Professional Documents
Culture Documents
Practical Manual
PREPARED BY:
Kalid Hassen
REVIEWED BY:
Elias Cherenet (Ass prof)
SEPTEMBER, 2020
HARAMAYA, ETHIOPIA
Prepared By: Kalid Hassen
Table of Contents
INTRODUCTION ..................................................................................................................... 1
1.7.1 Introduction to the principles of the Analytic Hierarchy Process (AHP) .................19
1.9.2 Refine the optimal areas using Majority Filter tool ......................................................28
i|Page
Prepared By: Kalid Hassen
ii | P a g e
Prepared By: Kalid Hassen
6.4 Adding fields to the attribute table and filling necessary data’s ............................... 78
6.4.1 Length.....................................................................................................................................78
iii | P a g e
Prepared By: Kalid Hassen
Geostatistical Analyst............................................................................................................... 89
Soil Erosion Assessment using Revised Universal Soil Loss Equation (RUSLE) ................ 105
8.8 soil loss A calculation from the RUSLE equation ................................................... 120
Algorithm for Automated Mapping of Land Surface Temperature Using LANDSAT 8 Satellite
Data ........................................................................................................................................ 122
iv | P a g e
Prepared By: Kalid Hassen
Supervised Image classification Using support Vector Machine (SVM) .............................. 132
v|Page
Prepared By: Kalid Hassen
INTRODUCTION
Dear trainee, I am very glad to be able to introduce a spatial analysis and Environmental
modelling. Spatial analysis is in many ways the crux of GIS, because it includes all of the
transformations, manipulations, and methods that can be applied to geographic data to add
value to them, to support decisions, and to reveal patterns and anomalies that are not
immediately obvious in other words, spatial analysis is the process by which we turn raw data
into useful information. If GIS is a method of communicating information about the Earth’s
surface from one person to another, then the transformations of spatial analysis are ways in
which the sender tries to inform the receiver, by adding greater informative content and value,
and by revealing things that the receiver might not otherwise see. Some methods of spatial
analysis were developed long before the advent of GIS, and carried out by hand, or by the use
of measuring devices like the ruler and spatial analysis using GIS is in many ways its logical
successor.
This training manual aims to guide students how to explore ArcGIS Desktop software produced
by ESRI (Environmental System Research Institute) for different spatial analysis. The manual
intended to help identify different spatial problems and solutions. Each section in the manual
introduces a general group of functions in ArcGIS, providing step by step instructions for using
a set of tools with screen captures showing those steps.
This manual tries to cover the presentation and implementation under ArcGIS and its spatial
analysis tool box, Geostatistical Extension, Archydro Tool and Landsat Tool was used through
case studies dealing with site selection using both Vector and raster data’s, Finding alternative
route Analysis, hydrological analysis using Hydrology toolset and Archydro Tool,
Geostatistical Analysis, Network analysis and RUSLE soil erosion model, Land surface
Temperature modeling as well.
Hence, I recommend interested students, staffs, researchers and all concerned to use this
manual to familiarize themselves with spatial Analysis and modelling in GIS in particular to
their field of specialization. I always appreciate feedback, whether you found a typo or spelling
mistake or want to suggest a better way of explaining particular concepts and techniques. The
best way to succeed with GIS is to make learning how to use it a collective process.
Kalidh84@gmail.com
Kalidh84@yahoo.com
Kalid.hassen@haramaya.edu.et
Kalid Hassen, Haramaya University
1|Page
Prepared By: Kalid Hassen
2|Page
Prepared By: Kalid Hassen
Assuming you are employed as GIS analyst in the urban planning department in Dire
Dawa city, Ethiopia, as a GIS expert you are asked to develop site suitability map of the
town to support the future urban development planning activity proposed by the planning
department. The planning department is interested to know information on suitable area
for School in the city. This is because the existing urban structure is not enough to
accommodate the demand of the population hence by constructing new school the
associated problem in the town will be minimized. The department also set criteria to fulfill
for the candidate site to be selected. The final suitable site to be selected for building new
school should not to be located on steep slope, forest land, and not near to existing school.
On the other side the planning department wanted to locate the new school near to
recreational sites and roads infrastructure. In addition to this the size of the final site should
be 2000 square meters. Based on the above stated criteria and requirement, find the most
suitable site for building new school. Therefore, based on the project requirement develop
a site suitability model consisting of a detailed procedure to select the best site as stated in
the criteria. During the weighted overlay in most case the most important parameter will
get higher influence than the least important and optionally we can also set to use equal
influence but it should be fully justified. The common scale to be used during reclassifying
the raster data will be 1 to 10 by 1. We will use this scale range to combine all the input
data set during weighted overly to select the best/ suitable site.
Our input data set
Land use Raster dataset representing the land-use types over the area
DEM Raster dataset representing the elevation of the area
School location Feature class representing point locations of existing schools
Recreational sites Feature class representing point locations of recreation sites
Roads Feature class representing the linear road network for the city of
Dire
City boundary Indicate the extent of the city
3|Page
Prepared By: Kalid Hassen
1.1.2 Description
In this exercise, you will prepare for analysis then creating a geodatabase for your results.
Next, you will derive slope, distance to recreation sites, and distance to existing schools,
and then reclassify these derived datasets to a common scale from 1 to 10. You will drive
weight them using AHP method in ArcGIS_Ahp extension and the land use dataset
according to a percentage of influence and combine them to produce a map displaying
suitable locations for the new school. Finally, you will select the optimal site for the new
school from the alternatives.
Steps:
4. Click the Connect to folder button in the Select the map's geodatabase window.
5. Browse to and click the working copy of the Spatial Analyst folder just created.
6. Click OK.
7. Click the New File Geodatabase button.
8. Name the new file geodatabase Scratch.
4|Page
Prepared By: Kalid Hassen
9. Click Add.
10. Click OK
Both our current and scratch workspaces are set to your Scratch.gdb geodatabase. For this
workflow, access data from the Dire Dawa.gdb, which is in the Spatial Analyst folder and
contains your data; write your data, by default, to your Scratch.gdb.
