points is
7
referred to as the barrel distortion, while the positive radial displacement is referred
to as the pincushion distortion (MA et. al., 2004). The tangential distortion causes that
every object in the image is moved from theoretical position in a direction
perpendicular to radial distortion direction. In good quality lenses the value of
tangential distortion is about 10 times less than radial distortion (LUHMANN et. al.
2006, REMONDINO et.al. 2006), so its effects are mostly negligible (HUGEMANN
2010). The distortion model could be determined empirically in calibration process in
which distortion parameters are estimated based on least squares polynomial fitting.
For this purpose, calibration object equipped with a number of survey marks with
known 3D coordinates should be established.
Beside camera kit calibration, hemispherical image based modelling of terrain
obstacles involves some image segmentation to determine which of the segments
refers to the obstacles, and which refers to the open sky. Image segmentation is a
term used in computer vision to determine a process of partitioning digital image into
multiple segments. Those separated segments are also called superpixels. The goal of
segmentation is to simplify or change the representation of an image into something
that is easier to perform analyses (LINDA et al.). Image segmentation can be used to
determine objects such as lines, curves, areas and others. The result of segmentation
is that every pixel in an image is designated using specified criteria. Pixels with the
same designation are part of a single superpixel (segment). After segmentation image
becomes a set of segments or a set of contours extracted from it. Each segment differs
from others by colour, brightness, texture or by other attribute depending on
segment’s properties. Image segmentation is used in many fields, e.g. in. medical
imaging, object or face recognition, fingerprint recognition, satellite images analysis
or traffic control systems. There are many methods of segmentation such as
thresholding, clustering, edge detection, region-growing, graph partitioning and
other. Each of which uses different algorithms and criterias to determine segments,
depending on what purpose it was developed.
The most appropriate method for hemispherical images segmentation to
terrain obstacles modelling is the thresholding method. This method is based on a
threshold value to turn a grey scale image into a binary image. This value can be
determined in various ways, using for example the maximum entropy method,
maximum variance or k-means clustering. It can be also determined visually. In this
case threshold value can be obtained on the basis of the resulting visual effect.
Thresholding is common step in an image analysis, where we need to distinguish
pixels by two different brightness area, for example between object (e.g. obstacle)
and background (e.g. the sky).
In this study fish-eye lense camera set calibration procedure, principles of
vertical images execution and dedicated methods of hemispherical images
segmentation and terrain obstacles modelling are presented. The algorithm realizing
developed methods has been implemented in a computer application called Obstacles
Modelling and Mission Planning Tool (MP+), which has also been described by the
authors herein. Subsequently the test models were developed for two GNSS
measurement points in different observational conditions. Terrain obstacles models
obtained from hemispherical photography are compared with one obtained from
laser scanning and both of the are used in advanced mission planning procedure. The
8
results of the mission planning procedure are also presented in the article and
compared with the actual GPS observations results.
1.1. Object and methods
To perform the study of hemispherical photography suitability for GNSS

mission planning two test points were selected. First of these points (named F) was
placed in forested area where positioning was complicated due to the tree canopy
directly above the measured point. Most of the trees in the neighbourhood of the
point were conifers and single deciduous trees had no leaves because of the season.
As a second test point (named C) the place in area between two buildings was
selected to simulate measurements at dense urban area (Fig. 1).
Fig. 1. Measurement points locations.

Resource: elaborated by PELC-MIECZKOWSKA.
At both measurement points vertical, hemispherical images were executed

and laser scanning measurements were performed in order to obtain spatial
information about the shape and the distribution of terrain obstacles around these
points. In addition 6 hour lasting GPS positioning sessions were conducted
simultaneously for both points. At each point the fixed height, the same for GNSS
antenna horizon, laser scanner horizon and the gold ring of fish-eye lense was set up.
Nikon D90 12.3 megapixel digital semi-professional camera with Sigma 4.5
mm f/2.8 EX DC HSM circular fish-eye lens (180 degree field of view) have been used
to obtain hemispherical images. The camera set was mounted on geodetic tripod and
carefully levelled and centred over measurement point. To indicate the north
direction the geodetic pole was used.
Subsequently, over the selected points Leica C10 scanner was placed so that
the height of his telescope coincided with the height at which the hemispherical
images were taken. The scanner was set to scan density 0.005 x 0.005 x 100 m. To
properly illustrate the obstacles which are above the GPS antenna, all objects in
vertical range 0∘ - 90∘ and horizontal range 0∘ - 360∘ were scanned. Measurement
9
lasted about 30 minutes during with 27 million points have been registered. As in
previously described hemispherical photography to indicate north direction the
geodetic pole was used.
The GPS measurement was performed on February 27th. Two high quality,
geodetic TopCon Hiper Pro GNSS receivers were used. During measurement session
lasting from 8:00 to 13:46 GPS time GPS observations were conducted
simultaneously for points F and C.
1.2. Camera set calibration
Hemispherical images suffer from large amount of distortion due to

optomechanics of fish-eye lenses. As it was mentioned in introduction, for derived
distortion model of camera set, some calibration process should be done. In the
presented research, distortions parameters are estimated based on least squares
polynomial fitting. To determine polynomial coefficients empirically the calibration
box with set of check lines was established (Fig. 2). The calibration box was designed
in a way to ensure mapping points of intersection the vertical with horizontal lines
over the entire surface of the image. The distance between lines equals 5mm, the
dimensions of box: 404x396x200 mm and the camera set was situated in the way that
the gold ring was within 171 mm from the wall perpendicular to the axis of the lens.
Fig. 2. Calibration box.

Besides radial and tangential distortion, in case of short (less than 2 m) line of
sight, the entrance pupil point position and its movement had to be taken into
account. Those values were determined experimentally with low power laser beam
technique. The position of entrance pupil point related to the angle of incoming ray
was referred to the position of golden ring marked on the lens. For Sigma 4.5 mm
f/2.8 EX DC HSM circular fish-eye lens that offset is small and ranged from -0.2 cm to
0.2 cm (Fig. 3).
10
Fig. 3. The entrance pupil point movement for Sigma 4.5 mm f/2.8 EX DC lense.
Based on determined entrance pupil point position, known line of site length
and line-to-line distances, reference values of elevation of intersection points was
calculated. The radial distance r of those points images were also measured in the
image. Afterwards, elevations (el) of those points were calculated having regard to
equisolid angle projection principals from the formula:
=2 . (1)
For calibration model determination, the odd-order polynomial radial distortion

model in general form:
( )= + + + + +⋯ (2)
was used therefore five polynomial coefficients were derived. The coefficients of
polynomial were calculated with 147 degrees of freedom on which the residual
standard error was only 0.185. Graph that shows the fitting of a polynomial is
presented below (Fig. 4).
11
Fig. 4. Polynomial fitting.
Resource: elaborated by TOMASZEWSKI.
Calculated polynomial coefficients were applied to MP+ application described

in detail later in this article.
1.3. Hemispherical image
The hemispherical photographs shown below were taken at measuring

points. Digital images were taken with a resolution of 4288x2848 pixels but due to
fisheye lense geometry, image of the subject was in the shape of a circle with a radius
of 1142 pixels (Fig. 5).
Fig. 5. Hemispherical images: on left - point between the buildings, on right - point in the forest.
To be suitable for Terrain obstacles modelling colour images should be

transformed into binary image. First step to prepare the binary image is to separate
obstacles from the sky explicitly by segmentation. It can be simply done by
thresholding. For this purpose popular, open source graphics software GIMP was
used, but this functionality will be implemented soon in the next version of MP+
application. As a result of segmentation a binary image is created, where black colour
denotes an obstacle and white an open area (Fig. 6). The thresholding value was
determined visually by hand in such a way as to leave more massive elements as
12
obstacles. While those by which GPS signal can pass through, such as small branches
of trees, were deleted. The threshold value for both images is different and equals 160
for image on left and 133 for image on right (as denoted in Fig. 5 and presented in Fig.
6).
Fig. 6. Binary images created on the dasis of hemispherical photograps.

Resource: elaborated by BEDNARCZYK.
1.4. Laser scanning data processing
On the basis of performed measurements two point clouds representing

objects disturbing GNSS signals above selected points were obtained. Immediately
after the measurement data has been uploaded into the Leica Cyclone suite (Fig. 7).
Fig. 7. Obtained point clouds.

Resource: elaborated by TOMASZEWSKI.
13
Coordinates of points in obtained clouds were in scanner coordinate system
with origin in the centre of the telescope. Such data could not possibly be used for
determining the GNSS signal obstructions. Therefore, the coordinates of all the points
have been exported to files and then converted into topocentric horizontal coordinate
system with origin in the centre of scanners telescope. The values of the elevation of
all measured points were obtained from the known growth of coordinates relative to
the centre of the coordinate system. The azimuth of each point was defined with
respect to the points of scanned geodetic pole, which was placed to point true north.
Obtained topocentric horizontal coordinates of all points were then transformed to
cartesian coordinate system of the image to compare the laser scanning results with
hemispherical photography. Subsequent transformations are shown in the scheme
below (Fig. 8).
Fig. 8. Diagram showing the transformation process.

On the basis of the previously obtained cartesian coordinates the images of

measured obstacles were obtained (Fig. 9).
Fig. 9. Binary images created on the basis of laser scanning.

Resource: elaborated by PELC-MIECZKOWSKA, TOMASZEWSKI.
14
1.5. Computer application for mission planning
Nowadays, there are many computer applications for mission planning, but
none of them has a functionality of taking into account obstacles derived from
hemispherical images or any other accurate data acquisition methods. That is why,
for the purposes of this work, a dedicated computer application had to be created,
called MP+. Main window of described application is shown in Fig. 10. In general, the
application is combining two functions: one associated with loading the image, setup
correction parameters and generating obstacles model while the second one is
related to loading the parameters of the satellite constellation and calculating the
number of available satellites and the value of PDOP coefficient. As a results of joining
obstacle modelling and conventional mission planning, predicted number of available
satellites and PDOP values, taking into account terrain obstacles, are obtained. . To
make the work easier, application interface provides the functionality of displaying
polar grid, north direction and satellites positions with labels. It was created using
Code Gear Rad Studio 2007, in Object Pascal programming language.
Fig. 10. MP+ application - main window with data loaded.

Preparing and loading the image

Application can load an image stored in device independent format, such as
BMP file (also known as device independent bitmap - DIB). DIB defines bitmaps in
various colour resolutions from 1bit per pixel to 24bits per pixel. The main purpose
of DIB is to allow bitmaps to be moved between different devices, besides it is
commonly known. This makes it to be quite versatile format, and therefore it has
been chosen by the authors. Obstacles model creation requires two types of images.
One is coloured (e.g. truecolour 24bit/px) bitmap for visual setup of the hemisphere
15
parameters, the second is binary image, segmented by thresholding (as mentioned
before), to create the model itself. Image loaded into the application is the basis of
created model and further mission planning. Both images (coloured and binary)
should be converted into DIB format and so prepared can be loaded into MP+ by
choosing File->Open from the main menu.
Setting up model and correction parameters

After preparing image data, user needs to setup some parameters associated
with calibration of the image. They can be entered in the Model settings window (Fig.
11). All angular values in application should be entered in degrees.
Fig. 11. MP+ application - model settings window.

First of all the colour image must be loaded to setup the hemisphere. In this
step colour image is needed because, it’s easier to identify details on it, than on a
binary version. The hemisphere parameters are:
 Azimuth shift - an angular value indicating the north direction measured from
vertical axis of the image. It can be entered by hand or indicated by pointing
cursor on an image.
 Radius - radius of the hemisphere in pixels, it can be entered or pointed on an
image. It’s size should be taken as a radius of a circle seen on the photograph.
 Angular range - hemisphere angular range, counted from the horizon to the
zenith. It depends on the lens characteristics.
 After that, other calibration parameters can be entered such as:
 Image geometric centre shift - these are two values of X and Y in pixels,
denoting the shift of the central point of the hemisphere from the centre of an
image. It should be measured while calibration process.
16
 Distortion correction parameters - these are values of polynomial coefficients,
determined while lens calibration process.
 To complete obstacles model setup, also are needed to enter:
 Image information such as filename and resolution
 Camera information such as name, type, lens type and focal length
Entered parameters are needed by the application algorithms to properly
calculate values of azimuth and elevation from the image, taking into account lens
projection and distortion. They can be saved into text/xml file, and loaded when
needed. Files are named with *.xmo extension to distinguish them from other xml
datastores and make more associated with MP+ application. The structure of this file
allows to store all of the model parameters described above. It is consisted of XML
tags, which makes it self-descriptive format. Every value is tagged as follows:
<[parameter name tag]> [parameter value] </[parameter name tag]>
for example:
<Aparat>NIKON D90</Aparat>
stores type of camera used for taking a photo. Sample content of the file is shown in
Fig. 12.
Fig. 12. MP+ application - content of the *.xmo file.

Obstacles model
Once set up and saved model parameters data in *.xmo file can be used many
times for different images taken with the same lens and camera kit and with the same
resolution. It simply can be loaded by pressing Load button in Model settings window
(Fig. 11). To create the obstacles model and start mission planning a binary,
segmented image must be loaded. It becomes a basis of obstacles model. Every black
pixel in the image denotes the obstacle (1 or true), every white pixel the open sky (0
or false) (Fig. 13).
17
Fig. 13. The principle of creating an obstacles based on binary values.
Elevations and azimuths of every pixel in the image is calculated in real time,
on the basis of parameters stored in *.xmo file. This allows to specify which location
on the image denotes the obstacle. In order the relationship between the files was
unequivocal, application gives the *.xmo file the same name as the image file (only
extension is different). After loading both binary image and *.xmo files (Fig. 14) one
gets a complete obstacles model for further mission planning, which can be stored
and loaded when needed. If there are both files created, application loads them
together while opening an image file.
Fig. 14. Obstacles model dataset schema.

Mission planning
When there’s an obstacles model defined a mission planning can be performed. To do
that, we need the predicted positions of satellites above the point of measurement,
expressed as the azimuth and elevation. Such information could be derived from
almanac data provided for instance in YUMA, SEM or any other data format as well as
navigation message received during GNSS positioning or some IGS product. Besides
satellite orbit parameters coordinates of measurement station and the time range for
which planning has to be done are necessary. At present, the test version of the
application azimuths and elevations data are loaded from file prepared externally.
Now, position of each satellite from every epoch will be applied to the obstacles
model, which allows to determine the obstacle value. In this way, satellites on
18
positions behind obstacles can be filtered from calculations for mission planning.
Application automatically calculates SVS and PDOP parameters for each epoch. SVS
denotes number of available satellites, while PDOP is one of DOP (Dilution of
Precision) parameters which is the geometry factor expressing composite effect of
relative satellite-user geometry on the GPS solution error (Kaplan et.al. 2006).
Predicted values of DOP coefficients can be calculated from satellite azimuth and
elevation angles (Banerjee et.al. 1996) as follows:
−cos ( ) ( ) − ( ) ( ) − ( ) −1
−cos ( )cos ( ) − ( ) ( ) − ( ) −1
= (3)
⋮ ⋮ ⋮ ⋮
− ( ) ( ) − ( ) ( ) − ( ) −1
Where Az and el denotes azimuth and elevation angles. From A matrix, Q matrix is
calculated as:
=( ) (4)
If the diagonal elements of Q are qxx, qyy, qzz, then PDOP is:
PDOP= + + (5)
In order to provide the influence of obstructions on the satellite visibility and
PDOP value, the SVS and PDOP diagrams with and without obstructions are
generated by application. Example SVS/PDOP chart is shown in Fig. 15.
Fig. 15. MP+ application - SVS/PDOP resulting graphs.

19
Results of mission planning can be saved into vector graphics in enhanced
metafile (EMF) format and also as a text file. In this way they can be processed in
external software. The format of the resulting text file is as follows:
[TIME] [SVS] [PDOP] - one line per epoch
[PRN] [AZ] [ELEV] [OBS_VAL] - separate line for each satellite
Where:
[TIME] - time of measurement (epoch) in format HH:MM:SS
[SVS] - number of satellites available - integer value
[PDOP] - PDOP value - decimal value
[AZ] [ELEV] - azimuth and elevation of the satellites in degrees - decimal value
[OBS_VAL] - obstacle value - boolean value, 0 means no obstacle, -1 obstacle
1.6. Results
As mentioned earlier, two test points C and F and test time window from 8:00
to 13:46 GPS time were established. Created MP+ application was used for advanced
mission planning procedure for that points and time. To demonstrate significance of
taking into account the shape and distribution of terrain obstacles in the GNSS
mission planning process, at first the graph of predicted availability of satellites and
PDOP coefficients values without regard to any terrain obstacles was generated. Due
to the relatively small distance between the measuring points C and F (less than 100
m), the constellations of satellites observed from both points can be regarded as the
same. As it can be seen in the figure 16 if not for the terrain obstacles, throughout the
test time window the circumstances at the measurement points are excellent for
carrying out GNSS measurements. During the measurement time average number of
visible SV is 12 and the minimal value is 11, while PDOP value do not exceed 2.
Fig. 16. Predicted number of available satellites and PDOP value without obstacles model.
Resource: elaborated by BEDNARCZYK, PELC-MIECZKOWSKA.
To meet the assumption of real terrain obstacles taking into account, for both
tested points two versions of obstacles models had to be created. One with
hemispherical images usage and the second with images prepared from laser
scanning measurement. In the second case there was no need to perform correction
of distortion, because the coordinates were computed directly from XYZ coordinates
of points cloud created by the laser scanner.
20
Inclusion of the terrain obstacles models resulted in a significant reduction of
the available satellites number with a simultaneous increase of PDOP values (Fig. 17
and 18). Moreover, these values and their time distribution vary considerably for
each of measurement points. Point C is characterized by a lower average number of
available satellites than the point F, and changes of this value, as well as changes of
PDOP values were less dynamic. This is because most of the curtains at point C were
buildings which completely block the GPS signal reception while point F was situated
under trees and satellites were only temporarily obscured by branches.
Fig. 17. Predicted number of available satellites and PDOP values for point C: a) obstacles model
from laser scanning, b) obstacles model from hemispherical image.
Depending on the data used for obstacles modelling (laser scanning and
hemispherical photography), charts of satellites available did not differ significantly,
especially in case of point C. The number of satellites available at point C ranged from
7 to just 3 and does not vary because of the method by more than 1 in particular
epoch. The value of PDOP coefficient repeatedly exceeded 6 or even 10 but one can
indicate time periods appropriate to the GPS measurements e.g. 9:17 to 9:50 or
11:53-12:06.
In case of F point, which is characterized by very difficult conditions,
predicted number of available satellites and value of PDOP changes rapidly during the
time. The difference between the number of visible satellites predicted with use of
different obstacles model are slightly larger. Despite the fact that, in this case can also
be distinguished periods of time in which accurate and reliable positioning will not be
possible (e.g. 11:30-12:19) , and such when the GPS session could succeed (e.g. 8:15-
21
8;35). During test time window the number of available satellites vary from less than
4 up to 11.
Fig. 18. Predicted number of available satellites and PDOP value for point F: a) obstacles model
from laser scanning, b) obstacles model from hemispherical image.
In the final investigation, the number of satellites actually available during the
static measurement session is presented. It turned out that both receivers mostly
recorded signals of the greater number of satellites that it were carried out from
mission planning. To understand this difference not only the number of observed
satellites was verified but also how many of them will be used to determine the
position by the SPP algorithm. First of all it was carried out that some received signals
was inappropriate to position determination. Despite that, the numbers of satellites
used for positioning in particular epoch was still greater, than the one predicted while
PDOP values were lower (Fig. 19). It should be noted that algorithm which was used
in the study was simple SPP computation. If the post-processing algorithm would be
applied, the part of most obstructed signals would be excluded from the computation
and graphs would become significantly similar to each other.
22
Fig. 19. Predicted number of satellites used by positioning algorithm and PDOP value: a) point C,
b) point F.
1.7. Conclusion
In this study author's, advanced GPS mission planning method was presented.
The distinguishing feature of this method is to enhance the well-known process of
GNSS mission planning for the introduction of terrain obstacles models. The obstacles
models were derived from hemispherical photographs and terrestrial laser scanning.
Both of those methods are very efficient in data on the neighbourhood of the
measurement point acquisition because they are quick and easy to perform. The great
advantage of those method is that obstruction model is in any case made exactly at
the point of measurement, which avoids errors caused by imprecise measurement
station coordinates. Unfortunately the laser scanning technique requires very
sophisticated than expensive device and advanced software to data processing. The
hemispherical photography made with semi professional camera kit allows to obtain
sufficient accuracy to support GNSS mission planning. The processing of data
obtained from hemispherical photography involves camera set calibration and image
segmentation which could be performed automatically by thresholding algorithms.
On the basis of comparative analysis of results from the two considered
methods, it can be concluded that they provide similar results. However, as it was
mentioned before, the process of data acquisition is significantly more expensive and
time-consuming in case of laser scanning. While taking a hemispheric image requires
a professional camera and software to automate the calculation. Despite the fact that
the purchase of photographic equipment incurs certain costs, however they will be
many times lower than the performance of laser scanning.
23
In fact, tests indicate a good reliability of the prediction of the number of
satellites resulting from MP+ application; when the numbers of visible satellites do
not agree however the difference is small and the prediction is mostly conservative.
Proper determination of obstacles above GPS measurement point, makes it possible
to perform efficient GPS measurements even at hard observational conditions.
1.8. Additional conclusions
As it was previously mentioned while performing research on advanced