Steps:
1. Click the menu Geoprocessing → Environments.
2. Click Workspace to expand the environment settings related to workspaces.
3. For Current Workspace, navigate to your Dire Dawa.gdb in your Spatial Analyst folder.
4. Click Add.
5. Click OK.
• your scratch workspace is already set to your Scratch.gdb geodatabase
5|Page
Prepared By: Kalid Hassen
6|Page
Prepared By: Kalid Hassen
You will first create a new toolbox to hold the models will create in this exercise & the
next.
Steps:
1. Create a new toolbox in your Spatial Analyst folder. Name the toolbox Site Analysis.
7|Page
Prepared By: Kalid Hassen
You will create a model to perform Spatial Analyst tasks. A model is built by stringing
tools together in ModelBuilder. Once your model is created, we can easily experiment with
parameter values, use different input data, run the model over and over again, and share it
with others.
Steps:
1. Right-click the Site Analysis Tools toolbox and click New > Model.
Steps:
8|Page
Prepared By: Kalid Hassen
4. Check the Store relative path names (instead of absolute paths) check box.
5. Checking this box sets all source paths referenced by the tool as relative to the
location of the toolbox; therefore, if the model is moved to a different directory it
will still be able to be executed successfully.
II. Specifying environment settings
Before start to perform analysis on our data, should set any relevant environment settings.
For more information on how to set environments and the hierarchy between analysis
environments, go to the analysis environment of the ArcGIS Spatial Analyst extension.
Because your environment settings apply to each process of this model only, you will set
environment settings for the model.
Steps:
1. Click the Environments tab.
2. Expand Processing Extent and check Extent.
3. Expand Raster Analysis and check Cell Size.
Tip: The Current Workspace and the Scratch Workspace are already set, as these
environment settings are inherited from the map document.
4. Click Values.
9|Page
Prepared By: Kalid Hassen
5. Expand Processing Extent. Set the Extent by click the drop-down and selecting same
as Layer elevation.
6. Expand Raster Analysis. Set Cell Size by click the drop-down arrow and selecting
same as Layer elevation. The cell size of your elevation layer will be applied to all
subsequent raster outputs. Your elevation dataset has the largest cell size (30 meters).
Caution: Setting a smaller cell size than your largest input will not mean we have more
detailed information in subsequent raster results; we will just have more cells of the same
value, which may affect your display and calculation speeds. Although the software does
not prevent it, it is considered incorrect to set a cell size smaller than your largest input
cell size.
7. Click OK on the Environment Settings window.
10 | P a g e
Prepared By: Kalid Hassen
11 | P a g e
Prepared By: Kalid Hassen
6. Right-click the output variable from the Slope tool and click Rename.
7. Type Slope Output and click OK.
Caution: Renaming an element label does not alter the name of the output on disk. A layer
will be added to the Table of Contents called Slope Output which references data on disk
called Slope out.
IV. Deriving distance
i. From recreation sites
To find locations close to recreation sites, must first calculate the Euclidean distance from
recreation sites.
Steps:
1. Hover the pointer over the Euclidean Distance tool connected to rec_sites. We can
easily see all the default parameters set for this tool. There is no need to adjust any of
these parameters.
2. Accepted the default for the Maximum distance, thus leaving this parameter empty.
Therefore, the edge of the output raster is used as the maximum distance. The Output
cell size is taken from the environment setting previously set to that of our elevation
12 | P a g e
Prepared By: Kalid Hassen
13 | P a g e
Prepared By: Kalid Hassen
Deriving datasets, such as slope, is the first step when building a suitability model. Each
cell in our study area now has a value for each input criteria (slope, land use, distance to
recreation sites, and distance to schools). Need to combine the derived datasets so can
create our suitability map that will identify the potential locations for the new school.
Hover, it is not possible to combine them in their present form for example, combining a
cell value in which slope equals 15 degrees with a cell value for land use that equals 5
(forest) and get a meaningful answer that can compare to other locations. To combine the
datasets, they first need to be set to a common measurement scale, such as 1 to 10. That
common measurement scale is what determines how suitable a particular location each
cell is for building a new school. Higher values indicate more suitable locations for the
school. Using the weighted Overlay tool, can weigh the values of each dataset, and then
combine them. Hover, the inputs for the weighted Overlay tool must contain discrete,
integer values. Land-use data is already categorized into discrete values; for example,
forest equals a value of 5, so we can add this dataset directly into the weighted Overlay
tool and assign each cell a new value on the common measurement scale of 1 to 10 (you
do this later in the tutorial). The values in the datasets you derived in previous steps are all
floating-point, continuous datasets, categorized into ranges, and they must first be
reclassified so that each range of values is assigned one discrete integer value. Potentially,
the value given to each range can be any number, provided we note the range that the value
corresponds to. This is because can weight these values within the weighted Overlay tool
14 | P a g e
Prepared By: Kalid Hassen
the next step after reclassifying the derived datasets. Hover, it is easier to weight the cell
values for derived datasets while reclassifying. In the weighted Overlay tool, can accept
the default and leave the scale values the same as the input values will reclassify each
derived dataset to a common measurement scale, giving each range a discrete integer value
between 1 and 10. Higher values will be given to attributes within each dataset that are
more suitable for locating the school.
Steps:
1. Locate the Reclassify tool in the Spatial Analyst Toolbox Reclass toolset. Click and
drag the Reclassify tool onto ModelBuilder in line with Slope Output. Add another
Reclassify tool in line with Distance to recreation sites and another in line with
Distance to schools.
2. Click the Add Connection tool. Use the connect tool to connect:
a. Slope Output to the Reclassify tool
b. Distance to recreation sites to the Reclassify (2) tool
c. Distance to schools to the Reclassify (3) tool
d. On the model toolbar, click the Select tool.
e. Click the Auto Layout button, and then click the Full View button.
I. Reclassifying slope
It is preferable that the new school site be located on relatively flat ground. We'll reclassify
the slope output, slicing the values into equal intervals. We'll assign a value of 10 to the
most suitable range of slopes (those with the lost angle of slope) and 1 to the least suitable
range of slopes (those with the steepest angle of slope) and linearly rank the values in
between.
Steps:
1. Open the Reclassify tool connected to the Slope Output variable.
2. Accept the default for the Reclass field parameter so the Value field will be used.
3. Click Classify.
4. Click the Method drop-down arrow and click Equal Interval.
5. Click the Classes drop-down arrow and click 10.
6. Click OK.
7. Click Reverse New Values.
• Reversing the values applies higher new values to the values representing less steep
slope, since these areas are more suitable for building.
15 | P a g e
Prepared By: Kalid Hassen
16 | P a g e
Prepared By: Kalid Hassen
9. Rename the output variable from the Reclassify tool to Reclassed distance to
recreation sites.
Tip: To resize the output variable element, Reclassed distance to recreation sites, click the
element and move the mouse pointer over one of the blue handles surrounding the element,
then click and drag to resize the element so all text can be seen.
III. Reclassifying distance to schools
It is necessary to locate the new school away from existing schools to avoid encroaching
on their catchment areas. you will reclassify the Distance to schools’ layer, assigning a
value of 10 to areas farthest from existing schools (the most suitable locations), assigning
a value of 1 to areas near existing schools (the least suitable locations), and ranking the
values in between linearly. By doing this, you will determine which areas are near and
which areas far from exist schools.
Steps:
1. Open Reclassify (3) tool.
2. Accept the default for the Reclass field parameter so the Value field will be used.
3. Click Classify.
4. Set the Method to Equal Interval and the number of Classes to 10.
5. Click OK.
• We want to position the school away from existing schools, so we will assign larger
numbers to ranges of values that represent locations farther away, because these
locations are most desirable. Since the default assigns high New values (more
17 | P a g e
Prepared By: Kalid Hassen
suitable locations) to higher ranges of Old values (locations farther away from
existing schools), we do not need to change any values at this time.
6.Accept the default for the Output raster parameter.
7.Click OK.
8. Rename the output variable from the Reclassify (3) tool to Reclassed distance to
schools.
IV. Executing Reclassify
Steps:
1. Right-click each of the variable outputs Reclassed slope, Reclassed distance to
recreation sites and Reclassed distance to schools and click Add To Display.
2. Click the Run button to execute the three Reclassify tools in your model.
3. On the toolbar, click the Save button.
4. Examine the layers added to were ArcMap display.
5. Locations with higher values (with low gradient slopes, close to recreation sites, and
away from existing schools) are more suitable than locations with lower values (with
steeper slopes, far from recreation, and close to existing schools).
18 | P a g e
Prepared By: Kalid Hassen
The Analytic Hierarchy Process (AHP) was introduced by SAATY [1] and is a very
popular means to calculate the needed weighting factors by help of a preference matrix
where all identified relevant criteria are compared against each other with reproducible
preference factors.
The following table shows a simple comparison matrix of order 3 where 3 criteria C1, C2
and C3 are compared against each other. In the direct comparison of the criteria C1 and
C3, for example, criterion C1 has been regarded strongly more important than C3, hence
a value of 5 has been assigned to the corresponding matrix position. The transpose position
automatically gets a value of the reciprocal value, in this case 1/5 which equals 0.2.
19 | P a g e
Prepared By: Kalid Hassen
Next, the assigned preference values are synthesized to determine a ranking of the relevant
factors in terms of a numerical value which is equivalent to the weights of the factors.
Therefore, the eigenvalues and eigenvectors of the square preference matrix revealing
important details about patterns in the data matrix are calculated. The above square matrix
of order three gives three eigenvalues with which three eigenvectors each having three
vector components can be calculated. It is regarded sufficient to calculate only the
eigenvector resulting from the largest eigenvalue since this eigenvector contains enough
information to provide by its eigenvector components the relative priorities of the factors
being considered [2].
Of course, the values of the pair-wise comparison matrix will normally be well considered
and not set arbitrarily. However, people’s feelings and preferences remain inconsistent and
intransitive and may then lead to perturbations in the eigenvector calculations. Such
inconsistencies might be of the form that a factor Ai being preferred over another factor
Aj with Aj being preferred over a factor Ak is not preferred over Ak (Ai must be preferred
over Ak in this case). Therefore Saaty [1] provided the consistency ratio CR which is a
single numerical index to check for consistency of the pair-wise comparison matrix. It is
defined as the ratio of the consistency index CI to an average consistency index RI, thus
𝑅𝐼
CR =
𝐶𝐼
Values for RI [2] (with n = order of matrix)
(n) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
(RI) 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 1.51 1.48 1.56 1.57 1.59
Following the illustrations of Saaty [1] who calculated RI values up to a matrix order of
15, matrices with an order greater than 8 have an RI order of magnitude value of about
1.45.
20 | P a g e
Prepared By: Kalid Hassen
The consistency index CI can be directly calculated from the preference matrix with
𝜆 𝑚𝑎𝑥 − 𝑛
CI =
n−1
Where
λmax: greatest eigenvalue of preference matrix
n: order of matrix
Based on expert knowledge and experience, Saaty & Vargas [2] recommend a revision of
the preference matrix if the consistency ratio CR exceeds a value of 0.1.
1.7.2 Using AHP extension
Steps:
1. Push the button. A window appears which lets you define all criteria which you
consider relevant for your decision.
Please note: The extension expects classified raster datasets. Therefore, only integer
rasters will appear as list items in the classified map raster layers list box. You must take
care that all raster’s are classified within the same class value range (no checking is
performed if
the classified criteria raster’s have the same class range).
2. Click on a list item and the push the button to define it as a criterion which will
be used in the subsequent analysis.
21 | P a g e
Prepared By: Kalid Hassen
3. Please note that you can change the descriptor of the used layers
4. If you are finished with your criteria definitions click on the button.
The following window appears:
5. Enter your preference values in the preference matrix by double clicking on the
corresponding matrix cell or by hitting the <Enter> key. A window will appear asking
you to enter a value.
22 | P a g e
Prepared By: Kalid Hassen
• Having entered a valid value, the transpose position in the matrix is automatically
inverted. Elements on the matrix’ main diagonal will remain unchanged. Entering
values higher than the maximum preference value (which is 9) will be reset to 9
values smaller or equal to 0 will be set to one. The ‘pref. table’ button displays a
window which shows a ‘legend’ of the preference values.
6. To calculate the weights, push the button. The calculated weights are displayed
in the text box which is underneath the compute button. The button lets you view the
details of the calculation in terms of the resulting eigen values and more.