mission planning, two nearly six hour sessions were conducted. Subsequently
position of the receiver for each measurement epoch was determined. The algorithm
for determining receivers position erased observations from all satellites that caused
the biggest errors of the final result. On the basis of the results of mentioned
algorithm, and analysis on advanced mission planning, consecutively determined
numbers of satellites that:
 should be observed by the receiver without any obstacles,
 were seen by the receiver,
 were processed by the positioning algorithm.
The results of these analyses are shown in the sky plots below (Fig. 20 and 21).
Fig. 20. Sky plot of satellites on point F: a) All satellites, b) registered by the receiver c) used by
algorithm
Resource: elaborated by BEDNARCZYK, PELC-MIECZKOWSKA, TOMASZEWSKI.
Fig. 21. Sky plot of satellites on point C: a) All satellites, b) registered by the receiver c) used by
algorithm
Resource: elaborated by BEDNARCZYK, PELC-MIECZKOWSKA, TOMASZEWSKI.
24
As it can be seen in the figures above, used SPP algorithm was able to cut off
most of the measurement data that were obstructed. However after deeper analysis
of SV visibility on point C one can notice that receiver registered data from satellites
that were covered by the building. This is likely due to the fact of multipath error.
Building behind which the measurement data were provided is lower and the signals
from the satellites find their way to the zenith of the antenna after reflection from the
adjacent wall. Reflected signals were strong and therefore has not been eliminated,
neither by the receiver nor by the computational software. If the obstructed data
collected during this session would be used for precise positioning, computational
software that performs post-processing would also use the reflected signals for
vector calculation. Use of such data would significantly decreased the accuracy of
obtained coordinates. However that hemispherical images of the sky above
measurement points can be used as an additional information regarding correctness
of the measurements made by the receiver. On the basis of performed images, one
can specify satellites from which the signal was disturbed and then remove them
from the subsequent calculations. Based on the above analysis, it can be concluded
that the hemispheric images can be used as a tool for pre-processing of the
measurement data before performing post-processing.
25
2. POSSIBILITIES TO VERIFY SPATIAL DATA OF REAL
ESTATE USING ARCHIVAL AERIAL PHOTOGRAMMETRIC
IMAGES
Since 2010, Poland aims to create an ICT database comprising spatial data sets of
spatial information infrastructure including, among others, the register of land and
buildings, geodetic records of public utilities as well as topographic database, so
detailed as to ensure the creation of standard cartographic studies. For cities and for
rural areas with dense development and areas intended for development, these are to
be topographic databases, so detailed as to ensure the creation of standard
cartographic studies at the scales of 1:500-1:5000, harmonized with other databases.
The concept of harmonization of data sets refers to the activities of a legal, technical
and organizational character, aimed to bring consistency to these collections of data
and adaptation to their common and combined use.
Unfortunately, the quality of the data stored in the national geodetic and cartographic
information center and their mutual consistency is often insufficient and does not
allow for a correct definition of real estate boundaries. This is largely the result of the
manner of creating the register of land and buildings in Poland. Thus, the first part
presents the legal bases and the sources of data capture in establishing the record of
land and buildings.
Further, surveying and legal proceedings were described which result in determining
the course of real estate boundaries. The data on boundary location captured
therefrom are used to update the spatial data sets of spatial information
infrastructure. Since data sources regarding the location of a real estate boundary
which are kept in the poviat geodetic and cartographic information centers are often
insufficient, it was decided to look for other sources.
The aim of this study was to examine the extent to which archival aerial
photogrammetric images can contribute to the process of determining the position of
real estate boundaries. Aerial photographs are a fully objective archival material
documenting land development. Their usefulness to determine real estate boundaries
at a specific point in time (on the day of taking them) is limited, though. Aerial
photographs do not offer a possibility to identify the actual course of a real estate
boundary, however, it is possible to recognize the location of elements such as fences,
hedges, field margins, buildings and roads.
This study analyzes the possibilities of using archival aerial images to determine the
location of a real estate boundary based on specific cases from the conducted
research.
2.1. A historical outline of creating the land cadastre in Poland since 1945
When the land cadastre, further converted into land records in 1955, was being
established after the war, the data from the Austrian cadastre, whose documentation
covered approximately 14% of the Polish territory, and the Prussian cadastre, which
covered approximately 40% of Poland, were used among other things.
The first post-war legal act regulating the issues of land cadastre was the decree of 24
September 1947 on the land and building cadastre (DEKRET, 1947). This decree,
since its entry into force, i.e. since 29 September 1947, repealed all applicable
26
regulations issued by the partitioning powers. Cadastral surveys were to be
maintained in “consistency with the status in the field” (art. 6). The requirement of
maintaining the cadastral survey “in permanent consistency with the entries in land
and mortgage registers” (art. 7, section 1) by surveyors is worth emphasizing. Only
the decree of 1947 regulating the cadastre contained a provision on maintaining
consistency between the cadastral survey and the land and mortgage register.
Over the next three years there was no secondary legislation published for this
decree. This resulted in a situation in which certain government departments and
business entities who needed various cadastral data began to establish their own land
register. In December 1950, the Regulation of the Minister of Construction was issued
on the procedure regarding organizing, establishing and running of land and
buildings cadastre (ROZPORZĄDZENIE, 1950). The presidia of the national councils
were obliged to run the land and buildings cadastre. It was assumed that in order to
prepare inventories and descriptions of land and buildings for establishing a unified
cadastral survey, the following sources will be used:
 for the lands with the post-Austrian or post-German cadastre - data from this
cadastre, after their updating,
 for the areas without the cadastre - measurement data held by surveying
authorities, state and local government agencies, as well as public institutions and
private individuals, providing that the surveying authority approves these data as
suitable to establish the cadastre,
 for the areas without either the cadastre or measurement data - the results of
measurements and investigations of the surveying authority.
In practice, neither the provisions of the decree nor of the Regulation were
implemented (FEDOROWSKI, 1974).
The normative act regulating land records, which was the first implemented one, was
the decree on land and buildings register issued on February 2, 1955 (DEKRET,
1955). Keeping records was left in the hands of the presidia of poviat councils. As of
the decree's entry into force, i.e. 15 February 1955, the land cadastre was converted
into land register. Data on land and buildings, used as the basis for economic
planning, assessment of taxes and benefits, purchase and compulsory supply, making
entries in land and mortgage registers and meeting the economic needs, could be
based solely on the records. It was assumed that until basic cadastral maps were
prepared, any existing maps and statistical materials could be used for registration
purposes.
For the implementation of the provisions of the decree, two separate departmental
instructions were issued - instruction of the Minister of Agriculture of 21 April 1955
on establishing and running land register, rules for preparing land records, and fees
for drawing copies of maps and for copies and excerpts from records and documents
which are components of the cadastral survey in the areas of settlements, housing
estates and cities which are not poviats (INSTRUKCJA, 1955), and the instruction of
the Minister of Public Utilities of 8 September 1956 on establishing and running land
register, rules for preparing land records and fees for drawing copies of maps and for
copies and excerpts from records and documents which are components of the
cadastral survey in the area of the capital city of Warsaw, the city of Lodz and the
cities which are poviats (INSTRUKCJA, 1956). The bases for establishing a land
register were, inter alia, the maps and documents of the existing land cadastre, the
27
maps and measurement and classification records prepared in connection with the
redevelopment of the agricultural system, as well as other maps and surveying
documents (consolidating, delimiting). While establishing the register, both Prussian
and Austrian cadastral documents were used, as well as land survey records created
after the war.
The existence of two separate instructions led to dichotomy and differences in
interpretation. Therefore, in 1969, the ordinance of the Ministers of Agriculture and
Public Utilities on the land register was issued (ZARZĄDZENIE, 1969), which repealed
both departmental instructions. The cadastral data of individual plots included (§8):
the location, boundaries, area, type of land, class of agricultural land or other
classified land, identification of an owner or a holder. The concept of a parcel was
defined as follows: “a parcel is an area of land limited by land which is the subject of a
separate possession” (§7 section 1). Basing on the person of a holder led to a lack of
consistency between cadastral data of a parcel and its legal status. The register
demonstrated not only the person of an owner, but also the person who actually held
the land. Pursuant to §24 section 2 (ZARZĄDZENIE, 1969), a person who presented a
copy of the land and mortgage register, a notarial deed, a court judgment or a final
administrative decision was entered as the owner. The procedure of entering the
owner was based solely on the submitted documents. An autonomous possessor (§24
section 2, clause 2a) was also entered as the holder, as well as the person who
submitted the above-mentioned documents, but there was a discrepancy between
their content and the actual status, or when the documents were questioned by the
persons concerned (§24 section 2, clause 2b). That situation required the actual
possession of the land to be confirmed by a government administrative authority.
Thus, the provisions of Ordinance allowed questioning the documents confirming the
right of ownership, and thus the law itself.
This Ordinance was the basic secondary legislation act used in establishing the land
register in Poland. It was used until the end of 1996, which is more than six years
after the entry into force of the Geodetic and Cartographic Law (USTAWA, 1989). The
main problem of this Ordinance was limiting the cadastral data only to the land. No
form of recording data on buildings was provided. The second problem was the fact
that the provisions did not provide for carrying out an investigation regarding the
person of the owner – the owner was entered on the basis of the documents which he
or she provided or, if they failed to do so, they were entered as the land holder.
On 1 July 1989, the Act of 17 May 1989 - Geodetic and Cartographic Law - came into
force (USTAWA, 1989), which regulates the issues of land and buildings in Chapter 4.
The data contained in the register of land and buildings are set out in Article 20 of
Geodetic and Cartographic Law. At the time when the Act came into force, for land
these were (art. 20 section 1): the location, boundaries, surface area (instead of the
previous area), the type of arable land and soil classes and, additionally, designation
of land and mortgage registers or a set of documents, and for buildings - location,
purpose, utility function and general technical specification. Moreover, pursuant to
art. 20 section 2 of the Act, the owner was entered into the register, and in the case of
state and municipal land - other natural and legal persons, who were in possession of
the land, and their place of residence or registered seat. Therefore, instead of the
holders, the register included the owners (co-owners) of land and buildings,
perpetual users (co-users) and autonomous possessors (co-possessors). The
28
cadastral data also included information about the plot being entered in the register
of monuments (art. 20 section 2 clause 3).
On 17 December 1996, the Regulation of Ministers of Spatial Planning and
Construction and of Agriculture and Food of 17 December 1996 on the register of
land and buildings was issued (ROZPORZĄDZENIE, 1996), which came into force on 1
January 1997. Pursuant to §4 of this regulation, cadastral subjects were plots of land
and buildings. In accordance with §2 (ROZPORZĄDZENIE, 1996), the register
comprised data regarding the rights (including subjective rights). The descriptions
used data collected in the national geodetic and cartographic information center. The
information on individual rights resulted from the data contained in the land
registers and other documents.
Among the major changes in running of the cadastre, introduced by the Regulation on
the register of land and buildings of 1996 (ROZPORZĄDZENIE, 1996), is identifying
the owner, and not the holder, as the person to be entered in the register of land and
buildings. The second major change is the definition of a record parcel, which refers
to the definition of real estate contained in article 46 of the Civil Code (USTAWA,
1964), providing bases to synchronize the cadastre with the land and mortgage
register. This Regulation was in force until 30 September 1999. It was repealed by the
Act of 13 October 1998 – the provisions introducing the acts on reforming public
administration (USTAWA, 1998).
2.2. Current regulations on determining the cadastral data
Effective since 2 June 2001, the new regulation on the register of land and buildings
(ROZPORZĄDZENIE, 2001) provided the bases for a comprehensive modernization of
land records and for creating a register of buildings and premises from scratch. On 31
December 2013, a significant amendment to this regulation came into force. The
changes are designed to provide a possibility to supply data from the cadastre to the
database of topographic objects set forth in art. 4 section 1b of Geodetic and
Cartographic Law [USTAWA, 1989] and to create a computer database of the
cadastre. However, to make it possible, it is necessary to carry out a comprehensive
modernization of land records in Poland.
The following shall be subjects to registration (§10 ROZPORZĄDZENIE, 2001):
 figures and descriptive data regarding land and buildings as well as premises,
 data on real estate owners and their respective shareholdings in ownership rights,
as well as the dates determining acquisition and loss of that right.
In the case of land with an unknown owner, the cadastre contains the data of the
persons or entities who hold this land in autonomous possession. Pursuant to §36
(ROZPORZĄDZENIE, 2001) the course of boundaries of record parcels is entered into
the register based on geodetic documentation accepted into the national geodetic and
cartographic information database, and prepared:
 in delimitation proceedings,
 for the purpose of real estate subdivision,
 in land consolidation and land exchange proceedings,
 in proceedings for consolidation and subdivision of real estate,
 for the purpose of judicial or administrative proceedings, and then used for the
final judgment or final administrative decision,
29
 in establishing the real estate cadastre and the register of land and buildings, under
the previous legislation,
 by the Border Guard if this documentation defines the course of State boundaries
with the accuracy appropriate for the records,
 as a result of geodetic detailed measurement of the existing or restored boundary
markers or determined boundary points.
In the absence of the above-mentioned documentation, or if the data contained
therein are not reliable, the data on the course of the boundaries of record parcels are
captured by means of field surveying or geodetic photogrammetric measurements
preceded by determining the course of these boundaries. Such a determination of the
course of record parcel boundaries, including the position of the boundary points
defining them, may be based on aerial imaging, satellite imaging or on an
orthophotomap, provided that this imaging or orthophotomap have a resolution
ensuring visualization of the field details that may be relevant in determining the
course of these boundaries.
In accordance with §61 (ROZPORZĄDZENIE, 2001), a numerical description of
cadastral boundaries, cadastral units and record parcels shall be made using
respective sets of boundary points whose position relative to the 1st class basic
geodetic control has been determined based on geodetic detailed measurements with
mean errors not exceeding 0.30 m, in the way ensuring the mapping of the position
and shape of these spatial objects and the mutual relationship between them. The
position of boundary points used for a numerical description of objects is defined in
the Cartesian coordinate system 2000.
However, when carrying out modernization of the cadastre, whose one of the aims is
to convert the cadastral map stored in the analog form to the electronic form, in case
of inability to determine the position of boundary points with the above accuracy, the
numerical description of record parcel boundaries shall be made using boundary
points, whose position relative to the 1st class basic geodetic control will be
determined based on the existing materials of the national geodetic and cartographic
information database, with the mean errors not exceeding:
 3.0 m - where boundary points belong to record parcels located in rural areas,
outside dense development,
 0.60 m - where boundary points belong to record parcels located in urban areas
and in dense development in rural areas.
The data defining the position of the boundary points with the mean error greater
than 0.30 m relative to the 1st class basic geodetic control, which were used for the
numerical description of the record parcel boundaries, shall be replaced in the
process of records modernization with the more accurate data available to the
authority, even if the new, more accurate data are still determining the position of the
boundary points with the mean error greater than 0.30 m, and it is not possible to
determine the position of these points with a mean error not exceeding 0.30 m.
2.3. Procedures for determining the course of real estate boundaries
As it was mentioned above, the course of boundaries of record parcel is included in

the records based on appropriate surveying documentation accepted into the
national geodetic and cartographic information database. If there is no such
30
documentation, or if the data contained therein are unreliable or do not meet the
requirements of technical standards, the data on the course of boundaries of record
parcels is captured as a result of field surveying or photogrammetric measurements
preceded by determining the course of these boundaries in the field. The course of
boundaries of record parcels is determined by a qualified surveyor in the presence of
the parties involved. Pursuant to §39 (ROZPORZĄDZENIE, 2001) the course of
boundaries of record parcels, including the position of the boundary points defining
them, is determined by a surveyor on the basis of joint statements of the owners or
perpetual users of these parcels, or people who hold these parcels subject to
autonomous possession, confirmed by their joint statement made to the protocol
determining the course of boundaries of record parcels. In the case where liable
parties fail to submit a joint statement to the protocol determining the course of
boundaries of record parcels, the course of boundaries of record parcels, including
the position of the boundary points defining them, shall be determined by a surveyor
on the basis of the last peaceful possession, if that possession is not inconsistent with
the information contained in available documents defining the legal status of the land
within the boundaries of these parcels. However, in the case where peaceful
possession can not be identified or it is inconsistent with the information contained in
the available documents defining the legal status of land, the course of boundaries of
the parcels covering this land, including the position of the boundary points defining
them, are determined by a surveyor after examining the position of boundary
markers and traces, and after conducting an analysis of all available documents
containing information that is relevant in this regard, including statements of the
parties involved and the witnesses.
Fixed boundary points are marked by a surveyor on the ground in a way that enables
their measurement. Permanent stabilization of these points can be made only upon
the initiative and at the expense of the parties involved, and it may relate only to the
points determined based on a joint statement of the parties. The findings regarding
the course of boundaries of record parcels shall be recorded by a surveyor in the
protocol. An integral part of the protocol determining the course of boundaries of
record parcels are field sketches drawn up by a surveyor, which in addition to the
standard content shall include:
 the position of boundary points relative to field details situated on the boundaries
being determined or in their immediate vicinity, such as buildings, fences, poles
and field margins,
 the numbers of record parcels,
 the signatures of persons involved in the activities undertaken to determine the
course of boundaries, or annotation made by a surveyor on their refusal to sign.
When the process of determining boundaries of record parcels is based on aerial
images, satellite images or an orthophotomap, information and signatures are
contents of copies of this imaging or this orthophotomap.
Another procedure aimed to determine the course of real estate boundaries is
delimitation. Delimitation of a real property is defined in the Geodetic and
Cartographic Law as procedure that is meant to determine the course of real property
boundaries by identifying the position of boundary points and lines, fixing these
points on the ground using boundary markers and preparing relevant documents. If
necessary, all or some of the boundaries of the specified real estate with the adjacent
31
real estate or other land shall be subject to delimitation. This procedure is performed
by the mayor, however, the activities to determine the course of boundaries shall be
performed by a qualified surveyor authorized by the mayor. In accordance with
article 31 of Geodetic and Cartographic Law (USTAWA, 1989), when determining the
course of boundaries, boundary markers and traces, maps and other documents, as
well as geodetic control points are taken into account. Administrative proceedings on
the delimitation of a real property are concluded with the issuance of the final
decision. Geodetic documentation of this procedure accepted into the national
geodetic and cartographic information database forms the basis for restoration of this
boundary in case it is obliterated on the ground.
In a situation where the real estate boundary has not been determined in delimitation
administrative proceedings or if the party dissatisfied with its course appeals the
decision, the matter shall be referred to the court. The court is competent to carry out
the delimitation, also in a situation where there is a case pending on the property
ownership rights, and determining the course of the boundaries is necessary to
resolve the matter. In such a case, the court also includes the decision on the real
estate delimitation in its judgment.
The court hears the cases on real estate delimitation in non-litigious proceedings. The
criteria for determining the boundary course by the court are different from those in
administrative proceedings. Pursuant to art. 153 of the Civil Code (USTAWA, 1964), if
the boundaries of the land have become contentious, and the legal status can not be
established, the court shall determine the boundaries according to the last peaceful
possession. However, in the case where peaceful possession can not be identified and
the delimitation proceedings did not lead to a settlement between the parties
concerned, the court establishes the boundaries taking into account all the
circumstances; at the same time the court may grant a corresponding cash
compensation to one of the owners. The proceedings in the field are performed by a
surveyor - court expert, as ordered by the court. The surveyor also prepares
surveying documentation, which is accepted into the national geodetic and
cartographic information database before the delimitation judgment has been issued
by the court. Enforcement of judgments on real estate delimitation is also carried out
in the presence of a surveyor.
2.4. Aerial images - interpretation
Aerial images are a fully objective archival material documenting land development.
Their usefulness to determine real estate boundaries at a specific point in time (on
the day of taking them) depends on many factors, though. Aerial photographs do not
offer a possibility to identify the actual course of a real estate boundary, it is however
possible to recognize the location of elements such as fences, hedges, field margins,
buildings and roads. These elements can show what the actual status of land use was
at a given time. The most probable course of boundaries will be represented by field
margins, fences and hedges. A possibility of a good interpretation of their position is
limited by factors such as image resolution, image quality, and view obstructed by
vegetation.
At present, photographs are processed digitally. In the case of photographs originally
taken on film, it requires scanning them. This is done with the use of special
32
photogrammetric scanners to ensure high geometric scanning accuracy of about 12
m (KURCZYŃSKI, 2014). The required geometric resolution should be adjusted to
the resolution of the film and, according to (KURCZYŃSKI, 2014), the scanning pixel
should not be greater than 10 m. In most scanners for aerial photographs, the
detector size ranges from 6 do 8 m. However, in practice, the scanning with a
resolution so high is rarely performed, as digital images then have large sizes, and
photographic emulsion particle size is visible in the image. Most frequently, the
scanning pixel of 12 do 16 m is used. In Poland, photographs were mostly taken at
the scales from 1: 8000 to 1:30000. Having taken into account the size of the scanning
pixel, the ground sampling distance is therefore from 10 cm to 45 cm. Sometimes the
photographs were taken at larger scales, even about 1: 3000. The ground sampling
distance is then about 5 cm.
The quality of archival photographs depends on the used film and on the
photochemical processing. In Poland, image quality was variable. There are examples
of photogrammetric flights with a large scale and high-quality images, and a few years
later, at a similar scale, the images are definitely of an inferior quality.
In Poland, since the beginning of the twenty-first century, more and more aerial
photographs are taken using photogrammetric digital cameras. Because of their
higher radiometric resolution, the image readability is clearly better than that of the
photographs on film. In recent years in Poland photographs were taken with ground
sampling distance of about 20 cm. Only in the areas of large cities photographs were
taken with a higher resolution, i.e. with ground sampling distance of about 10 cm.
While the resolution of images and their readability affect the quality of
interpretation and the accuracy of measurement, the crowns of the trees and larger
shrubs often limit the ability to recognize the position of general landscaping
elements. Fig. 22 shows the same piece of land photographed in April (a) and August
(b). In the places pointed by the arrows in the photo taken in April, the fence is clearly
visible, because the trees are defoliated. In the image taken in August (a few decades
earlier, in fact), it is virtually impossible to recognize the course of the fence..
Fig. 22. The same pice of land in the photo taken: a) in April, b) in August
Resource: CODGiK.
33
The ability to identify the location of fences or other landscaping elements also
depends on the time of day at which the image was taken. An important element
associated with the objects protruding above the surface of the land are their
shadows. The cast shadow (the shadow of an object located on the adjacent object)
could be either a positive or a negative effect. A major problem in the interpretation
of aerial photographs is the fact that the objects there are visible from above, which
often makes them difficult to recognize or they are hardly visible, especially if these
are vertical objects with a small cross section (e.g. poles, fence posts). An appropriate
direction of sunlight can result in the shadows being visible, which are an oblique
projection of an object. Fig. 23 and Fig. 24 present a shadow of a fence visible on the
ground. It allows for an easy recognition that there is a fence at all, and how it is
constructed.
However, the cast shadow may also be a great hindrance for interpretation. Large
objects (buildings, large shrubs and trees) often cast a shadow which is so strong that
it masks the photo image of the area. An example of such a situation has been
presented in Fig. 23. Both photographs were taken on April 26, but the “a” photo in
2009, while the “b” photo in 2012. Arrows point to the fence, which in “b” is
completely invisible, as it is located in the strong shadow. In “a”, which was taken
about 1.5 hours later than “b”, the shadow of tall shrubs and trees is cast next to the
fence, and therefore the fence is clearly visible.
Fig. 23. The same piece of land in photographs taken on April 26: a) in 2009, at about 1.5 hours
later than b, b) in 2012.
Resource: CODGiK.
Recognition of a fence or hedge is relatively simple. The measurement of land

boundaries is more difficult when they are not marked with any fence. The
identification of the exact position of the line representing the boundary between
adjacent lands may pose a problem. If adjacent parcels are covered with the same
crops (e.g. wheat), quite frequently the boundary between them is unrecognizable in
the images. If agricultural land is adjacent to a dirt road, then this boundary is clearly
34
visible, but it is uncertain whether it is the actual boundary of the plot. Fig. 24
illustrates a boundary between a dirt road and agricultural land (field, pasture,
meadow). Often, the type of land use on the same parcel varies over time, and
consequently, depending on the method of cultivation, the boundary between
agricultural land and the road is moved (Fig. 24. a - 1998 b - 1998).
Fig. 24. the same piece of land in images taken in different years – change to the appearance of
the boundary between the dirt road and agricultural land.
Resource: CODGiK.
It also happens that the actual boundary of agricultural land is not clearly visible, and
it is difficult to identify in a single image. Stereoscopic observation of images proves
to be helpful then. It allows recognizing even small slopes, which are often located on
the border between the road and the plot. Fig. 25 presents a situation where a part of
the plot (a) is planted with grass (grassland). This makes it impossible to identify a
boundary between this land and the dirt road in a single image (Figure 25, left). Using
stereoscopic observation, a boundary of the slope has been drawn, which has been
established as the boundary between grassland and the road. The situation in the part
marked with the letter b in Fig. 25 helped in this identification. The plot in this place
35
was plowed. The boundary of the plowed fragment coincided with the slope
boundary. Without the stereoscopic observation it would not be possible to recognize
it.
Fig. 25. The upper and lower edge of the slope: a) unrecognizable in a single photograph, b) being
at the same time the boundary of the agricultural land.
Resource: CODGiK.
2.5. Capture of spatial data from aerial photographs
Capture of accurate spatial data from a single image is not possible. It is required to
have two images (the stereogram), which include photographs of the area of interest.
For the measurements on stereograms, a software called digital stereoplotter or
photogrammetric workstation is used.
The measurement itself must be preceded by a procedure of fitting the stereogram
into the spatial reference system. For this purpose, it is also necessary to know the
internal orientation elements of the camera (principal distance, principal point
position), which the photos were taken with (KURCZYŃSKI, 2014). Airborne
photogrammetric cameras are subject to the calibration process, during which these
elements are determined. Photogrammetric camera design ensures the constancy of
these elements. In order to achieve high accuracy of fitting these image stereograms
into the spatial reference system, photographs taken with professional airborne
photogrammetric cameras should be used. Archival aerial photographs in CODGiK
resources were taken in the vast majority with such cameras. In the analog
photogrammetric cameras, basic information about the values of the interior
orientation elements were photographed on the edge. Thus, when using digital copies
of these images, the information can be read. In the images taken with digital
cameras, the information about the values of the interior orientation elements are not
36
photographed. They are recorded in an additional calibration report, which must be
purchased along with the photographs.
The placement and orientation of a camera in space requires knowledge of exterior
orientation elements of images. At present, these elements can be determined at the
time of taking photographs, but not always with satisfactory accuracy. In order to
obtain the maximum possible accuracy of the camera orientation in space, the
photographed elements with known spatial coordinates - the so-called ground
control points - should be additionally used. In the case of using archival images,
especially from analog cameras, the determination of the spatial orientation of the
camera must be based solely on the use of ground control points.
Ground control points could be terrain details identified in the photographs. It is
important that their identification, both in images and in the field, is unambiguous. It
is also important for the position of a given point in the field to remain unchanged
between the time of obtaining its spatial coordinates and the time when the
photograph was taken. If the time interval is small, various details may serve as
control points. These may include: building corners, corners and posts of fences,
electricity and telecommunications poles, corners of lawns, sidewalks, bridges and
culverts, sewer manholes and drains. Any rebuilding or relocation of these elements
leaves traces which remain visible in the field for some time. In contrast, it is different
when the time interval between taking photographs and determining spatial
coordinates of ground control points takes a few years, or even decades. A small
change in the position of the majority of these elements will generally be impossible
to recognize. Quite frequently, aerial photographs do not offer a possibility to
accurately recognize the small elements such as poles, fences and culverts. Only
rebuilding of a building is easy to notice, as in the photographs they take up a
relatively large space. It must always be taken into account, though, that a change in
the position of a detail may not be noticed, and this detail will be used as ground
control point. Therefore, when studying archival images, the number of control points
should be significantly higher than when new images are analyzed. The minimum
number of control points for the orientation of the stereogram of aerial images is
three. Five to six control points are used as a rule. With archival images, there should
be more control points, even a dozen or so. Through the analysis of deviations at
stereogram orientation, this allows for the detection of ground control points that
have been displaced.
During a standard study of a stereogram of aerial images, control points are
distributed over the entire surface of the stereogram. Extrapolation outside the area
covered by control points carries the risk of accumulation of errors. Verification of
spatial data of real estate on archival photographs is often performed only in a small
area. Then control points can be distributed only around the study area. An example
of such a study has been given in Fig. 26.
37
Fig. 26. The upper and lower edge of the slope: a) unrecognizable in a single photograph, b) being
at the same time the boundary of the agricultural land.
Resource: elaborated by WRÓBEL.
Spatial coordinates of control points are best captured by performing surveys in the
field. Unfortunately, this involves high costs. Besides, when studying archival images,
it often turns out that the objects visible in the photographs are no longer in the field
or they have been rebuilt. Thus, to capture spatial coordinates of control points, the
existing large-scale maps are used. This solution is definitely cheaper, and sometimes
the fact that these maps have been out of date for many years enables one to find
objects that no longer exist in the field.
Stereogram of aerial images allows to capture spatial data with an accuracy
depending on the scale of images (for photographs from the analog camera) or on
ground sampling distance (for images from a digital camera). It is assumed that clear
elements with good visibility can be measured in the photographs with an accuracy of
about one pixel. Linear elements such as fences or land boundaries, which are
theoretically clearly visible in the photograph, can be measured with an accuracy
somewhat better than one pixel. However, in practice, it is often not possible due to
shadows deteriorating stereoscopic observation.
The accuracy of determining the spatial coordinates of the measured points depends
on the accuracy of their measurement on the photographs and on the accuracy of
determining the spatial orientation of the images. The latter one can be measured by
the root mean square error (RMS) of fitting the image stereogram into the control
points (absolute orientation in the digital stereoplotter).
Fig. 27 illustrates a graph of the RMS of detailed fitting of ground control points
obtained with absolute orientation of the stereograms with different ground
sampling distance. The X, Y coordinates of ground control points were captured from
large-scale maps. Heights of the control points, if they were not on the large-scale
maps, were captured from a topographic map. The accuracy of the height of control
38
points is less important, because when verifying real estate boundaries, usually only
the X, Y coordinates of the analyzed landscaping elements are required.
Generally, for most of the analyzed cases, the RMS of detailed fitting of ground control
points was similar to ground sampling distance. In five cases (in the Figure marked as
a circle instead of a disk), RMS was much larger than ground sampling distance. All of
these cases concerned a single piece of land, for which the large-scale map was
probably inaccurate. Regardless of the quality of images, fitting the stereogram into
the control points was significantly less accurate than for other areas.
Fig. 27. Dependence of the root mean square error (RMS) of the stereogram absolute orientation
on ground sampling distance.
Resource: elaborated by WRÓBEL.
2.6. Conclusions
Archival aerial images are useful for determining real estate boundaries on the day of
taking them. Many factors can affect the accuracy of the results. The scale of
photographs (ground sampling distance), image quality, time of year and time of day
in which the photographs were taken are essential. The time of year determines a
possibility to recognize landscaping elements located under the trees, and the time of
day determines a possibility to recognize objects located in large shade. The scale of
the photograph, and thus the ground sampling distance, affect both the ability to
identify analyzed objects and the accuracy of determining their spatial position.
The accuracy of determining spatial position of the measured points is influenced by
the accuracy of measurement on the photographs, and additionally by the accuracy of
fitting the image stereogram into the spatial reference system. The accuracy of fitting
the stereogram depends both on the ground sampling distance and on the accuracy of
determining the spatial coordinates of ground control points used for the fitting. The
conducted experiments have proved that by capturing control points coordinates
from the existing large-scale maps, sufficient accuracy of fitting the stereogram can be
39
achieved at the level of the ground sampling distance. However, this is possible only if
the maps which are being used are not subject to some systematic errors. If there is a
suspicion that the maps were prepared inaccurately, it is necessary to determine the
coordinates of the ground control points by direct surveying. The direct
measurement of ground control points (GRABEK, 2014) allows for a significant
increase in the accuracy of the absolute orientation of the stereogram in the digital
stereoplotter. Therefore, if it is possible, a direct measurement of ground control
points is recommended. In this way, however, it is not possible to measure the objects
which are already non-existent in the field, and these often have to be used to fit
images which are a few or a few dozen years old. Then, large-scale outdated maps
remain to be used.
40
3. DETERMINATION OF COORDINATES OF OBJECTS USING
KINEMATIC PHOTOGRAMMETRY FORMULAS WITH
DIGITAL CAMERA AND GPS-RECEIVER
Modern development of photogrammetry requires a radically new approach
to determine coordinates of points of objects. In many fields of science and technics
phenomenon are considered in dynamic but not static conditions. Therefore actual
problem of photogrammetry is to determine not only position of object in space but
in time as well. The solution of this problem became possible owing to modern means
used in geodetic and photogrammetric production. This is stipulated first of all due to
application of GPS systems, combined with digital camera. This give possibility to
include in the corresponding equations a number of additional relevant values
obtained regardless of survey. However, to obtain accurate results in kinematic mode
it is necessary to use appropriate mathematical processing.
Fundamentals on the strategic direction of kinematic photogrammetry are
presented in (DOROZHYNS'KYJ, 2002). Study highlighted radical changes related to
application of new technologies in geodesy and photogrammetry, first of all, it
concerns GPS receivers that have considerably changed both technologies of
photogrammetric data processing and possibilities of wider application of
photogrammetric methods in monitoring processes.
The transition from analogue photographs to digital systems also had significantly
changed the approaches and opinions in modern photogrammetry. Actually, the
analysis of these two trends (GPS and digital photogrammetry) allowed to form a
concept of development of new direction, called navigational digital photogrammetry.
Development of the concept of navigational digital photogrammetry and
monitoring of kinematic processes of researched objects changes both the
mathematical models of photogrammetric processing of images and the
possibilities of quantitative and qualitative assessment of these changes (GEL'MAN,
NYKYTYN, 2004).
On the other hand, range of applications of descriptive geometry theoretical
methods to engineering problems, in particular, for study of deformations of
structures, landslide processes, etc. has been broadened.
Thus, the study aims to develop functional relations of linear kinematic projecting
applied to different combinations of straight types of motion of image elements
and to use the method to determine coordinates of object points on their
representation on the digital image. It should be noted that we consider the case
when the camera is moving.
Traditionally, development of methods of projection representations in
descriptive geometry proceeds by summarizing dimensions, metrics and structure of
operational spaces (areas of definitions and the values of representations) and
projection apparatus and representation operators. They become feasible and
effective due to using methods and techniques of projective and linear geometry and
introduction of kinematic categories to descriptive geometry. Such diverse
development had led to statement of the problem of creation of a unified theory of
projection representations which would combine all existing methods. The least
studied component of this theory is its kinematic aspect. However, this is not about
41
kinematic interpretation of existing static means but about the really kinematic
descriptive geometry, where the process of representation has time characteristics
and is determined by the laws of motion of all types (GLOTOV et al., 1996)
Let’s establish relations between kinematic modes of point motion (speed, acceleration),
projection center motion and order of ray projection surface. Let’s derive and study
equations of primary and secondary projections of trajectories of prototype А when:
uniform motion of А and uniform motion of center Р; uniform motion of A and P in a
special position when A0 ( x01 , O, z01 ), P0 ( x02 , y02 , O ) (Fig. 28).
Fig. 28. Trajectory of a prototype of point A.
Resource: elaborated by HLOTOV.
Based on the derived equations it is necessary to determine the types of trajectories of

prototype. Let f 1 ( y  y1 ; z  k z  x  bz ) i f 2 ( x  x2 ; z  z 2 ) are rectilinear trajectories
according to prototype of (points) А і centre of projection Р that begin to move at
the same time: point А - uniformly with velocity V0, and the center Р - equable
changing with an initial velocity V0 and acceleration а.
The given conditions of movement are obtained from linear congruence of s(f1;f2)
with focal figures f1 і f2. Radial surface   Ai, Pі is considered as a uninterrupted set of
42
projection rays {s}(s i  Aі; Pі), corresponding to the instantaneous position of the
points in А і Р. For an arbitrary point S ij ( x, y, z)   , lying on the ray si  Ai , P at
time moment t we have the following relations:
 a t2 
Y  V0  t  

 2  y  y 01
(6)
x  V0  t  cos k z x  x01  V0  t  cos k z
z  V0  t  sin k z z  z 02
 (7)
x  V0  t  cos k z x02  x
x  x 02  x01 , y  y 02  y01 , z  z02  z01 . 

After substituting of the value:,
z  x 02  x   x  z  z 02 
t    x, z  
V0  z  z 02   cos arctgk z  x02  x   sin arctgk z 
into (3.1) we obtain the equation:

  a 2 
y  V0    x, z   2    x, z     x  x01  V0    x, z   cos arctgkz  
  
 x  V0   x, z   cos arctgk z    y  y 01   0, (8)

defining surface of the third order.
If the prototype center A and projection center P are doing uniform motion in a
straight line with speeds accordingly V1 і V2 , so the surface is a quadric surface
(unicell hyperboloid, or in the extreme case - hyperbolic paraboloid), then equation
(8) takes the following form:
y  V2    x, z   x  x 01  V1   x, z   cos arctgk z  
 x  V1    x, z   cos arctgk z    y  y 01   0, (9)
x  z  z0   z  x02  x 
where, -  x, z   .
V1  z  z02   cos arctgkz  x02  x   sin arctgkz 
In general, when the path, which А or Р pass during time t, is equal V0  t  a  t n , so

2
the surface σ has order n+1. Now we introduce for consideration plane of
projections π with fixed coordinate system η, ζ in it. Define initial position of
representation elements A0  x01 , y 01 , z 01 , P0  x02 , y 02 , z 02  і  0 x  0,  y,   z  and
straight trajectories of points А і Р respectively. Assume also that in the process of
movement of plane π the x-axis is its directional vector.
43
Let the representation elements begin simultaneous movement: the prototype А
and plane π - uniform motion with velocity V0, а centre Р equable changing with an
initial velocity V0 and acceleration а. Then in each moment of time ti projection f1' of
straight f1 on the plane π in its position  i  x  V0  t i  , that corresponds to this time
moment, is the intersection of the radial surface σ with plane πi..
Parametric equations y = (t), z = z (t) of infinite sets of secondary projections:
1    
A1 ,1 A2 ...; 2 A1 , 2 A2 ...;
(re-projecting from a moving plane π on the immovable planes of projections xy and

xz) of point
A in its successive positions on f1 are parabolic branches of the third order of type:
y  d1  x 3  d 2  x 2  d 3  x  d 4 ,
According to Newton classification, namely:
 y 02  l 2  y    x01  m1  V0  t    x02  m2  V0  t    y 01  l1  y , (10)
z  z 02  n 2   x  m1  m2   x 02  m2  V0  t   z  n1  n2 , (11)
where
x  x1  x2 ; z  z1  z 2 ,
 a t2 
m1  V0  t  cos  1  cos 1 ; m2  V0  t    cos  2  cos  2 ,
 2 
 a t2 
n1  V0  t  cos  1  sin 1 ; n2  V0  t    cos  2  tg 2  cos  2 ,
 2 
 a t2 
l1  V0  t  cos  1  tg1  cos 1 ; l 2   V0  t    cos  2  tg 2  cos  2 ,
 2 
here
 і  arctgkіy ;  і  arctgkіz ;  i  arctgtg і  cos  і , i=1,2.
When uniform motion of all representation elements, equations (10) and (11) define
the quadratic parabolas.
In very specialized position of trajectories of prototype A0 x01 ,0, z 01  and center
P0 x02 , y02 ,0 :
f1  y  y1 , x  0, f 2  y  y2 , z  0 ,
and uniform motion of all elements with the same unit velocity (V = 1) equation (10),
(11) - y  yt  and z  zt  are as follows:
1
y   y 2    x1  , (12)
x
44
1
z   z1      x 2   , (13)
x
where ν - the coefficient of linear dimension.

In the canonical form equiations (12), (13) will have the following form
2
 1   1 1 2 
  2   x1  y 2   x   y  x   4  x1  2 y   x2  y 2 , (14)
    
2
 1   1 1 2 
  2  x2  z1   x   z  x   4  x 2  z1   x2  z1 . (15)
    
Thus, in this case the projection of trajectory of prototype A - congruous quadratic
1
parabola (14), (15) with the parameter  x .
2
3.1. Determination of coordinates of digital photogrammetric survey points
As noted above, application digital cameras and GPS for the implementation of
aerial or ground-based photogrammetric survey when camera is moving (e.g. taking
photography from water carrier, road vehicle of aircraft), makes it possible to apply
the principle of linear operators to determine coordinates of object points. Indeed,
the combination of two independent devices that are used to perform the same task
should lead to obtaining qualitatively new results. In terms of mathematical
processing, the data obtained by GPS (coordinates of projection centres and time of
taking photography) make it possible to increase the number of equations, which in
turn should improve the accuracy of determination of object point coordinates.
Then similarly to the structure of presented algorithm we obtain the formulas for
digital ground based photogrammetric survey (similar formula can be obtained for
aerial survey) (GLOTOV, PUL'KEVYCH, 2006).
Input formula for first image (DOROZHYNS'KYJ, 2002):
x1 a x  a f  a13 z1
X  X S1  (Y  YS1 )  X S1  Y 11 1 12
y1 b11 x1  b12 f  b13 z1
, (16)
z1 c11 x1  c12 f  c13 z1
Z  Z S1  (Y  YS1 )  Z S1  Y
y1 b11 x1  b12 f  b13 z1
for second image:
x2 a x  a 22 f  a 23 z 2
X  X S2  (Y  YS2 )  X S2  Y 21 2
y2  b21 x 2  b22 f  b23 z 2
. (17)
z 2 c 21 x2  c 22 f  c 23 z 2
Z  Z S2  (Y  YS2 )  Z S2  Y
y 2 b21 x2  b22 f  b23 z 2
45
Between the exposures of first and the second images there was movement, so
formulas (16) and (17), while taking into account parallelism of systems x', y', z', and
x, y, z can be written as follows:
xt1
X  X S 1  Y  YS 1  ,
f
z
Z  Z S 1  Y  YS 1  t1 ,
f
x
X  X S 2  Y  YS 2  t 2 ,
f
z (18)
Z  Z S 2  Y  YS 2  t 2 ,
f
x x
X S 2  X S 1  Y  YS 2  t 2  Y  YS1  t1 
f f
x x x x
 B X  Y t 2  YS 2 t 2  Y t1  YS 1 t1 ,
f f f f
1   at  
Y  ft  V0   cos   YS 1 xt1  YS 2 xt 2 
pt   2 
.
finally we get:
at 2 at 2
(V0 t  ) cos cos  f (V0 t  ) sin  f
2 Y 2 .
Y ;
pt zt1  z t2
Thus, the introduction of three additional parameters (speed of projection centre

movement, time and acceleration) allows the introduction of two additional
equations.
Therefore, the obtained formulas (18) provide an opportunity to determine spatial
coordinates of image points.
Now, we define the a priori estimation of accuracy by using the formula (18). Since
the inclination angles of image do not exceed 3°, the formula can be simplified using
approximate values of direction cosines (LOBANOV, 1968). We thus will consider
only terms of the first order of smallness. In addition, the survey is made from water
surface, then z1=z2=z. Partial derivatives are:
46
X xt1 2
Y

f
, X
 1
 Y  YS1  1  x12 ,
 f 
Y Y  YS2
 z,
 2 pt X
1
 Y  YS1   x12z ,
Y ft 2  f 
 , X z
a 2 pt  YS1  Y  ,
 1 f
Z zt
 1 , X xt
Y f
Y

 YS 2  Y 2 ,
f

Z x z
 Y  Y S 1  1 21 , 2
 1 f X
 2
 Y  YS 2  1  x22 ,
 f 
Z  z 2 

 Y  Y S1  1  ,
1
 f 2  X
 2
 Y  YS 2   x22z ,
Z x1  f 

 Y  Y S1  , X z
 1 f
 2
 YS 2  Y
f
,
Z z t2
 ,
Y f
YS  Y  2 
Z
 x z
 Y  Y S 2 2 22 , 
Y
 1  f  x1 ,
 1 pt  f 
 2 f 
Z  z 2  Y Y  YS1  x1 z 

 Y  YS2  1  , 1
 
p t  f 
,
 2
 f 2 
Y (Y  YS1 ) z
Z x  ,
 Y  Y S 2  2 .  1 pt
 2 f
Y  YS 2  2 
Y
  f  x2 ,
 2 pt  f 
 
Y Y  YS 2  x2 z 
  
 2 pt  f 
Taking these values of the derivatives into consideration, we obtain averaged square
errors of spatial coordinates of terrain points:
mX 
x t21
mY2 
Y  YS 2  

1
f 
2
x12  2
m 
x12 z12


m21  z12 m21  
2 2   f 
1 2

f f f
 
 2 
1 2  x12  2 x2z 2 
 xt21 mY2  Ymax   f  m1  1 21 m21  z12 m21  
f f  f 
  
2

1 2 
xt21 mY2  Ymax

f 
x12  z 2 f 2  x12
 1
 m 2
,
  f   1
f f2
 
47
mZ 
z t21
m Y2 
Y  YS 2  x12 z12 m 2
1
 z2
 f  1
 2 
 m  x12 m 2  
2 2  f 2 1   1 1 
f f   f  
 2 2 2 
z t21 2
Ymax  x1 z1 2  x12  2 
 m Y2  2  2 m 1  f  m1  z12 m 21  
f 2
f  f  f  
   
2

1
z t21 mY2 2
 Ymax


f 
x12  2 2 2
  z1 f  x1  m 2
,
f   f  f2  1
  
2 2
   
1  2 
 BY xt  f V0 t  at  cos    x12  2 x12 z12 2 2 2 
mY      f  m1  m 1  z m1  
p t2 

2
 2  
  f  f2 
 
2 2
   
 at 2   x22  2 x 22 z 22 2 2 2  f 2 t 4 pt2
=
  BY xt1  f V0 t   cos     f   2m  m 2  z m2   cos 2  ma2
 2   f f 2
 4
     
2 2
 
1   2  
 B x  f V t  at  cos    x12  x12  f 2 2  2
   f   z1 m1 
pt2  Y t2  0 2   f  f2 
     
2 2
 
  at 2    x 22  x2  f 2
 f 2 t 4 pt2
  BY xt1  f V0 t   cos     f   2 2 z 22 m22  cos 2  ma2 
 2   f   f  4
    
2 2
 
1   at 2    x 2  x2  f 2
 f 2t 4 p t2 cos 2  2
 2 BY xt  f V0t  cos  
  f   z 2 m  ma
p t2 
 2   f  f2  4
   
The above formulas differ from classical by presence of arguments of velocity

and acceleration. In our case, for experimental works we used digital camera
Olympus E20p: x  40 mm, z  30 mm, with equivalent focal length 86 mm, and
image processing was planned to be done using digital photogrammetric station
"Delta 2": mf = mx= mz = 0,005mm, mα=mω=mκ=2″. The distance to the photographed
object ranged from 50 to 200 meters. Speed of the boat was 1 m/s and acceleration
was 0.1 m/s2 . Angular elements of exterior orientation (EXO) were chosen according
to reasons of horizontal survey (3˚ to 5˚), as the survey complex was equipped with
appropriate level. Table 1 shows the results of averaged square errors (ASE) for the
above mentioned parameters.
48
Table 1. Values of ASE of spatial coordinates when changing distance
Y (m)
50 100 150 200
ASE
0,059 0,091 0,127 0,165
mY 0,082 0,123 0,172 0,223
mZ 0,048 0,084 0,124 0,162
Resource: elaborated by SIEJKA.
Analyzing the values of errors, it should be noted that they fully meet the
requirements of large-scale mapping.
3.2. Topographical plans of the Vernadsky Antarctic station
Analysing topokinematic works in Antarctica, it should be noted that Russia

shows most of the activity in this area. A large amount of cartographic works had
been performed for development of GIS on the territory of East Antarctica. It had
been developed 24 trapezoids of topographic maps in scale 1: 100000 with total area
- 34 thousand km2 and 57 sheets in scale 1: 200 000 with total area - 330 thousand
km2.
As the input cartographic material maps in scale 1: 100000 and 1: 200000 had been
used. They were created by topographic and geodetic enterprises of USSR and Russia
at different times based on aerial survey, which was carried out by Soviet Antarctic
expeditions. Before start of the work, the geographic names have been tested and
refined using the dictionary issued by Central Scientific Research Institute of Geodesy,
Aerial survey and Cartography (1989). Regarding to cartographic data of base
stations Bellingshausen, Novolazarevskaya, Mirniy, Progress and part of the station
Union there were used topographical plans in scale 1: 500 and 1: 2000, created in
1984-2000. Altogether 33 trapezoids with total area of 29 km2 had been made.
When creating a digital topographic base the ellipsoid WGS-84 and UTM map
projection were used. "Conventional signs for topographical and general geographic
maps of Antarctica" and cartographic base in format ArcGis 8.3 with the addition
Arcmap were used for representation of geographic features.
Thus, Russia today has a cartographic component for GIS of Eastern Antarctica, which
created for the activity zone of Soviet and Russian Antarctic expeditions.
Further work planning is offered in three stages:
 creation of unified catalogue of points of geodetic network, located in East
Antarctica on the WGS-84 ellipsoid, but also possibly in coordinate systems
and projections adopted in the project partner countries;
 creation of GIS using software ArcView or ArcInfo with special thematic layers
– it is the most complicated and long-term stage in the technological scheme
of GIS creation.
Chilean Institute of Military Geography and Antarctic Institute of Chile had created
topographic plans in scale 1: 5000 in digital form [26]. Topographical plans were
created using geodetic methods.
Thus, analysing the recent mapping works implemented in Antarctica, it should be
noted, firstly a tendency to large-scale mapping, and secondly peculiarities of
representation and application of these plans. Based on these tendencies, it should be
noted the relevance of this problem (GREKU, 2000).
49
Different methods of topographic survey can be used for creation of topographical
plans of Argentine archipelago and adjacent coast of the Antarctic Peninsula. We will do
basic analysis of main methods.
Geodetic methods, as it was already noted, are not sufficiently technological, have a
large amount of field works and in most cases are life threatening in terms of labour
safety.
Based on the above mentioned analysis and tasks it was proposed digital terrestrial
photogrammetric survey, which has significant advantages over classical methods.
The results presented in Table 1, show that for the distance 200-400 m to an object,
any focal length and using various digital cameras in all survey cases, errors of
coordinates do not exceed permissible values for plan in scale 1:1000 (GUIDE ON
TOPOGRAPHYC SHOOTING, 1977).
3.3. Device Connecting EXO with GPS
To verify the proposed algorithm a device enabling the testing of the above
technologies was designed (BABIJ et al., 2006).
The idea of connection is as follows: the interval meter for camera control is replaced
with epoch GPS controller changes. As known, the change of epochs may be
established at every second with virtually unlimited interval. Thus the signal during
the transmission from the controller to the modem is used for the exposure of the
aerial camera and testing its duty cycle. The connection of GPS to EXO flow-chart is
shown on Figure 29.
Fig. 29. Connection of GPS to EXO Flow-chart.