• The given preference values are a numerical way to express the importance or
dominance of one criterion over another. Here, the criterion slope is regarded
strongly more important than criterion school (value 7 in the first row); you can
see that criterion Slope receives by far the highest weight 58%, LULC gets about
14 %, distance from recreational site 20 %.
Note: The consistency ratio ‘CR’ of 0.04 is less than Saaty’s recommended threshold
of 0.1, therefore the result is acceptable.
23 | P a g e
Prepared By: Kalid Hassen
7. Push the ‘OK’ button. If you want a report file to be created check ‘create report file’
(default); then specify a filename.
You are now ready to combine the reclassified datasets and land use to find the most
suitable locations. The values of the reclassified datasets representing slope, distance to
recreation sites, and distance to schools have all been reclassified to a common
measurement scale (more suitable cells have higher values). The landuse dataset is still in
its original form because we can weigh the cell values for this dataset as part of the
weighted overlay process. Slope values that are less than 4 (the least suitable because they
are too steep) as restricted so these values can be excluded. If all datasets reclassify equally
important, we could simply combine them, giving each equal influence; hover, we have
been informed that it is preferable to locate the new school close to recreational facilities
and away from other schools. You will weight all the inputs, assigning each a percentage
of influence based on previous step. The higher the percentage, the more influence a
particular input will have in the suitability model. We will assign the inputs the following
percentages of influence:
Steps:
24 | P a g e
Prepared By: Kalid Hassen
9. Add the land use layer, this time setting the Input field to LANDUSE. Click OK.
• You will now weight the scale values of the landuse layer so they are comparable
with the other inputs. A lower value indicates that a particular land-use type is less
suitable for building.
10. Change the default Scale Values for the landuse layer to the following values:
• Shrub---5
• Barren land---10
• Built up---3
• Agriculture---9
• Forest---4
25 | P a g e
Prepared By: Kalid Hassen
11. We'll now assign a percentage of influence to each raster; based on how much
importance (or weight) each should have in the final suitability map.
12. In the % Influence column, type the percentages for each of the input raster’s:
• Reclassed distance to schools to 7
• Reclassed distance to recreation sites to 20
• Reclassed slope to 58
• land use to 15
Tip: Move the mouse pointer over the name of an input raster to view the entire name.
13. Accept the default for the Output raster parameter.
14. Click OK.
Steps:
1. Click the Auto Layout button, and then click the Full View button.
2. Rename the output variable from the weighted Overlay tool to Suitable Areas, click
OK.
3. Right-click the Suitable Areas variable and click Add to Display.
4. Run the weighted Overlay tool.
5. On the toolbar, click the Save button.
• Examine the layer added to ArcMap display. Locations with higher values indicate
more suitable sites areas that are on less steep slopes of suitable land-use types,
closer to recreational facilities, and away from existing schools. Notice that the
areas we marked as restricted have a value of zero.
On our layer, each pixel has a value that indicates how suitable that location is for a new
school. Pixels with the value of 9 are most suitable, and pixels with the value of 0 are not
suitable. Therefore, the optimal site location for a new school has the value of 9. Another
criterion for an optimal location is the size of the suitable area. A suitable location would
include several pixels with value of 9 being connected.
26 | P a g e
Prepared By: Kalid Hassen
You will use a conditional expression in the Con tool to extract only the optimal sites. It
has been decided that those sites that are considered optimal must have a suitability value
of 9 (the highest value in the suit areas output). In the conditional expression, all areas with
a value of 9 will retain their original value (9). Areas with a value of less than 9 will be
changed to NoData.
Steps:
1. Click and drag the Con tool, located in the Conditional toolset, into ModelBuilder.
2. Open the Con tool.
3. Click the Input Conditional Raster value drop-down arrow, click the Suitable Areas
variable.
4. Type the condition Value = 9 for the value of the Expression parameter.
5. Click the Input true raster or constant value drop-down arrow and click the Suitable
Areas variable. If the condition we enter is true, then the value of the cells of the Input
true raster or constant value will be applied to the cells of the output raster.
6. Leave the value for the Input false raster or constant value parameter blank. Leaving
the Input false raster or constant value parameter blank will apply the default. The
default is that if any value in the Input conditional raster that doesn't meet the
condition, we enter will be assigned NoData in the output raster.
7. Accept the default for the Output raster parameter.
8. Click OK.
9. Click the Auto Layout button, and then click the Full View button.
10. Rename the output variable from the Con tool to optimal areas and click OK.
11. Right-click optimal areas and click add To Display.
27 | P a g e
Prepared By: Kalid Hassen
Examine the layer added to your ArcMap display. These are the optimal site locations for
the new school. There are many single cells representing optimal locations. These 30-
meter cells are too small for the school site. You will clean up the result, removing these
small areas, using the Majority Filter tool.
Steps:
1. Click the Majority Filter tool, located in the Spatial Analyst Tools Toolbox
Generalization toolset, and add it to ModelBuilder.
2. Open the Majority Filter tool.
3. Click the Input raster drop-down arrow and click the optimal areas raster variable.
4. Accept the default Output raster parameter.
5. Click the Number of neighbors to use drop-down arrow and click EIGHT.
• This option specifies the number of neighboring cells to use in the kernel of the
filter. The kernel of the filter will be the eight nearest neighbors (a 3-by-3 cell
window) to the present cell.
6. Accept the default to use the MAJORITY as the Replacement threshold. Using
MAJORITY as the Replacement threshold means five out of eight connected cells
must have the same value for the present cell to retain its value.
7. Click OK.
8. Rename the output variable from the Majority Filter tool to Filtered optimal areas.
9. Right-click the Filtered optimal areas and click Add To Display.
28 | P a g e
Prepared By: Kalid Hassen
We have discovered the optimal sites for building the new school. All the locations in the
Filtered optimal areas layer are suitable. The last step in this exercise is to locate the best
site out of the alternatives. Then you will locate the best site based on area. An optimal
school site is 20000 square meters square meters. You will first convert the Filtered
optimal areas raster to a feature class inside a geo database so you can use the area field
that is generated. You will use the select Layer by Attribute tool identify the optimal site
from the alternatives, based on area.
29 | P a g e
Prepared By: Kalid Hassen
I. Rasterizing
Steps:
1. Open the Raster to Polygon tool in the Conversion Tools toolbox, inside the From
Raster toolset.
2. Click the Input raster drop-down arrow and click the Filtered optimal areas raster
layer.
3. Leave the default for the Field parameter as VALUE.
4. Accept the default Output polygon features parameter path, but change the name
5. Leave the default checked to Simplify polygons. The polygons will be simplified to
reduce the stair-step effect when a raster is converted to a polygon.
6. Click OK.
30 | P a g e
Prepared By: Kalid Hassen
Now we have located a new site with 20000 square meters for School according to the
city urban office.
31 | P a g e
Prepared By: Kalid Hassen
32 | P a g e
Prepared By: Kalid Hassen
Dire Dawa Health bureau officials decided to build new hospital, when building a new
hospital, there are criteria that should be followed in order to find the most appropriate
location. These criteria consist of required proximities or distances from certain features.
Suppose, as a town planner, you have been assigned the task of finding the potential sites
for new hospital construction.
2.1.2 Description
In this exercise, you will able to develop suitable areas for hospital construction in Dire
Dawa city based on Vector spatial Analysis. Necessary input data are:
Existing Hospital Feature class representing the existing hospital in the study area
Residential area Feature class representing residential areas over the city
2.2 Buffering
Buffer is a zone around a map feature measured in units of distance or time. A buffer is
useful
for proximity analysis. In this exercise you need to buffer each of the shapefiles in order
to make them meet the criteria for hospital site selection.
Tip: find area of road and residential within 100 meter surrounding the city and 500 meters
away from Existing hospitals.
Steps:
1. Add Exsisting_hospital, residential and roads from your data directory
2. From the Analysis tools of ArcToolbox expand the Proximity toolset,
3. Double Click on the Buffer tool element from the proximity toolset
33 | P a g e
Prepared By: Kalid Hassen
• Repeat this with each of the remaining shapefiles, setting the distance based on the
criteria stated above.
Now that road and residential shapefiles have been buffered to only show the portions that
are applicable for hospital site selection, we need to merge these shapefiles with criteria to
be within a certain distance (residential, and road).
34 | P a g e
Prepared By: Kalid Hassen
Steps:
1. From the Data Management tools of ArcToolbox expand the General toolset,
2. Double Click on the Merge tool element from the General toolset
2.4 Erase
You have to exclude the land that is within a certain distance of existing hospitals; we must
remove these areas from the shapefile we created in the previous step.
Steps:
1. From the ArcToolbox window expand the Analysis Tools toolbox
2. Expand the Overlay toolset
35 | P a g e
Prepared By: Kalid Hassen
5. In the Input features dialogue box; use the dropdown arrow and select Shapefile
created in the previous step as an input data set.
6. In the Erase Features select Existing hospital 500km buffer using the dropdown arrow
Click on the browse fodder icon and store the output feature class
7. Click ok when finished.
2.5 Clipping
Clipping allows you to turn one Shapefile into a cookie cutter in order to cut out part of a
larger Shapefile. In the following exercise you extract suitable hospital site found in Dire
Dawa.
Steps:
1. From the ArcToolbox Window, Expand the “Analysis” tools.
2. Expand the Extract toolset on “Clip” tool element.
In the input Clip dialog box do the following arrangements.
a. In the input Features, use the dropdown arrow and select suitable hospital site
b. In the Clip Features box, use the dropdown arrow and select “Dire_Dawa”
c. In the output feature class option, click on the browse folder and save the data
36 | P a g e
Prepared By: Kalid Hassen
37 | P a g e
Prepared By: Kalid Hassen
38 | P a g e
Prepared By: Kalid Hassen
Assuming you are hired as GIS expert in the Haromaya woreda Transport Authority, the
office finds out a problem on the road which takes from bate to Harar city which there is
a huge traffic jam in Awady Town, causes wastage of fuels and air pollution as well it
increases stress and frustration among motorists and passengers. Unsafe driving was
observed as a main impact of traffic jam which may lead to road mishaps and hence
injuries, as a GIS expert you are asked to find alternate access route in this Area.
3.1.2 Description
In this exercise, you'll find the best route for a new access road from the Bate Town
location site to a nearby road intersection. The steps you might follow to produce such a
path are outlined below.
• Create the source dataset if necessary. The source is the Bate Town.
• Create the cost dataset by deciding which datasets are required, reclassifying them to a
common measurement scale, weighting them, then combining them.
• Perform cost distance analysis using the source and cost datasets as inputs. The distance
dataset created from this tool is a raster in which the value of each cell is the accumulated
cost of traveling from each cell back to the source.
1. To find the least costly path, you need a direction dataset, which can be created as an
additional dataset using the Cost Distance tool. This gives you a raster of the direction of
the least costly path from each cell back to the source (in this exercise, Bate Town).
• Create the destination dataset if necessary. In this exercise, the destination dataset is a
point at a road junction.
• Perform cost path analysis using the distance and direction datasets created with the Cost
Distance tool.
39 | P a g e
Prepared By: Kalid Hassen
You will create a new model, Find Best Route, inside our Site Analysis Tools toolbox. This
model will calculate the best path through the landscape from the source (Bate Town) to
the destination point, taking into consideration the slope of the land and the type of land
use the path will cross.
40 | P a g e
Prepared By: Kalid Hassen
Initially you will create and Import the source dataset such as Bate Town and a slope
dataset. Next you will create the cost dataset that will identify the cost of traveling over
the landscape from any location back to the bate town, based on the fact that it is costlier
to traverse steep slopes and construct a road on certain land-use types.
This section of your model will look like the following:
41 | P a g e
Prepared By: Kalid Hassen
42 | P a g e
Prepared By: Kalid Hassen
5. Click the Add raster row button. This time, add the landuse layer. Set the Input field
to Land use and click OK.
6. You will weigh the scale values of the land use layer so they are comparable with our
Reclassed slope dataset. A higher value indicates that it is costlier to build a road
through a particular land-use type.
7. Type the following Scale Values:
Shrub—3
Barren land—1
Built up—5
Agriculture—2
Forest—4
8. You will now assign an equal percentage of influence to each raster, since they are
both equally important in this analysis. Click Set Equal Influence to assign an equal
percentage of influence (50 percent each) to each input raster.