The implementation of this process takes place in the mode of providing

information to the modem, so the carrier frequency is used as the one transmitting
the relevant information. Input RTS signal from GPS controller is transmitted through
a protective diode to the divider built of fixed-value resistors. The choosing of epochs
happens by using containers. To convert the RTS signal of the radio-modem to a
logical "1", logic elements (amplifiers) with the properties of Schmitt trigger were
selected. This allows increasing the switching speed. In addition, this microcircuit
chip is used for its noise immunity and contains six independent one-way amplifiers
two of which are used for the inverted output. First DDI-amplifier inverts the pulse to
a logical "0" of TTL-level. The second amplifier amplifies the pulse and converts it into
a logical "1". The output stage is a classical scheme of the emitter follower. A key low-
frequency transistor, the output of which is protected from load capacitance by a
diode, is used. The device is powered with voltage 12V ± 10 %, and requires less than
50
0.1A current, power integrated circuits are made by a stabilitron tube and a diode.
The presence of +5.0V voltage on-chip is indicated by the light of a LED on the panel.
To implement the technological scheme of interaction of a GPS with a DC the
following processes must be performed:
 Bring GPS receiver into operating position, namely, combine the sensor with
the antenna and power supply and the controller, with help of buses connect
the modem to the controller.
 Connect the power coupling through an appropriate socket to the controller,
and the output circuit of the block to the DC.
 After the imposition of connection of the GPS receiver with satellites and the
start of counting epochs’ exposure of the DC is triggered.
3.4. Laboratory facilities for research
To determine the value of delay, a method has been proposed that bases on
the use of a device consisting of an electric motor, which is rigidly mounted on a stand
in an upright position (GLOTOV, 2005a, SIEJKA, GLOTOV, 2012). The engine has a
step-down gear, constructively placed together with it in a single case. The
transmission value is 1:20. At the end of the output gear shaft a bearing rod is fixed
horizontally by a thread joint (duralumin pipe with a diameter of 40 mm). At one end
of the rod with an axis perpendicular to the rod the antenna of the GPS receiver and
the camera are fixed, and on the other side – another GPS antenna. The entire
mechanism is set in a geodetic tripod. When applying electric current to the engine,
the rod begins to rotate around the axis and moving the antenna and camera in a
circle with a radius equal to half the length of the rod (Fig. 30).
Fig. 30. General view of the complex

However, after conducting research and analyzing the material a significant

difference in coordinates of the projection centres determined with the help
photogrammetric triangulation and GPS receiver was detected. Moreover, it was
observed that exposure of the camera is with a suitable delay, which clearly was
systematic. For these reasons the above described research laboratory complex of
GPS-receivers and DC proposed the authors was applied.
While moving the antenna complex a circle with a radius of 2 m was spun. The
angular velocity of the antenna was equal (for one circle) approximately 4 s. Base
51
station (permanent) was located at a distance of 30 m from the unit. Fixation rate of
epochs was equal to 20 s. This was done for reasons of multiplicity of rotation of the
device, so that the subsequent frame is carried out in the same place the original
camera position. GDOP parameter in the observations does not exceed 3 units. Length
of measurements was equal to 5.3 min. which corresponds with 16 frames (excluding
frames that were made in the fixed position). Figure 31 shows the arrangement of
points that were defined by the GPS receiver, and the difference of points (42-57) is
not due to the multiplicity of time of rod rotation.
Before the observation for 10 minutes, an initialization of the moving antenna in a
fixed position in relation to the base station (points 37-41) was performed. Then the
rotary motion of the antenna began. Observations were carried out by dual-frequency
receivers Trimble (base) and Trimble mobile. Processing of measurements was
performed using the software by Trimble.
After the first part of the experiment the measurements of the angles between the
centres of the shots that have been identified on the terrain at the appropriate
contours were performed. Thus, as the result of the study the angles that correspond
to the time shift exposure and fixation coordinates of the GPS receiver (see Table 2)
were calculated. The mean time delay was equal to 1.444 sec. To prove the sufficient
accuracy of the correction, we calculate a priori accuracy of the method. The
minimum temporal error is equal to 0.001sec., which corresponds to 5.4 angular
minutes.
Fig. 31. Fixation of coordinates by a GPS receiver (42-57) and fixed photography points (37-41)
52
Table 2. The values of time-delays and error of GPS
№ temporary
RMS GPS, s
points delay, s
1 1.444
2 1.443 -0.004
3 1.445 0.005
4 1.444 -0.003
5 1.443 0.001
6 1.442 -0.005
7 1.445 0.002
8 1.444 -0.000
9 1.442 -0.002
10 1.441 0.002
11 1.445 0.003
12 1.445 -0.005
13 1.444 0.001
14 1.441 -0.001
15 1.443 0.000
16 1.441
Thus, it is sufficient to measure the angles to be measured by the centers of

the shots with this accuracy, and therefore, using goniometric work precise
theodolites and tacheometer-rangefinders with which no more errors in the
determination of the amendment will be made. For positioning error of the GPS
receiver, it was as shown in table data - 0.0025 sec., since accuracy is on average 1 cm
in linear terms, which corresponds to 21 angular minutes.
In fact, the accuracy of the research was equal to 0.004 sec. that corresponds at the
carrier velocity of 40 m/sec. to the accuracy of the planned coordinates’
determination of 0.15 m. Survey consisted of field and office works. The purpose of
field work is to obtain photographic (digital) information about the terrain in the
form of stereoscopic pairs, as well as data defining elements of their exterior
orientation (DOROZHYNSKY et al., 2004).
To prepare the field survey work it was carried out geodetic support on area
where survey is performed (BAHMACH et al., 2003). Special marks for reference and
control points with size 40x20cm were made. The working field of a mark is black
slanting cross with 5cm wide on a white background. Fixture stake was mounting on
back side of a mark to fix it on the ground surface r, and the upper part of a stake
coincides with center of a mark. This is done for reasons to put there GPS rod. Marks
were placed along the perimeter of the survey object, so as to ensure the sparse
planimetric and altimetric referencing (PAR) of images (Fig. 32). Coordinates of 20
contour points located around the object site were determined to ensure complete
PAR.
53
Fig. 32. Marks of planimetric and altimetric preparation, located on the coastline
Work began with photogrammetric survey of Peninsula Marina (i.Halindez)

from the motorboat, which was equipped with a tripod for fixing digitals camera.
Survey was executed with minimal waves with Yav = 50m, allowing create plans in
scale 1: 1000. Such way 32 stereo pairs of mentioned Peninsula were taken. In
addition, survey of planar territory of the peninsula Marina was done from the side of
the dome, making it possible to create a large-scale general plan of this object. For
these reasons, on the western slope of the dome five surveying baselines were fixed
shooting, their PAR was performed using GPS and 10 stereopairs had been taken.
However, severe ice situation did not allow performing full amount of works
according to developed technology. Then it was decided to carry out survey of the
island from the surrounding territories: islands Winter, Grotta, Corner and Skua.
Orienting device - theodolite 2T30m was replaced with more precise theodolite 2T2 in
order to withstand the planned accuracy. It can determine the angular exterior
orientation elements with accuracy 2" when allowable error is in 4" - 5 ". For a
complete mapping of territory the survey was carried out by two cases: normal and
equideflected, depending on possibilities of baselines location and camera
perspectives. Works began on the island of Winter, where 10 points of baselines were
fixed; 26 stereo pairs of researched shoreline were taken and PAR of these points was
made using GPS in static mode. In addition, for modelling relief and contours of this
part of island there was made survey with digital phototheodolite complex with
preliminary marking of 8 points for normal and equideflected cases of survey.
After completing field work on i.Winter photogrammetric works were
continued on the island Grotta, intermediate Islands and island Corner, where
similarly 9 points of survey baselines were fixed and their PAR was made. Altogether
28 stereopairs of eastern part of the island Galindez were taken.
54
Works were not conducted on the island Skua and from the side of the strait Pinola
due to the limited term of staying at the Vernadsky station and impossibility of
delivery equipment and workers because of ice situation.
In following seasonal expedition works related to terrestrial photogrammetric
survey of Galindez Island were continued and completed.
"Dead" zones were detected when implementation of digital photogrammetric
survey of the island shoreline. So for these areas topographical survey using GPS
receivers was applied in the central part of the peninsula Marina, specifying
situational position of elements of geodetic provision of station and for further
stratification of GIS.
Owing to the relief nature of the Argentine Islands archipelago there became necessary
to do topographic survey of western part of the island Halindez using GPS. However
this work had not been done as it was not possible to get directly to the object. Then
there was fulfilled digital photogrammetric survey of peninsulas Penguin and Pidgin.
Marking and referencing of control points of territories were performed using landing
from a motorboat. Then digital photogrammetric survey was implemented from two
baselines fixed on dominating heights for normal and equideflected survey cases (8
points were marked and 12 images were taken). As a result, the internal part of
western region of the island was fully provided with data for mapping in appropriate
scale.
Applying proposed technology of coastline mapping using a motorboat, the survey of
western part of the island with preliminary sparse referencing around whole plane
that was subject to survey (altogether 18 points were marked and referenced) was
implemented (GLOTOV, 2005b). The survey was done using the motorboat Zodiac.
Digital camera was fixed on the water vehicle and antenna GPS-receiver was mounted
on the camera. The motorboat was moving with appropriate speed according to
designed course, and in the moment of exposure it braked to determine precise linear
elements of camera orientation. In such way moving on the strait Pinola, survey of the
island shoreline was made (taken 21 images). Moreover, the preliminary analysis of
the captured data makes it possible to conclude about feasibility of further
application of the above described technologies. Index of route unstraightness, that
was less than 1% when allowable value is 5.7%, can serve as an example. In addition,
the combination of camera with GPS and especially their synchronization enable to
reduce significantly the number of control points using the coordinates of projection
centers of camera in the performance of phototriangulation.
To implement survey from southern part of the island, from the side of the
island Skua, stations for digital photogrammetric survey (long-term) were
established on this island and then were used to monitor glacier on island Halindez.
On these stations there were made photographic survey (8 images) and preliminary
marking and referencing of 4 PAR points.
Consequently, all data of island territory captured by digital photogrammetric survey
from a water vehicle or baselines points and by GPS were prepared for further office
processing to create topographic plan in scale 1: 1000 as a basis for GIS.
Office processing of terrestrial digital images was done at DPS "Delta 2" (SIEJKA,
GLOTOV, 2012). Let’s consider proposed technological scheme of processing of digital
images.
55
After setting out initial parameters in the windows "Camera" and "Ground Control
Points" orientation of images was implemented. Inner and relative orientations were
not executed because, firstly, obtained data are digital images, not scanned, and,
secondly, it is proposed to build the model and perform absolute orientation on two
separate images. The process "Absolute orientation" was implemented in full volume.
Absolute orientation parameters were set out, which include survey type and
orientation method. Approach of absolute orientation process was to implement
stereoviewing of mark (or cursor) on ground control points.
The next stage of the work was capturing information about terrain. The program
allows to collect vector information from raster images (manual vectorization).
Capturing of information was done in the following sequence:
 Outlines;
 Pickets;
 Contours.
In following seasonal expedition experimental research works on creating
large-scale plans of region of the Antarctic Vernadsky station were continued.
Altogether 350 stereopairs were taken from fixed and intermediate baselines, as well
as from water vehicle. Geodetic GPS receivers were used to obtaine coordinates of
240 points for field referencing of images and for the surveying of area (control
points).
In result of the works topographical plans in scale 1: 1000 of islands Galindez, Winter,
Skua, the eastern part of the island Peterman and the cape Hut of Antarctic coast had
been created.
3.5. Peculiarities of digital photogrammetric survey
Analysis of the technological scheme of creating large-scale plans, will

allow to determine the need for additions, improvements and modifications of the
investigation method. The task is to identify the peculiarities of the technological
scheme for creation of topographical plans (as a base for GIS) and to keep its as
possible lowest cost, i.e. the cost of field and office works not only for mapping
coastlines, but for whole area of the islands in the region. In general, as already
mentioned, about 350 stereopairs were taken with digital cameras Kodak DC 260,
Olympus E 20r and Sony DCS. Processing these images of researched territories
the plans in scale 1: 1000 had been created, which makes it possible to analyse the
peculiarities of the proposed process flowsheet of field and office works.
Slope of the mean plane of the object was taken into account during survey from
baselines. For this baseline, if possible, was located above the object and tilted
case of survey was applied (Fig. 33). Since distances to the mean plane do not
exceed 100-300 m, then this case does not significantly effect on the accuracy of
coordinates determination (GLOTOV, 2004).
As our experience shows, in most cases of large scale survey with using
photogrammetric straight intersection, does not always ensure accuracy of
coordinates determination. This is stipulated by phototheodolite design features:
 not rigid connection of digital camera with theodolite;
 possibility of shift of theodolite vertical circle at the moment of
photography;
56
 presence of digital camera aberrations.
Fig. 33. Digital image taken from baseline of shoreline (tilted survey case)
Therefore, we conclude about need for creation of production sample of digital

phototheodolite with a certain set of characteristics. First of all it should be chosen such
type of camera that its resolution would as high as possible and the sizes of CCD matrix
would be maximal. Cameras can be series like Canon EOS 350D, Olympus E-300, the main
characteristics of these digital cameras are listed in table 3. Note, when the maximum focal
lengths of standards lenses of these cameras, distortion practically does not exceed
allowable values (VOLCHKO et al., 2005, GEL'MAN, 1982, GLOTOV, 2004, GLOTOV, 2003,
GLOTOV, MAJOROV, 2000a, GLOTOV, MAJOROV, 2000b]. This is explained primarily that
only paraxial part of optical system is used. But at the same time distortion was accounted
pixel by pixel, according to the method we had developed.
Table 3. Basic specification for digital cameras

total amount
size Capacity
Performance optical of
CCD-matrix imagery ISO equivalent
Models exposure (s) pixels
(pixels) (Mb)
(million)
1/4000 - 60 s
Olympus E-300 3264 x 2448 23.3 100/200/400 (up to 8 8.89
minutes)
100, 200, 400, 1/4000 up to
Canon EOS350D 3456 x 2304 3.3 8.2
800, 1600 30
3.6. Characteristics of orienting device
During the researches precise theodolites Theo 010V and 2T2 were used for
reasons of excessive precision. Let’s justify this decision by appropriate calculations of
accuracy of absolute orientation elements for Digital Camera Olympus E20p with
57
equivalent focal length. For this we use formulas (19) that establish relations between the
coordinates of the point of vertical survey and coordinates of corresponding points of
image which has inclination angles (LOBANOV , 1968):
Δx 
Δ  ρ;
f
f  Δz
Δω  ρ; . (19)
f 2  z2
Δx 
Δκ  ρ.
z
Let’s calculated using these formulas values Δ , Δω , Δκ for f = 90 mm, x = 40
mm, z = 30 mm, Δx = Δz = 0,005 mm. We will obtain: Δφ = 11˝, Δω = 10˝, Δκ = 35˝. Thus, for
orienting device should be applyied theodolites with precision of angles determination
within ± 5 ".
In the capacity of orienting device it should be used a theodolite as it give possibility
to determine the coordinates of control points with maximum accuracy, because
accuracy of angles determination should be as high as possible (GLOTOV, 2005a,
GLOTOV, 2005b). For rapid identification of directions it is proposed to use electronic
total stations, such as total station for precision measurements of firm Sokkia SET
2110-322, its accuracy of the angles determination is 2 ", and distance determination
is 2mm + 2D.
It is appropriate to emphasize that digital phototheodolite should consist of
professional camera installed on orienting device and non-parallelism of axes can be
accounted analytically (GLOTOV, 2005c ). If necessary, the camera can be dismantled
to perform angular observations by electronic total station.
Coordinates of baselines were determined using GPS-receivers of geodetic class that
definitely made it possible to calculate coordinates with sufficient accuracy (GLOTOV,
2002). Note that receivers of the following types: Leica (SR 530), Topcon (ODYSSEY),
Trimble (4600LS) and others can be used.
In our study realisation PAR was performed to determine the coordinates of
reference and control points. Basically it was sparse referencing, as later
phototriangulation was performed. Let’s define the following peculiarities of this
process:
 viewing on unmarked point is complicated by the fact that it cannot be
unambiguously targeted on micro object by telescope of orienting device
and measurable viewing marks of DPS;
 marked points allow to maximize accuracy of images orientation on DPS.
During the topographic works the survey of shoreline, and if possible survey
of remote areas, was implemented from motorboat (Fig. 34). Advantages of this
method comparing with survey from baselines in this region are obvious:
 it is not always possible to place the baselines on an object of survey;
 application of water vehicle give possibilities to implement consistently
survey close to normal case of survey;
 there is also possibility to determine the coordinates of camera projection
centers using GPS.
58
Let’s briefly discuss the proposed method of determination of linear exterior
orientation elements. GPS-receiver is set on the earth's surface on the area of survey,
and camera and onboard GPS-receiver with antenna connected to the camera are set
on board of moving water vehicle. Preliminary the epochs, which correspond to
intervals of photography taking, are established at onboard GPS-receivers. Steering of
camera is implemented by turning it on by signal of epoch changes at onboard GPS-
receiver (Fig. 34). At the same time the spatial coordinates of camera center
projection are determined.
Fig. 34. Survey of the island Skua shoreline from the motorboat using the complex
To test the proposed algorithm the following experiment was implemented.