9. Accept the default for the Output Raster.
10. Click OK.
3.4 Finding the least costly path
To find the least costly path between the bate town, and the destination site, the Cost
Distance tool and the Cost Path tool will be used. There are two outputs from the Cost
Distance tool. The Output distance raster shows the accumulated cost of traveling from
any location (or cell) to the source (bate town). It does not contain information on which
way to travel from a particular cell to the source, just how much it will cost to get there
following the least costly path. The Output backlink raster shows which way to travel from
any cell, following the least costly path back to the source. Using these outputs as inputs
to the Cost Path tool, along with input destination data, you will calculate the least cost
path between the bate town and the destination site.
I. Setting up the model
Steps:
1. Add the Bate Town and the destination layers to ModelBuilder.
2. Add the Cost Distance and the Cost Path tools to ModelBuilder.
• Both the Cost Distance and the Cost Path tools are located in the Spatial Analyst
Tools toolbox in the Distance toolset.
3. Use the Add Connection tool to connect Bate Town and Cost surface to the Cost
Distance tool.
43 | P a g e
Prepared By: Kalid Hassen
4. Rename the Output distance raster variable to Output cost distance and rename the
Output backlink raster variable to Output cost backlink.
5. Use the Add Connection tool to connect destination, Output cost distance, and Output
cost backlink to the Cost Path tool.
6. Rename Output raster to Output cost path.
44 | P a g e
Prepared By: Kalid Hassen
45 | P a g e
Prepared By: Kalid Hassen
1. Click the Auto Layout button, and then click the Full View button .
2. Rename the new_route output variable to Output route.
3. Right-click the Output route variable and click Add To Display.
4. Run the Raster to Polyline tool.
• Examine the new_route layer that is added to our ArcMap display. The raster path
has been converted to a polyline.
46 | P a g e
Prepared By: Kalid Hassen
47 | P a g e
Prepared By: Kalid Hassen
48 | P a g e
Prepared By: Kalid Hassen
GIS is capable of combining large volumes of data from a variety of sources, it is a useful
tool for many aspects of water resources investigation and is emerging as a significant
support tool for hydrologic modeling. In GIS we typically simplify hydrological systems
by assuming overland flow (no water absorbed by soils) and channeled flow (flow
converges). Typically, we derive flow direction, check for sinks, calculate flow
accumulation and derive channel network and catchments.
Imagine that you have just been hired by the Dugda Bora office of water resource. During
the briefing on your first day of work, you assign to derive flow direction, check for sinks,
calculate flow accumulation and derive channel network and catchments select in. You are
quite unfamiliar with Hydrology analysis in GIS so you will have to catch up quickly. The
aim of this exercise is to make you familiar with the basics of GIS hydrology tool and, in
particular, to introduce you to ArcHydro Tool.
49 | P a g e
Prepared By: Kalid Hassen
4.1.3 Description
In this exercise, you will able to identify sinks, determine flow direction, calculate flow
accumulation, delineate watersheds, and create stream networks. The hydrologic modeling
tools in the ArcGIS Spatial Analyst extension toolbox provide methods for describing the
physical components of a surface.
In this exercise, you will become familiar with the generation of river rout files and
watershed files using DEM data. The raster data DEM that was saved in your working
directory will be used in this exercise to delineate the river route and water shed.
4.3 Fill
In GIS we typically simplify hydrological systems by assuming overland flow (no water
absorbed by soils) and channeled flow (flow converges) so we cheek for sinks, sinks are
A local minimum (depression) in a surface representation.
(1) Find the nearest adjacent cell with elevation the same as or lower than the sink hard
code a flow direction to it (radial search)
(2) Increase the elevation of the sink to that of a neighbouring cell check neighbouring cell
then drains (sink filling) (ArcGIS approach)
Steps:
50 | P a g e
Prepared By: Kalid Hassen
Time needed for the process depends on the size of the DEM data you use. A new raster
data Fill will be added into the data frame. The new raster data looks just like the base data
of DEM.
51 | P a g e
Prepared By: Kalid Hassen
The process of flow direction calculation is much faster than the process of filling; within
several seconds the process will complete, a new raster data will be added into the data
frame.
4.5 Flow Accumulation
Total flow accumulated at a given point on a surface equivalent to the drainage area
upstream of a given point. Flow accumulation is then calculated as the sum of upstream
elements draining to a cell for every cell count how many neighbors drain to it. When we
calculate flow accumulation for a cell, we can also calculate catchment boundary for that
cell.
• From the legend of the layer flow accumulation, you will find that the values in nodes
of this raster data layer changes in a range from 0 to over 30,000; the value here
indicates the number of nodes upstream. Therefore, it is possible to estimate that
nodes with relatively large value are main rivers. Cells with high flow accumulation,
greater than a user-defined threshold value, are considered part of a stream network.
Steps:
1. In the same Hydrology toolbox
2. Open the Flow Accumulation tool.
3. The Flow Accumulation window will appear.
4. In the Flow Accumulation window, set the necessary factors as follows:
• Input flow direction raster: flowdir
• Location of output surface raster you’re working directory/ Geodatabase
• Name of the output surface raster: flowacc
5. Click the OK button.
6. The process of flow accumulation calculation is also relatively fast, it takes several
seconds.
52 | P a g e
Prepared By: Kalid Hassen
• When the process has completed, anew raster data flowacc will be added into the data
frame.
4.6 River Route Delineating
4.6.1 Reclassification
Steps:
1. Click on spatial Analyst tool
2. Release the Reclass tool under spatial Analyst The reclassify window will appear
3. Select Flowacc as input raster
4. Click Classify... button.
• The Classification window will appear. The classification here should be set depending
on your purpose of river route delineation. Very small values don’t actually represent a
real river route, but give you an indication of how many nodes are upstream. Also,
different values represent different ranks of rivers. Therefore, you might have to try
changing the setting for classes several times to complete the procedure of
reclassification. In this exercise we simply set the Classes as 3 and the Break Values for
each class as 500, 5000 and the maximum value of your target layer. Click the OK button.