On the surveyed coastline, 47 control points were marked and their coordinates were
determined using GPS receiver in kinematic mode with accuracy: mX=mY= 0,01m, mZ=
0,04m. Then, photogrammetric survey of these points was performed from a
motorboat. Data processing was implemented on DPS by the classical scheme
(phototriangulation, mapping relief and terrain, etc.). It should be noted that distance
to the object was 120 - 150m. The motorboat speed - 1m/sec.
In parallel, calculations on the proposed algorithm were carried out. Then there was
done the comparative analysis of results of phototriangulation and results of
determination of points coordinates relatively to GPS measurements (the latter were
considered as theoretical). The results of mean square errors (MSE) are presented in
table 4.
59
Table 4. Values of MSE control points obtained according to calculations in phototriangulation
and according to the proposed algorithm
MSE (phototriangulation) MSE (proposed method)
mX (m) mY (m) mZ (m) mX (m) mY (m) mZ (m)
0,17 0,28 0,13 0.12 0.18 0.10
As you can see, the value of MSE when application of the proposed algorithm
is about 1.5 times smaller in comparison with a classic method. Thus, we can use
equation (18) when the implementation of photogrammetric survey by the complex
GPS - camera from a water vehicle for shoreline mapping.
In addition, having water shore line represented on image image, we can determine
the angular orientation elements. To solve the problem we use the condition of
collinearity. Considering small value of angles the equation systems can be
represented as follows:
p x2 z2
h  z    f    x , (20)
B f f
where h - the height of camera above water surface; p - the point parallax measured by
water shore line; B - base line of survey.
To determine the unknown  ,  ,  , it is enough to measure the value of x, z
and p in the points according to water shore line and solve the equations system (20), as
only angular elements are unknown. Baseline of survey is determined using coordinates
of the projections centers; h - the difference between Zs and Z at the moment of survey.
Thus, photogrammetric survey from a water vehicle allows to get excessive magnitudes
for further processing of measurement results, which in turn increases the accuracy of
digital images orientation.
During processing image on DPS there were difficulties with orientation. To avoid them it
is necessary: to implement survey in such way to provide overlapping of stereopairs, and
this is possible if apply equideflected case of survey; angle of object capturing should be
such that object (terrain) was as completely represented on image as possible and only in
bottom part image contents the water shoreline. Appropriate implementation - change
camera focal length.
3.7. Conclusions
Formulas (18), obtained on the base of linear kinematic projecting, give possibility to
analytically process digital images obtained by ground based photogrammetric
survey, without any constraints imposed on their spatial position.
Results of a priori estimation of accuracy show that the proposed algorithm is fully
compliant with the requirements of large-scale mapping.
Values of error mY does not exceed error of plane coordinates, because the value Y is
determined using these coordinates.
Experimental research works demonstrated the ability to perform survey with
subsequent processing and creation of large-scale plans using digital phototheodolite.
It was developed the method and implemented production tests of application of
digital camera in conjunction with GPS for determination of linear elements of images
60
external orientation when survey of islands coastlines of the Argentine archipelago of
Antarctic coast is implementing from water vehicle.
The sheetflow of terrestrial digital images processing on DPS "Delta-2" was
developed.
As a result of implemented works topographical plans in scale 1: 1000 of islands
Galindez, Winter, Skua, the eastern part of island Peterman and cape Hut of Antarctic
coast had been created.
61
4. OPTIMIZED TECHNOLOGY FOR GNSS HEIGHT
DETERMINATION USING ASG-EUPOS
There are a lot of geodetic methods for determining of heights or height
differences. Considering the most frequently used these methods are classified as
geometric levelling, trigonometric levelling, and GPS/Levelling.
Geometric levelling is the determination of the height differences by using level and
hold two vertical rods (Fig. 35). Height difference is obtained via simple
formula:
= − = − (21)
where: HB – height of point B,
HA – height of point A,
tB – measurement at the point B,
tA – measurement at the point A.
Geometric levelling is considered as a simple and yielding very good result method.
However, the practical applications have shown that carrying out of this method is
very difficult especially on the mountains areas and sensitive to different model
errors.
The preventative measures must be taken to eliminate or reduce model errors. If it is
not, these situations decrease the survey velocity and accuracy, thus the cost of
surveying rises (BANGER, 1981; CEYLAN et al., 2005; NIEMEIER, 1986; BAYKAL,
1989).
The level may be opto-mechanical or digital, which implies different levels of security
regarding possible blunders, and also different levels of precision. The precision may
range from 0.3 to 3 mm/km and more. The equipment has to be used at least by 1
observer and 1 helper for holding the rod. Generally geometric levelling requires 1
observer and 2 helpers for the staffs. When the team works along roads, it is often
mandatory to have one extra worker to protect from the traffic. Thus the team varies
from 2 to 4 people. The daily production depends strongly on the equipment and the
composition of the team, from 4 km/day to more than 30 km/day (KASSER, 2002).
Fig. 35. Geometric levelling

Resource: elaborated by DAWIDOWICZ.
62
Nowadays, also motorized geometric levelling applications have been done by
establishing survey hardware on the vehicles. Only disadvantage of this technique is
that the cost of instrument and vehicles is very high and level points must be on the
edge of the road (CEYLAN et al., 2005; NIEMEIER, 1986; BECKER, 1986).
Height differences in trigonometric levelling are computed by using vertical
angle and distance (Fig. 36). Height differences are computed via formula:
=ℎ + (90° − ) − ℎ (22)
where: d – slant distance from the instrument to the signal,
Z – zenith angle,
hi –height of instrument,
hb –height of signal.
According to the land, time and observing vertical angle, trigonometric
levelling can be classified as follows: - unidirectional trigonometric levelling, - jumped
trigonometric levelling, - reciprocal trigonometric levelling.
This methodology is generally much faster than geometric levelling and of lesser
accuracy due to refraction effects. An exception is the trigonometric levelling using
simultaneous reciprocal measurements. The advantage of this method is that it can be
motorized. Recently it has been widely developed and used. The motorized
trigonometric levelling is equal accuracy ( ≤ 2mm / km) and equal cost with the
motorized geometric levelling when the measurements are done according to some
rules: - the reciprocal and simultaneously vertical angle observations, - the reciprocal
distance measurements, - the short observation ranges (250-300 m), - using
calibrated instruments, - carried out by the experienced person (CEYLAN et al.,
2005). Additionally a productivity of trigonometric levelling stays at a high level even
in mountainous areas.
Fig. 36. Trigonometric levelling

GNSS levelling is the most recent and advanced method that is used in the
determination of heights. Three-dimensional coordinates or coordinate differences
can be obtained by GNSS in the geocentric Cartesian coordinate system. The Cartesian
coordinates are transformed to geodetic latitude, geodetic longitude, and ellipsoidal
63
heights according to selected reference ellipsoid, i.e. WGS84. Generally surveying
applications require heights that are related to a physically meaningful surface such
as the geoid, or at least some attempt at realizing the geoid such as a surface based on
locally observed mean sea level. Such physically meaningful heights take the form of
orthometric or normal heights. Relationship between ellipsoidal height and
orthometric height is shown in Figure 37 (CEYLAN et al., 2005). Heights differences
in GNSS levelling are calculated using equation:
= ℎ − (23)
where: ℎ – ellipsoidal height differences,
– geoid undulation differences.
GNSS levelling has also several drawbacks, e.g. the benchmarks do not have to be
along roads, but require an open sky above them, which is not suitable in dense urban
areas (we shall remember that most surveying works are performed in such areas).
Fig. 37. GPS levelling

The final accuracy depends on the accuracy of GNSS measurements and of geoid
model. Although it is possible to reach millimeter horizontal relative accuracy levels
over tens, or even hundreds of kilometers, vertical GNSS accuracy is not so easily
obtained. The baseline vertical component is more sensitive to many disturbing
factors, for example: antenna phase centre variations or tropospheric refraction.
Accuracy of GNSS measurements is degraded by many factors. Below are
presented a complete classification of the GPS error sources (FIGURSKI, 2001):
1) imprecise knowledge of the satellite orbit parameters,
2) the measurement technique used (Static, Fast-static) and geometry
parameters of the network,
3) instrumental errors:
a) associated with the satellites:
- the on-board clock error,
- the variation of antenna phase centre,
64
- accuracy of the satellite ephemeris,
b) related to the receiver:
- the clock error,
- orientation and location of the antenna phase centre,
- the antenna height,
4) propagation disturbances:
a) the tropospheric refraction,
b) the ionospheric refraction,
c) the indeterminacy of the initial number of phase cycles,
d) the asymmetry of the constellation of satellites in the horizon,
e) the multipath,
f) the relativistic effects,
5) the adopted values of physical constants:
a) models of the physical phenomena,
b) the system parameters,
c) the transformation parameters between the reference systems,
d) the earth’s polar motion parameters,
6) the numerical accuracy and random errors.
The second factor determining the finall accuracy of GNSS levelling is an accuracy of
geoid modelling. An excellent knowledge is required of the geoid as GNSS provides
only geometric observations, and height is a geopotential information. If the geoid is
not available, the surveyor will have the possibility to use GNSS on a limited zone by
measuring the discrepancy between the official altitude and the ellipsoidal height.
The transformation can be achieved in different ways. First of all, a sufficient number
of common points are required to perform the transformation. This number depends
on the work to be carried out. If sub-centimeter level accuracy is required, a large
number of common points are needed. Another criterion is the precision of the local
height system. Lack of a large number of common points is a problem in practice. In
such a case, the number of points in the local levelling net must be increased, which
causes additional field survey, time consumption, and cost. There are many
interpolation methods solving the surface-fitting problem.
In Interpolation with Weighted Average the height of an interpolation point is
calculated by the weighted arithmetic average of the heights of the reference points
around it. The weight of a reference point is a function of the horizontal distance from
the interpolation point.
The aim of Interpolation with Polynomials is to express the surface with a polynomial
function.
The aim of the Multiquadric Interpolation method improved by HARDY (1971) is to
represent the surface with a single function. First, a trend surface is determined. A
polynomial or a harmonic series or a trigonometric function can be used as the trend
surface. The experiences have shown that first- or second-degree polynomials satisfy
the requirements (LEBERL, 1973).
Interpolation in Net of Triangles method uses the reference points as the vertices of a
net of nonoverlapping triangles that cover the interpolation area. Various algorithms
can be used for triangulation. The most common ones are optimal, greedy, and
Delaunay triangulation (YANLAK, BAYKAL, 2001).
65
4.1. European Position Determination System
The EUPOS (European Position Determination System) initiative is one of the

biggest projects introducing GNSS reference systems in Central and Eastern Europe.
The project was started in 2003 based on the success of the German SAPOS system.
The project unites 16 European countries into one GNSS reference system. All
countries build up their individual systems, but cross-border real-time data exchange
enables the separate systems to work as one region-wide system.
The EUPOS initiative created standards for establishment of national systems,
services compatibility and data exchange with foreign countries. Those standards
were basis for establishing the satellite positioning system ASG-EUPOS (Active
Geodetic Network EUPOS) in Poland. On 2 January 2007 the contract between the
Head Office of Geodesy and Cartography and the consortium: WASKO S. A., Geotronics
Polska Sp. z o. o., Trimble Europe BV., for building the ASG-EUPOS system was signed.
This was the beginning of the project realization (ORUBA et al., 2008).
The structure of the ASG-EUPOS is divided into three basic segments:
reference stations, management centre, user segment. These segments work together
and provide precise real-time positioning and post-processing applications. The
reference stations network currently (October 2014) consists of 101 Polish (77 with
GPS and 24 with GPS/GLONASS module) and 22 foreign regularly distributed stations
(Fig. 38).
Fig. 38. ASG-EUPOS

Resource: www.asgeupos.pl
66
The mean distance between reference stations is in average 70 km. Three
real-time correction services and two post-processing services are currently available
for users (Table 5). It is worth noting that at present ASG+ project is in its final stage.
The main objective of the project is to create supporting modules for currently
operating in ASG-EUPOS system services, create new system services and improve
the quality of positioning (FIGURSKI et al., 2011).
Table 5. ASG-EUPOS services

SERVICE SERVICE SURVEY DATA ESTIMATED MINIMUM HARDWARE
GROUP NAME METHOD ACCESS PRECISION REQUIREMENTS
Real-time NAWGEO kinematic GSM / 0.03 m (horiz) L1/L2 GNSS RTK receiver,
services RTK Internet 0.05 m (vert.) communication module
KODGIS kinematic 0.2 – 0.5 m L1 DGNSS receiver,
DGPS communication module
NAWGIS 1.0 - 3.0 m
Post- POZGEO static Internet 0.01 – 0.10 m L1 GNSS receiver

processing
services POZGEO D static /
kinematic
Resource: elaborated by DAWIDOWICZ based on www.asgeupos.pl.
NAWGEO is a fundamental ASG-EUPOS service which provides corrections for

real-time RTK measuring. Other services (KODGIS and NAWGIS) are intended for
users who do not require such high accuracy. Using the NAWGEO service, a user has
the possibility to take measurements in Single Base Station mode, or chose one of two
types of RTK corrections (Masterter and Auxiliary Concept or Virtual Reference
Station). A review of the real time services accuracy can be found e.g. in: BAKUŁA,
PRZESTRZELSKI, 2013; DAWIDOWICZ, 2012; EDWARDS et al., 2010; HADAŚ, BOSY,
2009; KOWALCZYK, TYMOSZCZUK, 2012; KUDRYS, KRZYŻEK , 2011; WIŚNIEWSKI et
al., 2013.
The POZGEO-D service is meant for users who utilize their own software to process
GNSS observations. In this approach, user downloads observation from ASG-EUPOS
reference stations for their own processing. Using the POZGEO-D service makes
possible to obtain higher accuracy from shorter observational sessions, although, it
requires some knowledge of GNSS observation processing and owning your own
software.
Using the POZGEO service, a user sends observation files in order to receive
automatically calculated coordinates of the measured points. The result are available
via the ASG-EUPOS website. Post-processing of the uploaded observations and
determination of the point’s coordinates are made independently in reference to the
six nearest system station. Because of the relatively long distances between stations
(~ 70 km) and the automation of the calculation process, obtaining high accuracy
requires long observation sessions (ECKLE et al., 2001; EL-MOWAFY, 2011; KADAJ,
2010). The final accuracy also depends on e.g. the measuring conditions, time of
observations or number of observed frequencies (DAWIDOWICZ et al. 2014).
67
4.2. Height measurement using ASG-EUPOS
So far a number of studies related to the height determination based on

measurements in the ASG-EUPOS network were conducted.
In DAWIDOWICZ (2013) the accuracy of the NRTK height measurements
using ASG-EUPOS system was analyzed. For analysis ten points in two test areas were
selected. On selected points several RTK measurement sessions, using three types of
correction available in the NAWGEO service, were conducted.
Generally, in this study, was found that RTK-NAWGEO height measurements have
good precision. The standard deviations in one of conducted test did not exceed 1.4
cm (with one exception) and the maximum height changes (with the same exception)
were within 5 cm range. In one case (Point 1 – MAC 24.05 session), an unexpected
jump in height in the 17 cm size occurred (Fig. 39). The reason for this jump was
unclear. This probably was done by undetected cycle slip or wrong ambiguity
solutions. Except this case results obtained for all other points had similar height
variation characteristics.
Fig. 39. Results obtained for two selected points

The standard deviation in another test, in the case of MAC and Single Station
measurement mode, clearly increased with the distance to the nearest CORS station,
for VRS correction it was almost stable and equoled 1.8 cm on average
(DAWIDOWICZ, 2013).
It was noted also that there were some systematic differences between heights
obtained using different types of corrections. Generally it was assumed that,
depending on the used type of correction heights can differ up to several centimeters.
There were also clear systematic differences between heights obtained using NRTK
and precise levelling. This means that used type of corection can have an impact on
obtaining heights (Table 6).
68
Table 6. Average height diferences (geometric levelling - average NRTK levelling) for first test
area
Point Average height differences [m]
number Session identification
24 March 25 March 24 May 28 May
VRS SS VRS SS VRS MAC VRS MAC
1 0.043 0.012 0.048 0.012 0.076 0.030 0.004 0.018
2 0.063 0.017 0.062 0.024 0.057 0.001 0.016 0.004
3 0.063 0.031 0.054 0.019 0.073 0.025 0.025 0.022
4 0.056 0.024 0.052 0.017 0.037 0.066 0.018 0.025
The normal height accuracy, measured as a difference between NRTK and precise
levelling heights, obtained in first test varied significantly depending on the
meseuring session. The accuracy was better than 2 cm for an average of 31% and
better than 5 cm of 71 % of the NRTK measurements. For measurements where the
largest differences occurred, fitting NRTK heights to a precise levelling network
significantly improved the results. This can be done through normal height
determination at least of the one point using geometric levelling. But it should
remembered that NRTK measurements often are fluctuating and their average also
can be biased. So it would be better to used two or more points. Such approach allows
for better control and adjustment of results.
If we perform measurements in the area where there are not bench marks a solution
could be measurement using all available types of corrections. The analysis of
obtained differences in the results could be helpful in identifying the best solution.
In DAWIDOWICZ (2012) the author performed an analysis of normal height
determination using the post-processing services: POZGEO and POZGEO-D.
Four points situated in UWM Olsztyn were selected for test measurements. A static
session was carried out on the test points. The following GNSS parameters were
assumed for that session: sampling interval 5s, minimum satellite’ elevation 10°, time
of measurement 4 hours. For the test points normal heights were determined by
precise levelling. Precise levelling was assigned to two 2nd order benchmarks of the
national levelling network.
Observations from static measurements carried out on test points have been
processed in several variants. Because cost-effectiveness is a requirement for most
geodetic projects several analyses were conducted into how the accuracy depends on
the baseline length and on the duration of the observing session (ECKLE et al., 2001;
PSIMOULIS et al., 2004). For this reason, four-hour session was divided into 4 one-
hour sessions and 8 half-hour sessions. Next, the four and one-hour sessions were
sent to the POZGEO service. Additionally, all measurement sessions were processed
with Topcon Tools v7.3 software (POZGEO-D service) in three variants: with LAMA
(~20 km baselines), with OLST (~3 km baselines) and with VRS (~0.5 km baselines)
stations as reference stations. VRS station were created with the POZGEO-D service in
the immediate vicinity of the measured points.
For all processing variants the normal heights were calculated using the “QGEOID-PG”
model. The normal height differences between heights obtained from precise
levelling and the heights obtained from satellite levelling are presented in Figure 40.
69
Height differences calculated from the results from one-hour sessions were within ±
2 cm range for 44% of the POZGEO solution, for 75% of the POZGEO-D LAMA (P-D
LAMA) solution, for 56% of POZGEO-D OLST (P-D OLST) solutions and for 50% of
POZGEO-D VRS (P-D VRS) solutions. With ± 3 cm range were 69% of the POZGEO
solution, 94% of the POZGEO-D LAMA solution, 94% of POZGEO-D OLST solutions
and 100% of POZGEO-D VRS solutions.
Height differences calculated from the results of half-hour sessions were within ± 2
cm range for 68% of the POZGEO-D LAMA solution, for 59% of POZGEO-D OLST
solutions and for 50% of POZGEO-D VRS solutions. With ± 3 cm range were 90% of
the POZGEO-D LAMA solution, 75% of POZGEO-D OLST solutions and 90% of
POZGEO-D VRS solutions.
Fig. 40. Normal height differences (geometric levelling - GPS levelling)

A one-hour static session proved to be too short for accurate determination of

ellipsoidal heights in the POZGEO service (height dispersion in the range of 7 cm).
Neither method provides determination of normal heights with ±2÷3 cm accuracy.
A better solution at the moment is either extending the measuring session for
processing in the POZGEO service or own observation post-processing (POZGEO-D
service). A half-hour session was sufficient to determine ellipsoidal heights with ±1
cm accuracy (variants with OLST or VRS as reference station). Additionally, because
of the short vectors between the measured points and the reference station, the
accuracy of the calculations did not depend on the type of receivers used for the
measurement (L1 or L1/L2). For those processing variants, the accuracy of
quasigeoid model has a significant impact on the value of normal height calculation. It
is hoped that this accuracy will be improved with the newest quasigeoid model for
Poland.
In DAWIDOWICZ et al. (2014) there were also presented an analysis of the
position determination accuracy using ASG-EUPOS POZGEO service. It is well known
that the final accuracy is e.g. the measuring conditions, time of observations or
number of measured frequencies dependent. Four consecutive days of GPS data were
70
processed to determine how the accuracy of derived positional coordinates depends
on the length of the observing session, the characteristics of horizon visibility on
points and the used in post-processing observations (L1 or L1+L2).
To analysis three points were selected. As a point A the ASG-EUPOS reference station
KROL in Olsztyn was adopted, assuming that this is the point with optimal observing
conditions without any obstructions. Points B and C are marked in an urban area,
where occurring trees or buildings limiting the number of observed satellites and
increases the risk of multipath (Fig. 41).
Fig. 41. Obstacles on measured points: a) A point, b) B point, c) C point.

The GNSS parameters which were adopted for measurement sessions: minimum
height of satellites above the horizon 10˚, measurement interval 1 s. Using two
frequency receivers allowed to divide observations into 2 variants depending on the
observed signals: L1+L2 observations and L1-only observations. 24-hour data were
subdivided into 6 time variants (0.5h, 1h, 2h, 3h, 4h and 6h).
Figure 42 presents RMS errors as a function of session duration and observing
conditions at the point. The positioning RMS errors are calculated from the
differences between the ‘true’ hGRS'80 coordinates with the estimated values. The
average position from the four 24-h L1+L2 sessions was adopted as the ‘true’
position.
The results show, as expected that both the length of sessions and observing
conditions affect the height determination accuracy. The impact of obstacles is
especially visible in the case of processing short L1+L2 observation sessions. For
longer sessions, the influence of observing conditions on the accuracy was clearly
reduced.
Generally it was concluded that the vertical accuracies of 4 cm are achievable in
POZGEO service provided 0.5 hours low multi-path dual frequency GNSS data. For
such sort sessions the accuracy clearly decreases for point measured under
conditions of strongly limited satellite availability or when L1-only observations are
available.
71
Fig. 42. RMS height determination statistics
It can be expected that ASG+ project will be completed and when all ASG-EUPOS
stations will be equipped in multi-GNSS receivers, significant improve of the quality
of POZGEO positioning occurs.
Additionally in DAWIDOWICZ (2011) author analyzed accuracy of heights
determination from static GNSS measurements using the Virtual Reference Station
(VRS). Studies have shown that the procedure of height appointment with the use
VRS may allow to reduce the length of observation sessions and improve accuracy
compared to the results obtained from the POZGEO service. In addition, because of
the short vectors between the virtual station and measured points, accuracy is not
dependent on the type of used receiver (L1 or L1/L2).
Although the use of the VRS station can improve the accuracy of position
determination some problems are with the estimation of the accuracy of the
determination of virtual stations (POZGEO-D service does not provide such
characteristics) - this is undoubtedly an important factor obstruct the use of this
technique in surveying. In addition, there are also certain problems of formal nature -
it is difficult to define VRS stations as typical geodetic control network point.
4.3. Typical levelling situations
Nowadays geodesy distinguishes some typical levelling situations. Fundamental

Levelling of the National Network is generally done directly by a national office. The
goal is to provide benchmarks everywhere in the country, with a variation of density
for benchmarks close to the population density and a high. This network must be
observed at the lowest cost (compatible with this precision) possible, and regularly
checked because of benchmarks destruction. The goal of Urban Densification
Network is to provide levelling over a large number of marks, some of them being
often natural ones (sewer plates, sidewalk borders, etc.), the other ones being
benchmarks with special attention paid to their conservation. In most cases, the
requested accuracy is high (1 mm to 5 mm relatively to the national levelling
network). Semi-Urban Network will be requested for the preparation of new works,
town housing developments etc. The required accuracy will be of the same type (0.5
72
to 1 cm relatively to the national network), but the density of the benchmarks will be
low.
Rural Height Determinations may be requested because the national network is not
dense enough, if some new investments are planned. The density will be low, but the
references will be perhaps very far from the site. Stability Monitoring is done in order
to check the movements or deformations of a bridge, a dam, a high building, the main
point will be the highest accuracy possible, with local references established only for
these works, possibly with no link to the national network.
Control and Real Time Guidance of Construction Machines appears more and more
important for productivity gains in civil engineering, and especially for the
construction of roads, highways or train lines. There are many possible specifications
of precision. The base layer thickness for roads should be monitored within 5 cm, and
the last layers, that are formed with quite expensive materials, should have a
thickness control to within 5 mm (KASSER, 2002).
4.4. Theoretical analysis of optimal technique
Generally the National Network should be observed with geometric levelling or

motorised geometric or trigonometric levelling. So far GNSS levelling accuracy is too
low for this type of task. For the Urban Densification Network GNSS will also not be
profitable. There are exist too many situations where the sky is not fully visible (close
to buildings, trees, etc.), and too many benchmarks impossible to pick up directly
with the antenna. In such areas an auxiliary tacheometer will be requested, limiting
the benefit of the GNSS advantages. For the Semi-Urban Network considering the low
density requested, the use of digital levels can be considered the best method
because of their low cost. There are also possibility to use of high precision
tacheometers with reciprocal simultaneous angle measurements especially if the area
is large and/or with difficulties of communications. For the Rural Height
Determinations, the GNSS will generally be the best economical solution, as soon as
the work to be performed is not too small. The use of differential GNSS may be
considered if the topography allows for it: it will provide a better security for the
quality of satellite measurements and the integrity of the collected data will be tested
before leaving the zone. For the Stability Monitoring, the use of optico-mechanical
levels should probably be preferred for their high precision. In a case of automatic
continuous monitoring digital levels and GNSS receivers may be used. For digital
levels, the required length of rod may be fixed, for example to a building, and
monitored automatically by the digital level controlled by a PC. GNSS receiver may be
permanently installed on a given device, with a reference station not too far away (e.
g. less than 1 km), a power supply and if necessary a data link. Considering the
possibility to filter the results, even vertical movements as small as 2 to 5 mm may be
detected over periods of several days (KASSER, 2002). Additionally while many
applications for GNSS surveying need to produce orthometric or normal heights in
vertical deformation monitoring the most important issue is to quantify a change in
height over time and whether any change is relative to the geoid or ellipsoid is not
particularly relevant. Using GNSS for deformation monitoring brings the normal
advantages of GNSS surveying such as no requirement for inter-visibility between
stations and the ability to measure large distances with high precision. Also,
73
deformation applications require many repeated observations over time and GNSS is
well suited to automated survey processes that can significantly reduce cost
(HIGGINS, 2010; NEILAN et al., 1997). For the Control and Real Time Guidance of
Construction Machines three methods may be considered: GNSS, laser equipment and
automatic tacheometers. All of these have "pros and cons". For example GNSS, in real-
time differential mode may provide an excellent permanent control as long as there is
no problems of "shadow" zones where the satellites cannot be received and generally
its accuracy is not sufficient for the last layers, as it cannot guarantee better than 1 cm
(KASSER, 2002).
4.5. Conclusions
In this study height determination techniques have been shortly compared with each
other according to used instrument, and measurement method. The study detailed
examines the limitations of GNSS heighting including those factors that affect the
GNSS height measurement itself and the associated issues of geoid. The possibilities
for GNSS heighting using continuously operating reference stations (ASG-EUPOS)
were also discused.
Each given type of work requires a careful analysis, as usual, and a regular re-
evaluation to the method that is optimal at a given date. But precise height
determinations are not only equivalent to geometric levelling. Focusing on GNSS
technology it can be considered that sometimes the GNSS may be used, sometimes
not.
For example if there is a geoid information at levelling in rural area in which density
of point is so low and in which shadow area is not formed the GNSS levelling method
should be chosen. If the deformation surveys are carried out in big structures as
bridge, dam, GNSS receivers may be used. In addition, the precision levelling method
should be chosen by using digital level with invar rods. In construction projects as
highway, railway, smoothing area, the GNSS levelling, the trigonometric levelling or
the geometric levelling may be chosen respectively.
Considering height determination using ASG-EUPOS, as expected, height accuracy
depends on many factors.
Focusing on real-time services it was noted e.g. that there are some systematic (up to
several centimeters) differences between heights obtained using different types of
corrections. There are also clear systematic differences between heights obtained
using NRTK and precise levelling. The normal height accuracy, measured as a
difference between NRTK and precise levelling heights, generally was better than 2
cm for an average of only 30% of the NRTK measurements. But for measurements
where the largest differences occurred, fitting NRTK heights to a precise levelling
network significantly improved the results. This can be done through normal height
determination at least of the one point using geometric levelling. But it should
remembered that NRTK measurements often are fluctuating and their average also
can be biased. So it would be better to used two or more points. Such approach allows
for better control and adjustment of results. If measurements in the area where there
are not bench marks are performed a solution could be measurement using all
available types of corrections. The analysis of obtained differences in the results
could be helpful in identifying the best solution.
74
Focusing on post-processing services it was founded that one-hour static session
proved to be too short for accurate determination of ellipsoidal heights in the
POZGEO service (height dispersion in the range of 7 cm). Neither method provides
determination of normal heights with ±2÷3 cm accuracy. A better solution at the
moment is either extending the measuring session for processing in the POZGEO
service or own observation post-processing (POZGEO-D service). A half-hour session
was sufficient to determine ellipsoidal heights with ±1 cm accuracy (for short
baselines – up to several kilometers). Studies have shown also that the procedure of
height appointment with the use VRS may allow to reduce the length of observation
sessions and improve accuracy compared to the results obtained from the POZGEO
service. Unfortunatelly in all GNSS measurements normal height accuracy depends
also of the accuracy of quasigeoid model. It is hoped that this accuracy will be
improved with the newest quasigeoid model for Poland.
Thus we encourage the surveyors not to overestimate the accuracy of GNSS
(discrepancy between the repeatability of GPS - a few mm - and its real precision) and
underestimate the problems posed by the different reference frames of GNSS and
national levelling network (accuracy of geoid modeling).
75
5. A RELATION BETWEEN A VERTICAL CRUSTAL
MOVEMENT FROM SPIRIT LEVELLING AND GLACIAL
ISOSTATIC ADJUSTMENT (GIA) FROM GRACE DATA IN
THE AREA OF POLAND
In the world and in the Europe are working on unification of height reference
systems. So far, the most popular method of determining the height is levelling
method. Levelling data is most of the long periods of observation. For this purpose,
you can also use GNSS data, however, the life of the station is still too varied in
different countries. Levelling data due to their nature and amount of work were
performed at different times and on different measurement areas. Levelling lines
were also conducted at various points and with different density. The result is that
the material in different countries and areas of these countries is not uniform. In
order to achieve the unification of data the best method is to use a model of vertical
crustal movements. This model makes it possible to bring data into one measurement
epoch. In developing the model of vertical movement should as much as possible to
remove from the development of systematic factors, large and random errors.
Vertical crustal movements are caused by human activities and natural factors. To
determine the effect of individual factors is very difficult due to the complexity of
these factors, measurement periods and gaps in the data. Introduction to the methods
of measuring satellite observations gives wide possibilities for their use.
The existing system gives you the possibility to designate GRACE global gravitational
field changes over time, in a grid with a resolution of up to 100 kilometer. The
observations usually are presented as so-called changes millimeters. equivalent to a
layer of water. The GRACE palpable changes are the result of changes in Earth's mass.
Observations are available online as the development of the spherical harmonics in
the form of files of GSM. The twin GRACE satellites to send monthly observations of
terrestrial stations since 2002. Raw data, accessible directly from the measurement
require additional studies to remove the so-called destriping effect. For this purpose,
filters are used, from the simplest to cut the length of signs, to more sophisticated,
anisotropic, directly tailored to the structure GRACE signal.
So big amounts of data make it possible to use them in research on vertical
movements of the Earth's crust. As a test area selected Polish area as an area lying in
the central Europe. In the area of Poland three major tectonic units adjoins: eastern
european precambrian platform, young western and middle paleozoic platform and
carpatian region (MIZERSKI, 2002). In the joint of this three areas there is a
Teisseyre’a – Tornquist zone, which is geologically active in moderate degree
(MIZERSKI, 2002).
In the Poland so far we developed three maps vertical crustal movements. As a test
model was used numerical study carried out in 2006. The model was produced on the
basis of precise levelling performed within about 20 years.
As the second model from the GRACE data model was used filtered simplest filter,
Gaussian filter with a radius of 300 km filtration. Monthly used in analysis solutions
across the whole period of the GRACE mission. Two models confronted and calculated
correlations of these models.
76
GRACE satellites send information about gravity changes in time and space. Data is
received by five calculating centers (German GFZ, U. S. JPL and CSR, French CNES and
Dutch DEOS) where is preprocessed, processed and post processed. Than data is
presented on several levels of processing (in our study we used data fom the level 2).
Such data is an ideal source for Glacial Isostatic Adjustment determination, which
presents trend in water/ice flow over the years. Such approach is possible due to the
fact thant GRACE observation are changes in water mass and are resultant of
redistribution of lithospheric masses, adjusting from the glacial loading of the last ice
age. The aim of the study was to test a model of a vertical crustal movement and to
check its relation to a model of GIA from GRACE observations. GIA values are
determine in exact location of Kowalczyk model to make correlation computation
possible.
5.1. GRACE data and its usage for GIA determination
Gravity Recovery and Climate Experiment (GRACE) mission started in 2002

and gives an information about gravity changes in time which means about mass
changes on the Earth. Month GRACE solution contains:
 raw geophysical signals acquired over land and ocean and inside the
Earth,
 errors of an atmosphere,
 measurement errors ,
 and gravimetric data elaboration strategies.
GRACE data is published as a GSM files which contain ΔClm(t) and ΔSlm(t)
coefficients of gravity expansion. Models created on a basis of a raw GRACE data
need an additional processing. Month solutions are acquired with least square
method so they contain errors. Errors in GRACE observaions are caused by the fact
that GRACE satellites are placed about 400 kilometers above the Earth. The higher
altitude over the Earth, the weaker gravity signal, so the signal measured by the twin
satellites is not exact gravity that is on the Earth. This impairment can be noted
during mappind as stipes arranged in north-south direction, as this bias is some kind
of noise in observations. The only way of destriping GRACE observation is filtering
the spherical harmonics. The second problem in using GRACE data is lack of data in
particular months. Gaps in month observation is caused by the repeated orbit of the
GRACE floe which makes final grid of poor accuracy.
A core of understanding and calculating GIA are deformations of the solid Earth, so
vertical crustal movement. This can be precisely observed thanks to GPS system.
Additionaly, a by-product of calculating GIA is a relative sea level, which is known not
only now but whole history of the sea level. For this purpose tide gauses values are
used (for a present relative sea level) but also submerged shorelinesare researched
by observing salnity for getting to know the past. A product from the GRACE satellite
essential for the GIA modelllinig is gravity (WHITEHOUSE, 2009).
Glacial Isostatic Adjustmen of the litosphere and mantle in fact is some kind of trend
in a form of corrections that occures in GRACE raw observation. Reoving GIA from a
gradiometric model is essential when determining global or regional equivalent
water thickness (EWT), but also ocean bottom pressure, ic sheet masses (WAHR,
ZHONG, 2013).
77
GIA is computed in milimeters of equivalent water thickness (changes of EWT per
year) that can be expressed with a formula (Veselinova Renglova, 2007):
N ( ,  , t )  N ( ,  , t ) GRACE  N ( ,  , t ) C  N ( ,  , t ) O  N ( ,  , t ) L (24)
In the formula, δN(φ, λ,t)GRACE means geoid change computed with spherical
harmonic coefficients C l ,m (t ) and S l , m (t ) of degree l and order m. This is calculated
by(Chao and Gross, 1987):
 l
N ( ,  , r )  R   (C l ,m cos m  S l ,m sin m ) P l ,m (sin  ) (25)
l 2 m 0
The GIA model needs to be filteres, the same as the raw GRACE data. In the presented
GIA model Gaussian filtering was used.
The question is why we shoud take into cosidereation glaciation/deglaciation now,
why these parameters interrupt satellite gravimetric observations. The eccentricity,
precession and tilt of the Earth relative to the ecliptic staying in relation to the
Milankovich cycle are resultant of theice sheet and mountain glaciers movement over
the years (glaciacion and deglaciation epochs). Moreover, the last deglaciation period
still can be noted by observing changes of relative sea level – so, mass changes
observed by the GRACE satellites. Those changes are not negligible; in the northern
part of the globe are more than 1 cm per year (PELTIER 1999). On the other hand,
mass changes strongly affect gravity field changes in a for of long-term trend. But not
only gravitational potential (system of Earth, water and ice load) is changed but
topography as well. In the process of deglaciacion a clash of two forces is observed:
ice and water as a response, both of them cause topography and gravitational
potencial separately. The clue of the process is a perfest balance between those two
phenomenons. The most important part of modelling GIA is defining an equation of
sea level which in fact is a difference in time between the geoid and topography
(BARLETTA, 2013).
As the GIA model is based on a historical, not measured but assumed history of the
Earth, the uncertainity of the model is about 20%.
The effect of the Earth mass changes are gravity changes over month. All changes of
the Earth's gravity field is detected by the twin GRACE satellites as it causes a small
change in distance between ships. To be able to recognize a slight disorders, a system
of a high precision is needed. This is fullfilled by a K-band ranging system (microvawe
band K). A GPS system is responsible for synchronization of time unit of the KBR.
Moreover, for the purpose of communication with the tracking station on a ground,
the S-band transmitter is used. Using lol-low-satellite-to-satellite tracking (ll-SST) let
achieving a few microns accuracy. Not only GPS allows for determining satellite
position on the orbit, but also ground based laser stations. Becose of using all
mentioned system, all changes between two satellites can be determine with a very
high presision. This distance precision in connection with a speed can be recalculated
prcisely into relative changes between ships. Not-gravitational acceleration are
calculated with on-board accelerometers (TAPLEY 2004).
As it was mentioned, GRACE data is representing global mass changes. One of the
sources of such phenomenon are ice sheet uplift and melting. Ice sheet changes of a
lithosphere can be expressed in a form of the Glacial Isostatic Adjustment (GIA).
78
Changes character is isotropic as rock masses are strived to equilibrium. Blocks of
stiff and light lithosphere are flowing on plastic astenosphere – if encumbered, they
are immersed, and opposite. Lithosphere surfaces immersing and uplifting are called
lithosphere vertical movements.
In a study Earth geopotential extension into spherical harmonic coefficient up to 180
degree and order from CSR observing center was used. As the raw data need to be
smoothed due to stipes occuring in the results. We used Gaussian filtering with a
radiud 400 km to compute post-process equivalent water thickness.
A variance between real surface of the Earth dependant on time and equipotential
surface is described by a formula below. The formula is called the equation of sea
level (PELTIER,2001).
~ ~
S ( ,  , t )  O( ,  , t )[G( ,  , t )  R ( ,  , t )] (26)
~ ~
Where S is a relative sea level, G geoid perturbation, R the Earth radius and O –
oceanic function.
Liquid center of a globe and the glaciations history are essential for GIA
modeling. GIA models can show many internal processes and can be used for geoid
parameters estimation, as:
 knowledge about ice sheet dynamics, lithosphere thickness estimation –
continental deformations as a result,
 climate – a source of climate changes in a past and future changes prediction
as a result,
 sedimentation stages – a history of seas and oceans, future changes prediction
as a result,
 sea and ocean level changes monitoring by observing ice melting and uplift –
interpretation and prediction of an ice sheet movement as a result,
 geometrical data accuracy improvement – better estimation of the gravity in
such places as Greenland or Antarctic as a result,
 sea shore migration, future sea shore location as a result,
 seismic and volcanic activity (WHITEHOUSE, 2009).
The GIA modeling is based on three components:
 The Earth model,
 Pleistocene glaciations model,
 Equation of sea level.
The Earth model is needed for isostasy influence simulation (FARRELL and
CLARK, 1976). On a basis of the GRACE, the GIA was estimated and depicted with
usage the following models:
 Pleistocene glaciations model published by PAULSON ET AL., 2007,
 Ocean model published by MITROVICA ET AL, 2005,
 The Earth model published by PELTIER, 2004.
Glacial Isostatic Adjustment data prepared on a basis of GRACE satellite measurement
and the Pleistocene glaciations model was computed in a grid form with half-degree
step for the whole area of Poland. The result of GIA analysis made by the authors is
presented in a Fig. 43. A figure is prepared in the ArcView software in exactly the
same scale and linear interpolation spatial resolution (0,02 km x 0,02 km).
79
Fig. 43. Glacial Isostatic Adjustment from GRACE data in Poland (Poland contours from:
http://www.eea.europa.eu/data-and-maps/data/biogeographical-regions-europe-2008/zipped-
shapefile-format-vector-polygon).
Resource: elaborated by BIRYLO
On a basis of figure 43 it can be said that the values range of Glacial Isostatic
Adjustment in Poland is smaller than the crustal vertical movements, i.e. -1,5 ÷ -3 mm
per year. Most of values of the post-glacial rebound are -1,5 ÷-2,7 mm per year – Fig.
44 (which is very similar to the model of the crustal vertical movements). But, when
analyzing GIA model, there are not noted outliers values as in a the model presented
in Fig. 44.
Fig. 44. Statistics of the Glacial Isostatic Adjustment from GRACE data, standard deviation in
dashed line.
80
5.2. Vertical crustal movement in Poland
Vertical movements of the earth's crust can be determined from various data
of different accuracy, and with different periods. This can be surveying, geological and
other data. Periods of designations can reach hundreds or thousands of years or
months, years or decades. The accuracy may vary from tenths of a millimeter to
several tens of meters.
Geophysics and anthropogenic phenomenon are a source of a crustal vertical
movement. Geophysical source of movement are ice melting or ice uplift, water
storage, sedimentation, erosion, mountain movement, mineral changes in lithosphere.
On the other hand, human being causes exploitation of natural resources, drying
and/or irrigation, communication.
Determined models of survey (levelling and GPS data ) and mareografics data,
embody these reasons and they are models with the highest accuracy. Two kinds of
crustal movements are distinguished: relative (if we determine one stable benchmark
which becomes a reference point), and observed crustal movements (related to an
average sea level) (KAKKURI, 1987).
If the eustatic changes of global mean sea and ocean level are added to an observed
one, the crustal vertical movements will become related to the geoid (KOWALCZYK,
2006).
To develop a model of vertical movement, first set the vertical movements relative to
each line network. We identified 360 common lines of levelling. The relative vertical
movements on these lines are in the range from 1.5 mm / yr to 1.5 mm / year, with
the exception of Jaroslaw and Inowrocław neighborhoods. In order to obtain the final
relative vertical movements nodal benchmarks, the entire network has been
adiustment parametric method. The relative vertical movements on benchmarks
nodes are in the range of -2.0 mm / year in Jaroslaw to +1.5 mm / year in Nysa, with a
standard deviation of 0.1 mm / year. One of the benchmarks, benchmark
Władysławowo, was taken as the benchmark for vertical motion is zero. The choice of
this benchmark as a reference benchmark, according to its close position relative to
mareografhic point in Wladyslawowo, enabling easy exchange of aligned relative
vertical movement on vertical movements relative to the sea level.
Usually, long-term changes in average sea level (vertical movements mareografhic
point), is determined by linear regression. The study proposes a method of the
moving average This method is the most resistant to the choice of model parameters
from which to a large extent depend on the results.
Connect networks of vertical movement relative to mareogrfhic point in
Wladyslawowo made in two variants assuming both a steady sea level. In the first
vertical movement was adopted mareogrfhic station in Wladyslawowo, determined
from data 1955 - 1999, which amounted to -2.98 mm / year. In the second - the
average vertical movement of the four mareografhic stations in Swinoujscie,
Kolobrzeg, Ustka and Władysławowo, which amounted to 2.05 mm / year. The
vertical movement relative to the benchmark in Wladyslawowo mereographic point
rated at 0.5 mm / year, which is to replace aligned relative vertical movements nodal
benchmarks for movements relative to the mean sea level were added to the first
amendment -3.5 mm / year - the first option, and -2.5 mm / year - the second option.
The information obtained about the vertical crustal movements are located at points
81
distant from each other an average of 50 km and cover an area of the country in an
uneven manner. For more information about the vertical movements of the rest of
Polish, should be at the points where there are no benchmarks, the values of these
movements, interpolated, or create a suitable model. There are many interpolation
methods that can be used in this case. It was decided to the of least squares
collocation method as it is a relatively new method takes into account the physical
properties of the studied phenomenon, and is relatively easy to program. The
interpolation method of co-location is important to the function of covariance.
The empirical covariance function of the data set containing only vertical movements
benchmarks node (235 benchmarks) and data containing vertical movements on all
benchmarks (approx. 11 thous.). In both variants of the variances and correlation
distances they are similar, which means that there should be no significant
differences in determining the model of one or the other data. In order to perform
interpolation, empirical covariance function analytic function should be replaced.
Markov was used function and the function of Hirvonen. These functions are used to
calculate the next vertical movements models.
After deleting the calculated baseline models received four graphic versions of the
observed vertical movements of the Earth's crust in the area Polish. Models in
versions 1 and 2 were created using Markov covariance function with establishing a
network of vertical movement relative to the station mareograficznej Władysławowo
vertical motion adopted respectively: -3.0 mm / year and -2.0 mm / year. Model
version 3 and 4 were created using Hirvonen covariance function with establishing a
network of vertical movement relative to the station mareograficznej Władysławowo
vertical motion adopted respectively: -3.0 mm / year and -2.0 mm / year. W
otrzymanych modelach odchylenie standardowe wynosi: – z zastosowaniem funkcji
Markowa ±0,13 mm/rok, – z zastosowaniem funkcji Hirvonena ±0,05 mm/rok.
Analyzing the resulting models obtained values of vertical movements do not differ
much in the corresponding versions, ie. 1-3 and 2-4. In most of the country's values of
vertical movements are similar, and the course of isolines is similar. In all models,
there are no positive vertical movements. The biggest negative vertical movements
were found in the area Inowroclaw and Rzeszow (below -5 mm / year), both area lie
in the so-called. zone of tectonic T-T. In the vicinity of Warsaw vertical movements
range from -2 mm / year to -4 mm / year - versions 1 and 3, and from -1 mm / yr to -3
mm / year - versions 2 and 4. In most parts of Polish movements corresponding
range from -2.5 mm / year to -4 mm / year and between -1.5 mm / year to -3 mm /
year. Vertical movements above -4 mm / year (version 1 and 3) and -3 mm / year
(version 2 and 4) occur in the near Elblag (Zuławy Wiśsane), Plock, Torun and
Wloclawek (TT zone border) and in the area of the western Polish from Wroclaw to
Legnica by the Zielona Gora to Gorzow Wielkopolski (Variscides is an area on the
border with the Czech massif). The smallest of the negative movements in all models,
was recorded in the regions of southern Bieszczady (Carpathian).
82
Fig. 45. Spirit levelling data in Poland.
Resource: elaborated by KOWALCZYK
Kowalczyk model of the crustal vertical movements in Poland is depicted in Fig. 46.
As the ultimate model of vertical crustal movements in the area of Polish adopted the
model developed using analytical Hirvonen covariance function of establishing a
network of vertical movement relative to the mareographic station Władyslawowo
vertical motion adopted as -2.0 mm / year.
Fig. 46. Crustal vertical movements in Poland - the Kowalczyk model.