53 | P a g e
Prepared By: Kalid Hassen
54 | P a g e
Prepared By: Kalid Hassen
55 | P a g e
Prepared By: Kalid Hassen
56 | P a g e
Prepared By: Kalid Hassen
57 | P a g e
Prepared By: Kalid Hassen
58 | P a g e
Prepared By: Kalid Hassen
The Arc Hydro tools are a set of public domain utilities developed on top of the Arc Hydro
data model. They operate in the ArcGIS environment. Some of the functions require the
Spatial Analyst extension.
The tools are accessed through the Arc Hydro Tools toolbar, where they are grouped by
functions into six menus, and seven buttons.
Menus
• Attribute Tools: functions allowing generating key attributes (fields) in the Arc
Hydro data model. Some of the tools require the existence of a geometric network.
• Point Delineation
59 | P a g e
Prepared By: Kalid Hassen
The tools have two key purposes. The first purpose is to manipulate (assign) key attributes
in the Arc Hydro data model. These attributes form the basis for further analyses. They
include the key identifiers (such as HydroID, DrainID, NextDownID, etc.) and the
measure attributes (such as LengthDown). The second purpose for the tools is to provide
some core functionality often used in water resources applications. This includes DEM-
based watershed delineation, network generation, and attribute-based tracing.
The functionality of Arc Hydro tools is expected to grow over time. They have been
implemented in a way that allows easy addition to their functionality, either internally (by
adding additional code) or externally, by providing additional functionality using key Arc
Hydro data structures.
5.2 Description
Different hydrologic modeling involves delineating streams and watersheds, and getting
some
basic watershed properties such as area, slope, flow length, stream network density, etc.
Traditionally this was (and still is!) being done manually by using topographic/contour
maps.
With the availability of digital elevation models (DEM) and GIS tools, watershed
properties
can be extracted by using automated procedures. The processing of DEM to delineate
watersheds is referred to as terrain pre-processing. We will use Arc Hydro tools to process
a DEM to delineate watershed, sub-watersheds, stream network and some other watershed
characteristics that collectively describe the drainage patterns of a basin. Thus, the purpose
of this training is in order to give a high light about how to delineate watersheds using
ArcGIS software.
The prime Objective of this Exercise is to create watershed boundaries from raster digital
elevation model datasets and analyze stream networks characteristics within those
boundaries
60 | P a g e
Prepared By: Kalid Hassen
Computer Requirements You must have a computer with windows operating system, and
the following programs installed:
1. ArcGIS
2. Arc Hydro tool
Steps
1. Go to ArcCatalog.
2. Create Folder
3. Right-click on the folder and select New->Personal Geodatabase.
4. Rename the Geodatabase. (WSA)
Steps
1. Set the Target locations (paths where the data will be saved) by selecting Set Target
Locations… from the ApUtilities menu on the ArcHydro toolbar.
2. Choose the Hydro configuration (HydroConfig), check the Map Level checkbox (to
apply the changes to this data frame) and set the paths to your folder, as displayed in
Figure below.
3. Click OK
61 | P a g e
Prepared By: Kalid Hassen
• Add the files and x.shp to the data frame using the Add Data button.
• Save the ArcMap document in your working directory.
Steps:
1. Project the image using the specified zone and hemisphere. If the image is new
and not projected you should do it as follows.
2. Go to Arc tool box → data management tools → projection and transformation
→ Raster → Project raster.
3. When ‘define projection’ window popup adds an image by locating in your folder
that you same it’s in.
4. Locate your data under’ Input data set → ‘specify the ‘coordinate system’ in to
projected coordinate system→ northern hemisphere → UTM → UTM
_WGS1984_Zone 37N proj. → Add → OK.
62 | P a g e
Prepared By: Kalid Hassen
5.3.2 Fill
This function fills the sinks in a grid. If cells with higher elevation surround a cell, the
water is trapped in that cell and cannot flow. The fill sinks function modifies the elevation
value to eliminate these problems; [3].
Steps
1. Go to Terrain Preprocessing > DEM Manipulation > Fill Sinks.
2. The output of the Hydro DEM is Fil. The default name of the output cans be
overwritten by the user.
3. Click “Fill All” button and then Click OK.
63 | P a g e
Prepared By: Kalid Hassen
This function computes the flow direction for a given grid. The values in the cells of the
flow direction grid indicate the direction of the steepest descent from that cell [3].
This step defines the direction of the steepest descent for each terrain cell. Similar to a
compass, the 8-point pour algorithm specifies the following eight possible directions:
1 = east, 2 = southeast,
4 = south, 8 = southwest,
16 = west, 32 = northwest,
Steps
1. Go to Terrain Preprocessing > Flow Direction
2. Confirm that the input of the Hydro DEM is Fil. The output of the Flow Direction
Grid is Fdr.
64 | P a g e
Prepared By: Kalid Hassen
3. Click OK
This step determines the number of upstream cells from any given cell. Upstream drainage
area at a given cell can be calculated by multiplying the flow accumulation value with the
cell area. This function computes the flows accumulation grid that contains the
accumulated number of cells upstream of a cell, for each cell in the input grid [3].
Steps
1. Go to Terrain Preprocessing > Flow Accumulation.
2. Confirm that the input of the Flow Direction Grid is Fdr. The output of the Flow
Accumulation Grid is Fac.
65 | P a g e
Prepared By: Kalid Hassen
This function computes a stream grid which contains a value of stream for all the cells in
the
input flow accumulation grid that have a value greater than the given threshold. All other
cells in the stream grid contain no data. The step classifies all cells with flow accumulation
greater than the user-defined threshold as cells belonging to the stream network [3]
Steps
1. Go to Terrain Preprocessing > Stream Definition
2. Confirm that the input of the Flow Accumulation Grid is Fac. The output of the
Stream Grid is Str.
3. Click OK
• Accept the default stream threshold to initiate streams with an upstream drainage
of 4851 cells.
Note: that the information tip serves as a guide for the user in specifying the threshold by
stating the total drainage area as number of cells or as sq.km, depending on the location of
the cursor.