Resource: elaborated by BIRYLO (based on Kowalczyk 2006)
83
Using this model, and vertical movements at junctions created map reperach vertical
movements on Polish territory. This map has been confronted with the previous map
vertical movements. Vertical movements of the two designations overlap, particularly
in central and north - eastern Poland (near Elblag), as well as in the triangle of
Wloclawek, Plock, Toru The smallest vertical movements on both maps of the area of
the Bieszczady Mountains. In the vicinity of Warsaw there were no such large vertical
movements as in the earlier designation. They are formed from -1 mm / year to 2 mm
/ year (previously: -2.5 mm / yr to -4.1 mm / year). Similar values demonstrated in
and around the Nysa from 0 mm / year to -1 mm / year (earlier than -3 mm / year).
For further analysis used a set number of vertical movements interpolated grid points
of latitude and longitude 20'x20 'with which it is possible to interpolate the vertical
movements of the Earth's crust in the area of Polish at any point in the country.
The statistics of the Kowalczyk model are shown in a Fig. 47. It can be denoted that
most of Poland is affected with vertical movements in a range -1,5 ÷ -3 mm per year.
Fig. 47. Statistics of the Kowalczyk model of the crustal movements, standard deviation in dashed
line.
Distribution of value is similar to a Gaussian curve. For a complete statistical analysis

of the Polish level data can be found in [KOWALCZYK and RAPINSKI 2013].
5.3. A relation between vertical crustal movement in Poland and GIA in Poland
Combining GRACE observations with GPS data in a for of the vertical crustal
movement can give us a new GIA model. So here in the study the authors want to
compare GIA model with a vertical crustal movement model; both, gravity changes
and crustal movement are temporal.
The aim of the study was to compare the two models, those presented in Fig. 43 and
46 and check whether there is a relation between them. For this purpose, the
difference between the two model was computed in the ArcView software. The mean
value of the Kowalczyk model was -0,25 mm per year, when GIA model was analyze, -
2,2 mm per year mean value was noted. After differencing the two models, mean
value in the whole area of Poland was 0,3. Standard deviation of the two tested
models, according to the table 7, was 0,4. when difference between two models was
studied, standard deviation was 0,5.
84
Table 7. Minimum, maximum, mean values and standard deviation of three models: Kowalczyk
model, GIA model from GRACE and differences between Kowalczyk and GIA model
Crustal GIA Difference
Vertical from between
Moveme GRAC models
nt E 2-3
1 2 3 4
Min. -6,4 -3,6 -1,0
Max. -1,0 -0,2 4,1
Mean -2,5 -2,2 0,3
St. dev. 0,4 0,4 0,5
The result of the differencing of the models is presented in a Fig. 48. First of all, GIA
grid file was interpolated in such a way that values of GIA in exactly the same 235
points of spirit levelling, then a linear interpolation was made to cover with values
whole area of Poland. Statistics of the interpolation are presented in a Fig. 49.
Fig. 48. A difference between crustal vertical movements from spirit levelling and mareograph
data and GIA from GRACE data. (Poland contours from: http://www.eea.europa.eu/data-and-
maps/data/biogeographical-regions-europe-2008/zipped-shapefile-format-vector-polygon).
85
Fig. 49. Statistics of a difference betwen Kowalczyk model of the crustal vertical movements and
GIA from GRACE, standard deviation in dashed line.
From the Fig. 48 and 49 very interesting study can be made. A satisfying information
is that most of area of Poland is covered with values about 0,3 ( i. e. -0,5 ÷ 1) which
means that the crustal vertical movement model and GIA model are well correlated.
In a region of parts of lodzkie, mazowieckie, podlaskie i zachodnio-pomorskie
voivodship values gain -1 ÷ -0,5. The biggest differences (-2,1 ÷ -2,5) are nearby the
Baltic sea.
5.4. Conclusions
Obtained from the a comparison of the two models differences can be very helpful in
guide further work on vertical crustal movements in the area, not only the of the
Poland. The results they give are likely to be using to evaluate the models received
from the GNSS observations. Receives an initial correlation indicates the the above
use.
The aim of the study was to analyze Kowalczyk crustal vertical movement model fro
spirit levelling and Glacial Isostatic Model from GRACE observation. A relation
between the models and a basis of the research was the fact that a huge influence on a
crustal movement have glaciations and deglaciations. The authors want to take a
chance of finding a new, quick and efficient method of a crustal vertical movement
determination by using satellite data. The two models were compared, statistical
characteristics were computed. On a basis of analysis it can be concluded that GRACE
satellite gravimetric observations can provide crucial support in defining crustal
vertical movement determination. Moreover:
 the mean value of the Kowalczyk model is -0,25 mm per year, when GIA
model is -2,2 mm per year,
 mean value of the difference between the two models for the whole area of
Poland was 0,3 mm,
 standard deviation of the two tested models is 0,4, but for the difference of
the tested models standard deviation is 0,5,
 the crustal vertical movement model and GIA model are well correlated –
Poland is covered with values about 0,3 ( i. e. -0,5 ÷ 1),
86
In some parts of Poland (i.e. lodzkie, mazowieckie, podlaskie i zachodnio-pomorskie
voivodship) values are -1 ÷ -0,5, the biggest differences (-2,1 ÷ -2,5) are nearby the
Baltic sea.
The biggest differently (area Inowroclaw and Rzeszow) result from business in these
areas of human activity. Similarly, in the vicinity of large cities (Warsaw).
The resulting comparison can help in identifying locations that require deeper
analysis of geological or look for reasons for the differences
A deeper analysis will require also the vicinity of Gdansk and Elblag, where the
difference between the models is relatively large.
Measure taken subject can continue creating additional comparisons using GNSS data,
levelling data and mereografic data in the various configurations. More and longer
periods of satellite and gravimetric observation can contribute to the wider work on
identifying the correlation between the created models. Very dense grid gravity data
from the GRACE mission can also be used to determine the correlation on smaller
areas or those where there are large vertical movements of the earth's crust.
Acknowledgements
A project is financed by the National Science Center, agreement number: UMO-
2012/07/N/ST10/03275 form 10th July 2013.
87
REFERENCES
ACKERMANN, S., ANGRISANO, A., DEL PIZZO, S., GAGLIONE, S., GIOIA, C., TROISI, S.: 2013,
Digital Surface Models for GNSS Mission Planning in Critical Environments, Journal of
Surveying Engineering. 10.1061/(ASCE)SU.1943-5428.0000119. SSN (print): 0733-
9453. ISSN (online): 1943-5428.
APONTE, J., MENG, X., HILL, CH., MOORE, T., BURBIDGE, M., DODSON A. 2009. Quality
assessment of a network-based RTK GPS service in the UK. Journal of Applied Geodesy,
3: 25-34.
BABIJ, L., GLOTOV, V., OZGOVYCH A. 2006. Research of coupling of GPS device and film cameras.
Collection of scientific works "Recent achievements in geodesic science and industry".
Lviv, (II) : 161-166.
BAHMACH P.G., GLOTOV V.M., LADANIVS'KYJ, J.V., TRETJAK K.R. 2003. Geodynamic network of
Argentine islands adjacent to the Ukrainian Antarctic station Vernadsky. Ukrainian
Antarctic Journal. №1 : 149-151.
BAKUŁA, M., OSZCZAK, S., PELC-MIECZKOWSKA, R.: 2009, Performance of RTK Positioning in
Forest Conditions. Journal of Surveying Engineering, 135(3): 125-130, ASCE.
BAKUŁA, M., PELC-MIECZKOWSKA, R., WALAWSKI, M.: 2012, Reliable and Redundant RTK
Positioning for Applications in Hard Observational Conditions, Artificial Satellites,
Journal of Planetary Geodesy, 47(1-2012), 23-33.
BAKUŁA, M., PELC-MIECZKOWSKA, R.: 2011, Reliable Technique of GNSS/RTK Positioning
under Severe Observational Conditions, Reports on Geodesy, 2(91): 67-73.
BAKUŁA, M., PRZESTRZELSKI, P. 2013. Technologia wiarygodnego pozycjonowania RTK
GPS/GLONASS. Przegląd Geodezyjny, 7: 3-9.
BANASIK P. 1999. Wyznaczenie przebiegu quasi-geoidy w rejonie Krakowa na podstawie
pomiarów niwelacyjnych i GPS. Rozprawa doktorska. Kraków.
BANERJEE P., BOSE A., 1996, Evaluation of GPS PDOP from elevation and azimuth of satellites,
Indian Journal of Radio and Space Physics, 25:110-113.
BARLETTA V. R., BORDONI A., 2013. Effect of different implementationsof the same ice history
in Gia modeling, Journal of Geodynamics Vol 71, (2013) pp 65-73.
BOSY, J., ORUBA, A., GRASZKA, W., LEOŃCZYK, M., RYCZYWOLSKI, M. 2008. ASG-EUPOS
densification of EUREF Permanent Network on the territory of Poland. Reports on
Geodesy, 2(85): 105-112.
CEYLAN, A., INAL, C., SANLIOGLU I. 2005. Modern height determination techniques and
comparison of accuracies. FIG Working Week 2005 and GSDI-8, Cairo, Egypt April 16-
21, 2005.
CHAO, B. F., GROSS, R. S., 1987. Changes in the Earth's rotation and low-degree gravitational
field induced by earthquakes. Geophys. J. R. astr. Soc. 91, No. 3, 569-596. DOI:
10.1111/j.1365-246X.1987.tb01659.x.
COLCORD, J. E.: 1989, Using fish-eye lens for GPS site reconnaissance, Journal of Surveying
Enginiering 115(3):347-352. ASCE, ISSN 0733-9453.
CZARNECKI, K. 1994. Geodezja współczesna w zarysie. Wydawnictwo Wiedza i Życie,
Warszawa.
D’INCÀ, D., ZATELLI, P.: 2004, New modules for satellite surveying planning in GRASS, Geomat
Workb, 3:1–14.
DAWIDOWICZ, K. 2011. Wykorzystanie wirtualnych stacji systemu ASG-EUPOS do wyznaczania
wysokości w statycznych pomiarach GPS. Acta Scientiarum Polonorum seria Geodesia
et Descriptio Terrarum, 10(4): 21-34.
88
DAWIDOWICZ, K. 2012. GNSS satellite levelling using the ASG-EUPOS system services. Technical
Sciences, 15(1): 35-48.
DAWIDOWICZ, K. 2013. Analysis of height determination using the ASG-EUPOS NAWGEO
service. Technical Sciences, 16(1): 19 – 40.
DAWIDOWICZ, K., KRZAN, G., ŚWIĄTEK, K. 2014. Urban area GPS positioning accuracy using
ASG-EUPOS POZGEO service as a function of session duration, Artificial Satellites, 49(1):
33-42.
DODSON, A. H., SHARDLOW, P.J., HUBBARD, L.C.M., ELGERED, G., JARLEMARK, P.O.J. 1996. Wet
tropospheric effects on precise relative GPS height determination. Journal of Geodesy,
70: 188-202.
DOROZHYNS'KYJ O.L. 2002. Analytical and digital photogrammetry. Lviv Politechnik, 163 pp.
DOROZHYNSKYY O., MINILEVSKYY H., GLOTOV V.M. 2004. Photogrammetric research
conducted at the Antarctic station Academic Vernadsky. XXth ISPRS-2004 Congress,
Istambul - Turkey, pp. 642-644.
ECKLE, M. C., SNAY, R. A., SOLER, T., CLINE, M. W., MADER, G. L. 2001. Accuracy of GPS-derived
relative positions as a function of interstation distance and observing-session duration.
Journal of Geodesy, 75: 633-640.
EDWARDS, S. J., CLARKE, P. J., PENNA, N. T., GOEBELL, S. 2010. An examination of network RTK
GPS services in Great Britain. Survey Review, 42(316): 107-121.
EL-MOWAFY A. 2011. Analysis of web-based GNSS post-processing services for static and
kinematic positioning using short data spans. Survey Review, 43(323): 535-549.
ERHU, W., HUA, CH., ZHIGUO, A.. 2006. VRS Virtual Observations Generation Algorithm. Journal
of Global Positioning Systems, 5(1-2): 76-81.
EULER, H.J., KEENAN, C.R., ZEBHAUSER, B.E., WÜBBENA, G. 2001. Study of a Simplified
Approach in Utilizing Information from Permanent Reference Station Arrays. Proc. ION
GPS 2001, Salt Lake City.
FARRELL, W.E., CLARK, J.A., 1976. On postglacial sea level. Geophysical Journal International,
Geophysical Journal of the Royal Astronomical Society Vol. 46, Issue 3.
FEDERICI, B., GIACOMELLI, D., SGUERSO, D., VITTI, A., & ZATELLI, P.: 2013, A web processing
service for GNSS realistic planning, Applied Geomatics, 5(1), 45-57.
FEDOROWSKI W. 1974. Ewidencja gruntów (Land registration). Państwowe Przedsiębiorstwo
Wydawnictw Kartograficznych, Warszawa.
FIGURSKI, M. 2001. Określenie wybranych parametrów atmosfery na podstawie permanentnych
obserwacji GPS i ich wpływ na dokładność opracowań pomiarów satelitarnych. BEL
Studio Sp. Z o.o., Warszawa.
FIGURSKI, M., BOGUSZ, J., BOSY, J., KONTNY, B., KRANKOWSKI, A., WIELGOSZ, P. 2011. „ASG+”:
project for improving Polish multifunctional precise satellite positioning system.
Reports on Geodesy, 2(91): 51-58.
FIGURSKI, M., KAMIŃSKI, P., KROSZYŃSKI, K., SZAFRANEK, K. 2009. ASG-EUPOS monitoring
with referemce to EPN. Artificial satellites, 44(3): 85-104.
FIRUZABADI, D., KING, R. W. 2012. GPS precision as a function of session duration and reference
frame using multi-point software, GPS Solutions16(2): 191-196.
GEL'MAN, R.N. 1982. The accuracy of the measurements from the images obtained by non-
metric cameras. Geodesy and cartography, № 7 : 29-32.
GEL'MAN, R.N., NYKYTYN M.J. 2004. Application of the principle of phototheodolites with
terrestrial digital photography. Geodesy and cartography. – M : 32-37.
GLOTOV, V.M. 2002. Device of coupling of GPS and film cameras. Collection of scientific works
"Recent achievements in geodesic science and industry" – L, pp. 346-348.
GLOTOV, V.M. 2003. Features of determination of the focal length of digital photototeodolite
cameras. Interagency scientific and technical collection «Geodesy, cartography and
89
aerial photo shooting» Science. - Engineering. collection "Geodesy, Cartography and
aerosurveing, №63, pp. 112-127.
GLOTOV, V.M. 2004. Definition of the distortion of the digital film systems. Bulletin of the
National University of Water Management and Natural Resources. Part 1, Issue 4(28) :
207-213.
GLOTOV, V.M. 2005a. The research of the method of determining of the spatial coordinates of
the center projection of imaging camera. Collection of scientific works. Recent
achievements in geodesic science and industry – L, pp. 125-129.
GLOTOV, V.M. 2005b. Features of digital shooting while creating large-scale plans of the
Antarctic coastline near the station "Akademik Vernadsky". Science and Engineering.
„Bulletin of geodesy and cartography”, Kyiv, 3 : 22-25.
GLOTOV, V.M. 2005c. Method of placing of digital cameras according to the guide devices.
Collection of scientific works "Recent achievements in geodesic science and industry"
– L'viv, (I) : 248-252.
GLOTOV, V.M. MAJOROV G.J. 2000a. Analysis of the metric properties of digital filming systems.
// Interagency scientific and technical collection. Geodesy, cartography and aerial
photo shooting., 60: 102-106.
GLOTOV, V.M., KALINOVS'KA, O.P., PUL'KEVYCH, I.G., SMIRNOV, J.I. 1996. Determining of the
spatial coordinates of the points of panoramic shooting using the methods of kinematics
design. Engineering geometry and engineering graphics., 59 : 43-47.
GLOTOV, V.M., MAJOROV G.E. 2000b. Method of calibrating of digital cameras. Bulletin of the
2nd Int. Scientific and Practical Conference. - Lviv-Krakow, pp. 163-169.
GLOTOV, V.M., PUL'KEVYCH, I. 2006. Definition of the coordinates of the points of digital
shooting using the methods of kinematics design. Modern achievements of the geodesic
science and industry. Collection of scientific works, Lviv, pp. 176-183.
GRABEK P. 2014. Badanie przemieszczeń na osuwisku w Kłodnem metodą fotogrametrii
lotniczej (Research in displacements on landslide in Kłodne based on aerial
photogrammetry). Praca dyplomowa. Akademia Górniczo Hutnicza w Krakowie.
GREKU, R.H. 2000. Study on the international project SCAR Geodetic Infrastructure in
Antarctica. Bulletin of the Eng. Antarctic Centre, Kiev, 3 : 141-149.
GROVES, P. D.: 2011. Shadow Matching: A New GNSS Positioning Technique for Urban
Canyons, Journal of Navigation 64(3):417-430.
Guide on topographyc shooting in a scale of 1: 5000, 1: 2000, 1: 1000, 1: 500. Phototeodolyte
shooting - M.: Nedra. 1977. - 136 p.
HADAŚ, J., BOSY, J. 2009. Niwelacja satelitarna GNSS z wykorzystaniem serwisu NAWGEO
system ASG-EUPOS. Acta Scientarum Polonorum Geodesia et Descripto Terrarum,
8(2): 53-66.
HÄKLI, P. 2006. Quality of Virtual Data Generated from the GNSS Reference Station Network.
Proc. of the XXIII FIG Congress, Germany, Munich, October 8-13.
HIGGINS, M. B. 2010. Heighting with GPS: possibilities and limitations.
http://www.fig.net/commission5/reports/gavle/hig.
HOFMANN-WELLENHOF, B., LICHTENEGGER, H., WASLE, E. 2008. GNSS – Global Navigation
Satellite Systems. Springer-Verlag Wien, Austria.
HOFMANN-WELLENHOF, B., MORITZ, H. 2005. Physical Geodesy. SpringenWienNewYork.
HUGEMANN, W.: 2010, Correcting lens distortions in digital photographs, Ingenieurbüro
Morawski+ Hugemann: Leverkusen, Germany.
KADAJ R. 2010. Application of Schreiber’s type difference observation system for elaboration of a
session of static GPS measurements. Biuletyn Wojskowej Akademi Technicznej, LIX(2):
85-106.
KADAJ, R., ŚWIĘTOŃ, T. 2009. Algorytm i oprogramowanie modułu automatycznego
postprocessingu (APPS) w polskim systemie satelitarnych stacji referencyjnych (ASG-
90
EUPOS). Zeszyty Naukowe Politechniki Rzeszowskiej, Budownictwo i Inżynieria
Środowiska, 51(262): 37—56.
KAKKURI, J., 1987. Character of the fennoscandinavian land uplift in the 20th century,
Geological survey of Finnland, Special Paper 2, Helsinki.
KANNALA, J., BRANDT, S. S.: 2006, A generic camera model and calibration method for
conventional, wide-angle, and fish-eye lenses, Pattern Analysis and Machine
Intelligence, IEEE Transactions on, 28(8), 1335-1340.
KAPLAN E. D., HEGARTY Ch. J. 2006, Understanding GPS, Principles and Applications, Artech
House Inc. ISBN-10: 1-58053-894-0.
KASSER, M. 2002. New Modern Height Determination Techniques. Report about the WG 5.2
Activities in 1998–2002, FIG XXII International Congress, Washington, D.C.USA, April
19–26 2002.
KOWALCZYK K, 2006. New model of the vertical crustal movements in the area of Poland,
Geodezija ir Kartografija, Vilnius, No. 4 vol XXXII, pp. 83-87, DOI:10.1080/13921541.
2006.9636702.
KOWALCZYK, K., TYMOSZCZYK, T. 2012. Opracowanie trójwymiarowego modelu dna stawu w
aplikacji SURFER przy użyciu danych z pomiaru GPS RTK. Zeszyty Naukowe
Politechniki Rzeszowskie, 283(59): 197-206.
KOWALCZYK K., RAPIŃSKI J, 2013. Evaluation of levelling data for use in vertical crustal
movements model in Poland Geodynamica et Geomaterialia, Vo. 10, No.4 (172),
pp.401-410, DOI: 10.13168/AGG.2013.0039.
KRYŃSKI, J., ŁYSZKOWICZ, A. 2006a. Regional quasigeoid determination in the area of Poland.
5th FIG regional Conference for Africa, Accra, Ghana, 8-11 March.
KRYŃSKI, J., ŁYSZKOWICZ, A. 2006b. Centimetre quasigeoid modelling in Poland using
heterogenous data. IAG Proceedings from 1st International Symposium of the
International Gravity Field Service (IGFS), 28 August – 1 September, Istanbul, Turkey.
KUDRYS, J., KRZYŻEK, R. 2011. Analysis of Coordinates Time Series Obtained Using the NAWGEO
Service of the ASG-EUPOS System. Geomatics and Environmental Engineering, 5(4):
39-47.
KURCZYŃSKI Z. 2014. Fotogrametria (Photogrammetry). Wydawnictwo Naukowe PWN SA,
Warszawa
LANDAU, H., VOLLATH, U., CHEN, X. 2003. Virtual Reference Stations versus Broadcast Solutions
in Network RTK – Advantages and Limitations. GNSS, Graz, 2003,
http://gootechmabna.com/catalouges/Papers/3sencore/15.pdf.
LINDA G. SHAPIRO AND GEORGE C. STOCKMAN (2001): “Computer Vision”, pp 279-325, New
Jersey, Prentice-Hall, ISBN 0-13-030796-3LUHMANN, T., ROBSON S., KYLE S., HARLEY
I.: 2006, Close-range Photogrammetry.
LOBANOV, A.N. 1968. Phototopography. Nedra, pp 268 p.
ŁYSZKOWICZ, A. 2009. Oszacowanie dokładności quasi-geoidy z modelu EGM08 na obszarze
Polski. Acta Scientarum Polonorum Geodesia et Descripto Terrarum, 8(4), 39-47.
MA, L., CHEN, Y., & MOORE, K. L. (2004). Rational radial distortion models of camera lenses
with analytical solution for distortion correction. International Journal of Information
Acquisition, 1(02), 135-147.
MITROVICA, J.X., WAHR, J., MATSUYAMA, I., PAULSON, A., 2005. The rotational stability of an
Ice Age Earth, Journal of Geophysical Int., 161, 491-506.
ORUBA, A., RYCZYWOLSKI, M., WAJDA, S., BOSY, J., GRASZKA, W., LEOŃCZYK, M., SOMLA, J.
2008. The Implementation of the Precise Satellite Positioning System ASG-EUPOS. In
ENC-GNSS 2008 Conference Proceedings, EUGIN, Toulouse (France).
PARKINSON, B. W., SPILKER, J. J. 1996. Global Positioning System: Theory and Applications. Vol.
I, American Institute of Areonautics and Astronomics, Inc., Washington, USA.
91
PAULSON, A., ZHONG, S., WAHR, J., 2007. Inference of mantle viscosity from GRACE and
relative sea level data. Journal of Geophysical Int.171, 497–508. doi: 10.1111/j.1365-
46X.2007.03556.x.
PAŻUS, R. 2011a. Aktywnie i wirtualnie z POZGEO-D. Magazyn geoinformacyjny GEODETA,
1(188): 20 – 24.
PAŻUS, R. 2011b. Aktywnie i wirtualnie z EUREF. Magazyn geoinformacyjny GEODETA, 4
(191): 24 – 28.
PELC-MIECZKOWSKA R.: 2014, Primary results of using hemispherical photography for
advanced GPS mission planning, 9th International Conference “Environmental
Engineering”, Procedia Engineering 01/2014.
PESCI, A., LODDO, F., CENNI, N., TEZA, G., CASUALA, G. 2008. Analyzing Virtual Reference
Station for GPS surveying: experiments and applications In a test site of the northern
Apennines (Italy). Annals of Geophysics, 51: 619-631.
PELTIER, W.R., 2004. Global Glacial Isostasy and the Surface of the Ice-Age Earth: The ICE-
5G(VM2) model and GRACE, Ann. Rev. Earth Planet. Sci., 32, 111-149.
PELTIER, W.R., 2001. Global Glacial Isostatic Adjustment and modern instrumental records of
Relative Seal Level history, Elsevier Inc. International Geophysics Volume 75.
PELTIER W. R., DRUMMOND R., Roy K., 2012. Comment on “Ocean mass from GRACE and
glacial isostatic adjustment by D. P. Chambers et al., J. Geophys. Res., 117, B11403,
doi:10.1029/2011JB00896.
PSIMOULIS, P. A., KONTOGIANNI, V. A., NICKTITOPOULOU, A., PYTHAROULI, S. I.,
TRIANTAFYLLIDIS, P., STRIOS, S. C. 2004. Estimating the Optimum Duration of GPS
Static Observations for Short Baseline Length Determination in Greece. Proceedings of
the FIG Working Week, Athens, Greece, 22-27 May.
REMONDINO F., FRASER C., 2006: Digital camera calibration methods: considerations and
comparisons. IAPRS Volume XXXVI, Part 5, Dresden 25-27 September 2006.
RETSCHER, G. 2002. Accuracy Performance of Virtual Reference Station (VRS) Networks.
Journal of GPS, 1(1): 40-47.
SIEJKA, Z. 2009. Wykorzystanie pomiarów GNSS do wyznaczania współrzędnych podstawowej
osnowy realizacyjnej na terenach oddziaływań górniczych. Archiwum Fotogrametrii,
Kartografii i Teledetekcji, 19: 387-396.
SIEJKA, Z., GLOTOV, V.M. 2012. The accuracy of estimation of geodetic coordinates in a mode
GPS-RTK using a Trimble R8. Geoinformational monitoring of environment GPS and
GIS technologies GPS. Coll. Science. Mater. XVII Int. scientific-technical. symposium.
Alushta - Lviv, pp. 34-39.
SNAY, R. A., SOLER, T. 2008. Continuously Operating Reference Station (CORS): History,
Applications, and Future Enhancements. Journal of Surveying Engineering, 134(4): 95-
104.
TAMISIEA M.E., MITROVICA J. X., DAVIS J. L., 2007. GRACE gravity data constrain ancient ice
geometries and continental dynamics over Laurentia . SCIENCE 316 881 – 883.
TAYLOR, G., LI, J., KIDNER, D., BRUNSDON, C., & WARE, M.: 2007, Modeling and prediction of
GPS availability with digital photogrammetry and LiDAR, International Journal of
Geographical Information Science, 21(1), 1-20.
TOPCON POSITIONING SYSTEM. 2006. Topcon Tools User’s Guide. Topcon Positioning Systems
Inc, May.
VANICEK, P., KRAKIWSKY, E., 1982. Geodesy, the concepts North-Holland.
VESELINOVA RENGLOVA E., 2007. A Dynamic Geoid Model for Canada. A thesis submitted to
the Faculty of graduate studies in partial fulfillment of the requirements for the
degree of doctor of philosophy. UCGE Reports Number 20261, 18.
VOLCHKO, P.I., GLOTOV V.M., CHYZHEVS'KYJ V.V. 2005. Technology of creation of large-scale
plans of local objects using digital stereophotohrammetric methods. Collection of the
92
materials of the tenth scientific and technical symposium “Geoinformational
monitoring of the environment - GPS i GIS – technologies'”.- L'viv, pp. 133 – 135.
WAHR A. G. J., Zhong S., 2013. Computations of the viscoelastic response of a 3-D compressible
Earth to surface loading: an application to Glacial Isostatic Adjustment in Antarctica
and Canada, Geophys. J. Int., 192, 557–572, doi: 10.1093/gji/ggs030.
WANNINGER L. 2003. Virtual reference stations (VRS). GPS Solutions, (7): 143-144.
WANNINGER, L. 1997. Real-Time Differential GPS-Error Modeling in regional Reference Stations
Networks. Proc. IAG Symp 118, Rio de Janerio, 86-92.
WANNINGER, L. 1999. The Performance of Virtual Reference Stations in Active Geodetic GPS-
networks under Solar Maximum Conditions. Proc. ION GPS 99, Nashville TN, 1419-
1427.
WANNINGER, L. 2002. Virtual Reference Stations for Centimeter-Level Kinematic Positioning.
Proc. of ION GPS 2002, Portland, Oregon, 1400-1407.
WEGENER, V., WANNINGER, L. 2005. Communication Options for Network RTK. International
Association of Geodesy. Working Group 4.5.1: Network RTK. Available at:
http://www.network-rtk.info/wegener/communication.html.
WHITEHOUSE, P., 2009. Glacial Isostatic adjustment and sea level change, State of the art
report, Technical Report TR-09-11, Durham University.
WIŚNIEWSKI, B., BRUNIECKI, K., MOSZYŃSKI, M. 2013. Evaluation of RTKLIB's Positioning
Accuracy Using low-cost GNSS Receiver and ASG-EUPOS. The International Journal on
Marine Navigation and Safety of Sea Transportation, 7(1): 79-85.
YANLAK, M., BAYKAL, O. 2001. Transformation of elipsoid heights to local levelling heights.
Journal of Surveying Engineering. 127(3): 90-103.
LEGAL REGULATIONS
DEKRET z dnia 2 lutego 1955 r. o ewidencji gruntów i budynków (Decree on land and buildings
register). Journal of Laws No. 6, item 32.
DEKRET z dnia 24 września 1947 r. o katastrze gruntowym i budynkowym (Decree on the land
and building cadastre). Journal of Laws No. 61, item 344.
INSTRUKCJA Ministra Gospodarki Komunalnej z dnia 8 września 1956 r. w sprawie zakładania
i prowadzenia ewidencji gruntów, zasad sporządzania wykazów gruntów oraz opłat za
odrysy z map, za odpisy i wyciągi z rejestrów i dokumentów stanowiących część operatu
ewidencyjnego na obszarze m. st. Warszawy, m. Łodzi oraz miast stanowiących powiaty
(Instruction of the Minister of Public Utilities of 8 September 1956 on establishing and
running land register, rules for preparing land records and fees for drawing copies of
maps and for copies and excerpts from records and documents which are components of
the cadastral survey in the area of the capital city of Warsaw, the city of Lodz and the
cities which are poviats). Official Journal No.98, item 1135.
INSTRUKCJA Ministra Rolnictwa z dnia 21 kwietnia 1955 r. w sprawie zakładania i
prowadzenia ewidencji gruntów, zasad sporządzania wykazów gruntów oraz opłat za
odrysy z map i za odpisy oraz wyciągi z rejestrów i dokumentów stanowiących część
składową operatu ewidencyjnego na obszarze gromad, osiedli i miast nie stanowiących
powiatów (Instruction of the Minister of Agriculture on establishing and running land
register, rules for preparing land records, and fees for drawing copies of maps and for
copies and excerpts from records and documents which are components of the cadastral
survey in the areas of settlements, housing estates and cities which are not poviats).
Official Journal No.38, item 379.
93
INSTRUKCJA TECHNICZNA G-2. 2001. Szczegółowa pozioma i wysokościowa osnowa
geodezyjna i przeliczanie współrzędnych miedzy układami. Główny Urząd Geodezji i
kartografii, Warszawa.
PROJEKT NOWELIZACJI ROZPORZĄDZENIA RADY MINISTRÓW Z DNIA 8 SIERPNIA 2000 R. W
SPRAWIE PAŃSTWOWEGO SYSTEMU ODNIESIEŃ PRZESTRZENNYCH (Dz. U. Nr. 70,
poz. 821).
ROZPORZĄDZENIE Ministra Administracji i Cyfryzacji z dnia 29 listopada 2013 r. zmieniające
rozporządzenie w sprawie ewidencji gruntów i budynków (Regulation of the Minister of
Administration and Digitization amending the Regulation on the land and property
register). Journal of Laws item. 1551.
ROZPORZĄDZENIE Ministra Budownictwa z dnia 6 grudnia 1950 r. w sprawie postępowania
przy organizacji, zakładaniu i prowadzeniu katastru gruntowego i budynkowego
(Regulation of the Minister of Construction was issued on the procedure regarding
organizing, establishing and running of land and buildings cadastre). Journal of Laws
No. 57, item 520.
ROZPORZĄDZENIE Ministra Rozwoju Regionalnego i Budownictwa z dnia 29 marca 2001 r. w
sprawie ewidencji gruntów i budynków (Regulation of the Minister of Regional
Development and Construction of 29th March 2001 on the land and property register).
Journal of Laws No.38, item 454, as amended.
ROZPORZĄDZENIE Ministrów Gospodarki Przestrzennej i Budownictwa oraz Rolnictwa i
Gospodarki Żywnościowej z dnia 17 grudnia 1996 r. w sprawie ewidencji gruntów i
budynków (Regulation of Ministers of Spatial Planning and Construction and of
Agriculture and Food on the register of land and buildings). Journal of Laws No. 158,
item 813.
USTAWA z dnia 13 października 1998 r. przepisy wprowadzające ustawy reformujące
administrację publiczną (Act of 13 October 1998 – the provisions introducing the acts on
reforming public administration). Journal of Laws No.133, item 872, as amended.
USTAWA z dnia 17 maja 1989 r. Prawo geodezyjne i kartograficzne (The Act of 17th May 1989,
Geodetic and Cartographic Law). Consolidated text - Journal of Laws of 2010 No.193,
item 1287 as amended.
USTAWA z dnia 23 kwietnia 1964 r. Kodeks Cywilny (The Act of 23rd April 1964 Civil Code).
Consolidated text - Journal of Laws of 2014, item 121, as amended.
WYTYCZNE TECHNICZNE G-1.12. 2008. Pomiary satelitarne oparte na systemie precyzyjnego
pozycjonowania ASG-EUPOS.
http://geogis.com.pl/download/wytyczne_g_1_12_21_04_2008_1.pdf.
WYTYCZNE TECHNICZNE G1-10. 2001. Formuły odwzorowawcze i parametry układów
współrzędnych. Główny Urząd Geodezji i Kartografii, Warszawa.
ZALECENIA TECHNICZNE: POMIARY SATELITARNE GNSS OPARTE NA SYSTEMIE STACJI
REFERENCYJNYCH ASG-EUPOS. 2011.
http://www.asgeupos.pl/webpg/graph/standards/Zalecenia_ASG_EUPOS_20110207.
pdf.
ZARZĄDZENIE Ministrów Rolnictwa i Gospodarki Komunalnej z dnia 20 lutego 1969 r. w
sprawie ewidencji gruntów (Ordinance of the Ministers of Agriculture and Public
Utilities on the land register). Official Journal No 11, item.98, as amended.
94
WEB SITES
Beidou Navigation Satellite System, http://en.beidou.gov.cn, access: 23.02.2015.
European Environment Agency, http://www.eea.europa.eu/data-and-
maps/data/biogeographical-regions-europe-2008/zipped-shapefile-format-vector-polygon
GPS World, http://gpsworld.com/the-almanac, access: 23.02.2015.
GPS.gov, http://www.gps.gov, acces: 23.02.2015.
GRACE Tellus, http://grace.jpl.nasa.gov
Information Analytical Centre, Federal Space Agency, http://glonass-iac.ru/en/, acces:
23.02.2015
95
LIST OF FIGURES
Fig. 1. Measurement points locations. ..................................................................................................... 9
Fig. 2. Calibration box. .................................................................................................................................. 10
Fig. 3. The entrance pupil point movement for Sigma 4.5 mm f/2.8 EX DC lense. ........ 11
Fig. 4. Polynomial fitting.............................................................................................................................. 12
Fig. 5. Hemispherical images: on left - point between the buildings, on right - point in
the forest. ............................................................................................................................................................. 12
Fig. 6. Binary images created on the dasis of hemispherical photograps.......................... 13
Fig. 7. Obtained point clouds. .................................................................................................................... 13
Fig. 8. Diagram showing the transformation process. ................................................................. 14
Fig. 9. Binary images created on the basis of laser scanning.................................................... 14
Fig. 10. MP+ application - main window with data loaded. ...................................................... 15
Fig. 11. MP+ application - model settings window........................................................................ 16
Fig. 12. MP+ application - content of the *.xmo file....................................................................... 17
Fig. 13. The principle of creating an obstacles based on binary values............................. 18
Fig. 14. Obstacles model dataset schema. .......................................................................................... 18
Fig. 15. MP+ application - SVS/PDOP resulting graphs............................................................... 19
Fig. 16. Predicted number of available satellites and PDOP value without obstacles
model. ..................................................................................................................................................................... 20
Fig. 17. Predicted number of available satellites and PDOP values for point C: a)
obstacles model from laser scanning, b) obstacles model from hemispherical image.
................................................................................................................................................................................... 21
Fig. 18. Predicted number of available satellites and PDOP value for point F: a)
obstacles model from laser scanning, b) obstacles model from hemispherical image.
................................................................................................................................................................................... 22
Fig. 19. Predicted number of satellites used by positioning algorithm and PDOP value:
a) point C, b) point F. ...................................................................................................................................... 23
Fig. 20. Sky plot of satellites on point F: a) All satellites, b) registered by the receiver
c) used by algorithm ....................................................................................................................................... 24
Fig. 21. Sky plot of satellites on point C: a) All satellites, b) registered by the receiver
c) used by algorithm ....................................................................................................................................... 24
Fig. 22. The same pice of land in the photo taken: a) in April, b) in August ..................... 33
Fig. 23. The same piece of land in photographs taken on April 26: a) in 2009, at about
1.5 hours later than b, b) in 2012. ........................................................................................................... 34
Fig. 24. the same piece of land in images taken in different years – change to the
appearance of the boundary between the dirt road and agricultural land. ...................... 35
Fig. 25. The upper and lower edge of the slope: a) unrecognizable in a single
photograph, b) being at the same time the boundary of the agricultural land................ 36
Fig. 26. The upper and lower edge of the slope: a) unrecognizable in a single
photograph, b) being at the same time the boundary of the agricultural land................ 38
Fig. 27. Dependence of the root mean square error (RMS) of the stereogram absolute
orientation on ground sampling distance. .......................................................................................... 39
Fig. 28. Trajectory of a prototype of point A..................................................................................... 42
Fig. 29. Connection of GPS to EXO Flow-chart. ................................................................................ 50
Fig. 30. General view of the complex .................................................................................................... 51
96
Fig. 31. Fixation of coordinates by a GPS receiver (42-57) and fixed photography
points (37-41).................................................................................................................................................... 52
Fig. 32. Marks of planimetric and altimetric preparation, located on the coastline .... 54
Fig. 33. Digital image taken from baseline of shoreline (tilted survey case) ................... 57
Fig. 34. Survey of the island Skua shoreline from the motorboat using the complex . 59
Fig. 35. Geometric levelling........................................................................................................................ 62
Fig. 36. Trigonometric levelling............................................................................................................... 63
Fig. 37. GPS levelling...................................................................................................................................... 64
Fig. 38. ASG-EUPOS ........................................................................................................................................ 66
Fig. 39. Results obtained for two selected points........................................................................... 68
Fig. 40. Normal height differences (geometric levelling - GPS levelling)........................... 70
Fig. 41. Obstacles on measured points: a) A point, b) B point, c) C point......................... 71
Fig. 42. RMS height determination statistics .................................................................................... 72
Fig. 43. Glacial Isostatic Adjustment from GRACE data in Poland ......................................... 80
Fig. 44. Statistics of the Glacial Isostatic Adjustment from GRACE data, standard
deviation in dashed line. .............................................................................................................................. 80
Fig. 45. Spirit levelling data in Poland.................................................................................................. 83
Fig. 46. Crustal vertical movements in Poland - the Kowalczyk model. ............................. 83
Fig. 47. Statistics of the Kowalczyk model of the crustal movements, standard
deviation in dashed line. .............................................................................................................................. 84
Fig. 48. A difference between crustal vertical movements from spirit levelling and
mareograph data and GIA from GRACE data.. ................................................................................... 85
Fig. 49. Statistics of a difference betwen Kowalczyk model of the crustal vertical
movements and GIA from GRACE, standard deviation in dashed line. ................................ 86
97
LIST OF TABLES
Table 1. Values of ASE of spatial coordinates when changing distance ............................. 49
Table 2. The values of time-delays and error of GPS ................................................................... 53
Table 3. Basic specification for digital cameras .............................................................................. 57
Table 4. Values of MSE control points obtained according to calculations in
phototriangulation and according to the proposed algorithm ................................................ 60
Table 5. ASG-EUPOS services ................................................................................................................... 67
Table 6. Average height diferences (geometric levelling - average NRTK levelling) for
first test area ...................................................................................................................................................... 69
Table 7. Minimum, maximum, mean values and standard deviation of three models:
Kowalczyk model, GIA model from GRACE and differences between Kowalczyk and
GIA model............................................................................................................................................................. 85
98
NOTES ON THE AUTHORS
Michał Bednarczyk, PhD,
University of Warmia and Mazury in Olsztyn, Poland
Faculty of Geodesy, Geospatial and Civil Engineering
Department of Land Surveying and Geomatics
Research interests:
The use of computer science in geodesy such as: GIS
applications and analyses, spatially enabled databases, web
applications, software development and engineering. Mobile
technologies in geodesy, applications for mobile systems
and devices. Programming in: Object Pascal,VBA, Java, C,
Python, PHP, Ruby.
Monika Birylo, Ph.D

Univeristy of Warmia and Mazury in Olsztyn
Faculty of Geodesy Geospatial and Civil Engineering
Research interests:
Evaluating and usage of gradiometric satellite missions,
especially GRACE mission.
Karol Dawidowicz, PhD,

Institute of Geodesy
Scientific editor of this book.
Research interests:
An application of GNSS measurement techniques in geodesy,
surveying and navigation, especially for precise height
determination. Currently my main research interests
include issues related to GNSS antenna phase center
corrections (PCC) modeling and precise point positioning
(PPP).
99
Volodymyr Hlotov Assoc. Prof.
University of Agriculture in Krakow, Poland
Department of Land Surveying
Faculty of Environmental Engineering and Land Surveying
Research interests:
Photo theodolite survey and research elements interior
orientation of photo theodolite, aerial survey works to
determine the deformation process of hydraulic structures
and volumes of mining mass masses and volumes of mining
in open mine.
Currently the subject of research is the problem of
kinematic processes in various fields of science and
technology.
Kamil Kowalczyk, Ph.D

Univeristy of Warmia and Mazury in Olsztyn
Faculty of Geodesy Geospatial and Civil Engineering
Research interests:
Crustal verical movement in Poland, kinematic vertical
referece system, RTK and RTN measurements, engineering
measurement, interdisciplinary in geodesy
Anita Kwartnik-Pruc, Ph.D.

AGH University of Science and Technology, Poland
Department of Geoinformation Photogrammetry and
Remote Sensing of Environment
Faculty of Mining Surveying and Environmental Engineering
Research interests:
Real estate economy, with particular emphasis on geodetic
and legal procedures implemented with the participation of
surveyors.
In 2010, she acquired professional level 2 qualifications to
perform independent functions in the field of geodesy and
cartography. She is the author of over 40 scientific
publications. She is the Member of the Polish Real Estate
Scientific Society and of the Association of Polish Surveyors.
The study has been carried out with financial support from the
statutory research No 11.11.150.949 and 11.11.150.005 AGH
University of Science and Technology.
100
Renata Pelc-Mieczkowska, PhD,
Research interests:
An application of GNSS measurement techniques in geodesy,
surveying and navigation, especially under severe
observational conditions.
Zbigniew Siejka Ph.D.

University of Agriculture in Krakow, Poland
Department of Land Surveying
Faculty of Environmental Engineering and Land Surveying
Research interests:
In his research he deals with the problems of positioning
accuracy and reliability using satellite positioning and
navigation GNSS systems in Geodesy and related disciplines.
He is conducting research on the numerical elaboration of
high-precision integrated with the use of permanent
reference stations including the setting out networks for
linear objects. He has professional qualifications in the field
of geodesy and cartography in the area of situation and
altimetric measurements and elaboration of their results.
Dariusz Tomaszewski, MsC.

The Faculty of Geodesy, Geospatial and Civil Engineering
Institute of Geodesy
Research interests:
In his research he deals with satellite measurement
techniques. The main topic of his study is the development
of algorithms used for integrated navigation (GPS/INS).
101
Andrzej Wróbel, Ph.D.
AGH University of Science and Technology, Poland
Department of Geoinformation Photogrammetry and
Remote Sensing of Environment
Faculty of Mining Surveying and Environmental Engineering
Research interests:
Photogrammetric documentation of architectural
monuments, engineering photogrammetry, industrial
surveying, remote sensing, termography.
The study has been carried out with financial support from
the statutory research No 11.11.150.949 and 11.11.150.005
AGH University of Science and Technology.
102
View publication stats

Book3 2015

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Book3 2015

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Modern Geodetic Techniques in Spatial Measurement

Book · September 2015

Pelc-Mieczkowska Renata Karol Dawidowicz

SEE PROFILE SEE PROFILE

Michał Bednarczyk Monika Birylo

SEE PROFILE SEE PROFILE

Baltic Geodetic Congress (Geomatics) View project

Water Tower in Ciechanów View project

The user has requested enhancement of the downloaded file.

MODERN GEODETIC TECHNIQUES IN

Zagreb, Croatia, 2015

ISBN 978-953-6129-45-4 Nacionalna knjižnica, Zagreb, Croatia

1.1. Object and methods

To perform the study of hemispherical photography suitability for GNSS

Fig. 1. Measurement points locations.

At both measurement points vertical, hemispherical images were executed

1.2. Camera set calibration

Hemispherical images suffer from large amount of distortion due to

Fig. 2. Calibration box.

For calibration model determination, the odd-order polynomial radial distortion

Calculated polynomial coefficients were applied to MP+ application described

1.3. Hemispherical image

The hemispherical photographs shown below were taken at measuring

To be suitable for Terrain obstacles modelling colour images should be

Fig. 6. Binary images created on the dasis of hemispherical photograps.

1.4. Laser scanning data processing

On the basis of performed measurements two point clouds representing

Fig. 7. Obtained point clouds.

Fig. 8. Diagram showing the transformation process.

On the basis of the previously obtained cartesian coordinates the images of

Fig. 9. Binary images created on the basis of laser scanning.

Fig. 10. MP+ application - main window with data loaded.

Preparing and loading the image

Setting up model and correction parameters

Fig. 11. MP+ application - model settings window.

Fig. 12. MP+ application - content of the *.xmo file.

Fig. 14. Obstacles model dataset schema.

Fig. 15. MP+ application - SVS/PDOP resulting graphs.

1.8. Additional conclusions

As it was previously mentioned while performing research on advanced

2.2. Current regulations on determining the cadastral data

2.3. Procedures for determining the course of real estate boundaries

As it was mentioned above, the course of boundaries of record parcel is included in

2.4. Aerial images - interpretation

Recognition of a fence or hedge is relatively simple. The measurement of land

2.5. Capture of spatial data from aerial photographs

Fig. 28. Trajectory of a prototype of point A.

Resource: elaborated by HLOTOV.

Based on the derived equations it is necessary to determine the types of trajectories of

x  x 02  x01 , y  y 02  y01 , z  z02  z01 . 

into (3.1) we obtain the equation:

 x  V0   x, z   cos arctgk z    y  y 01   0, (8)

In general, when the path, which А or Р pass during time t, is equal V0  t  a  t n , so

(re-projecting from a moving plane π on the immovable planes of projections xy and

where ν - the coefficient of linear dimension.

3.1. Determination of coordinates of digital photogrammetric survey points

for second image:

Thus, the introduction of three additional parameters (speed of projection centre

The above formulas differ from classical by presence of arguments of velocity

3.2. Topographical plans of the Vernadsky Antarctic station

Analysing topokinematic works in Antarctica, it should be noted that Russia

3.3. Device Connecting EXO with GPS