66 | P a g e
Prepared By: Kalid Hassen
This step divides the stream into segments. Stream segments or links are the sections of a
stream that connects two successive junctions, a junction and an outlet, or a junction and
the drainage divide. This function creates a grid of stream segments that have a unique
identification. Either a segment may be a head segment, or it may be defined as a segment
between two junctions. All the cells in a particular segment have the same grid code that
is specific to that segment [3].
Steps
1. Go to Terrain Preprocessing > Stream Segmentation
2. Confirm that the input of the Flow Direction Grid is Fdr and Stream Grid is Str.
The output of the Stream Link Grid is StrLnk.
3. Click OK.
67 | P a g e
Prepared By: Kalid Hassen
This step delineates a sub-basin for every stream segment. It determines the outline of the
catchments within the SRTM data area using the stream definition files. This function
creates a grid in which each cell carries a value (grid code) indicating to which catchment
the cell belongs. The value corresponds to the value carried by the stream segment that
drains that area, defined in the stream segment link grid [3].
Steps:
1. Go to Terrain Preprocessing > Catchment Grid Delineation
2. Confirm that the input of the Flow Direction Grid is Fdr and Link Grid is StrLnk.
The output of the Catchment Grid is Cat.
3. Click OK
68 | P a g e
Prepared By: Kalid Hassen
Steps:
1. Go to Terrain Preprocessing → Catchment Polygon Processing
2. Confirm that the input of the Catchment Grid is Cat. The output of the Catchment
is Catchment.
3. Click OK
69 | P a g e
Prepared By: Kalid Hassen
This step converts stream in the grid representation into the vector representation. This
function converts the input stream link grid into a Drainage line feature class. Each line in
the feature class carries the identifier of the catchment in which it resides [3].
Steps:
1. Go to Terrain Preprocessing > Drainage Line Processing
2. Confirm that the input of the Stream Link Grid is StrLnk and Flow Direction Grid
is Fdr.
3. The output of the Drainage Line is DrainageLine.
4. Click OK
70 | P a g e
Prepared By: Kalid Hassen
Steps:
1. Go to Terrain Preprocessing > Adjoint Catchment Processing.
2. Confirm that the input of the Drainage Line is DrainageLine and Catchment is
Catchment.
3. The output of the Adjoint Catchment is AdjointCatchment.
4. Click OK
This step is used to create a drainage point at the most downstream point in the catchment
(center of a grid cell with the largest value in the flow accumulation grid for that
catchment) [3].
71 | P a g e
Prepared By: Kalid Hassen
Steps:
1. Go to Terrain Preprocessing > Drainage Point Processing
2. Confirm that the input of the Flow Accumulation Grid is Fac, Catchment Grid is
Cat, and Catchment is Catchment.
3. The output of the Drainage Point is DrainagePoint.
4. Click OK
This helps to delineate a watershed for a user specified point (interactive) based on the
preprocessed DEM [3]
Steps:
1. Click on the Point Delineation icon in the ArcHydro toolbar to activate the
tool.
2. Confirm that the input of the Flow Direction Grid is Fdr, Stream Grid is
Str,Catchment is Catchment, Adjoint Catchment is Adjoint Catchment, and
Watershed is Watershed.
72 | P a g e
Prepared By: Kalid Hassen
73 | P a g e
Prepared By: Kalid Hassen
74 | P a g e
Prepared By: Kalid Hassen
Network Analysis
75 | P a g e
Prepared By: Kalid Hassen
Assume that you have a meeting somewhere far from your resident for 3 days, and you
know that there is a huge traffic jam on the area so you think to line out the shortest path
which takes to your destination based on your travel time. In addition, you want to have a
room (hotel) after finishing the meeting for those three days so you want to know the
distance that can be reached from a facility within a specified amount of time. These
questions can be answered by network analysis.
Your input data set are:
Road To identify the shortest path for your destination
Hotel location To construct service area polygon
6.1.2 Description
In this Exercise you will find the best route for the given order of stops based on travel
time and you will create a series of polygons representing the distance that can be reached
from a facility within a specified amount of time. These polygons are known as service
area polygons.
6.2 Concept of Network Analysis
A network is essentially a set of lines known as segments or edges connected or joined by
a set of vertices known as nodes or junctions. A GIS stores these edge and junction features
with their attributes. ArcGIS Network Analyst provides network-based spatial analysis
tools for solving complex routing problems. It uses a configurable transportation network
data model, allowing organizations to accurately represent their unique network
requirements.
A Network dataset is a GIS dataset that is designed to support network analysis. It typically
consists of lines representing the routes of flow in the network, augmented with other
features (such as junction points), topology, and attributes that model network-relevant
properties such as impedance and capacity of flow.
What does it solve?
• Best route
• Closest facilities
• Service Area
• Origin destination cost matrix
76 | P a g e
Prepared By: Kalid Hassen
Steps:
1. Go to Edit
2. Start editing
3. Click Edit Vertices and delete Vertices
4. Take other vertices to other line
5. Right click on that vertices and finish sketch
• Next you have to consider the end point of each line is the beginning of another
lines.
Steps:
1. Go to split and Click on the junction of that line
2. Do the same for all roads
77 | P a g e
Prepared By: Kalid Hassen
Note: If the coordinate system of the line is in GCS you have to convert in to projected
coordinate System. Otherwise in attribute table it is impossible to calculate the geometries
for features
6.4 Adding fields to the attribute table and filling necessary data’s
6.4.1 Length
I. Adding a field to a table (Length)
Steps:
1. Open the table (right-click on the layer, Open Attribute Table)
2. Click on Options from the lower part of the spreadsheet window
3. Type = “Float”→ Ok
4. Click Add Field.
5. Write the name of the new field as Length. The New field will be added as the last
column.
78 | P a g e
Prepared By: Kalid Hassen
79 | P a g e
Prepared By: Kalid Hassen
8. When finished the dialogue box should be filled out as below. Click the verify tab
and ensure that the syntax is correct.
9. Click on Apply tab and Close
6.4.3 Time
I. Adding a field to a table (Time)
Steps:
1. Open the table (right-click on the layer, Open Attribute Table)
2. Click on Options from the lower part of the spreadsheet window
3. Type = “Double” → Ok
4. Click Add Field.
5. Write the name of the new field as Minute. The New field will be added as the last
column.
80 | P a g e
Prepared By: Kalid Hassen
81 | P a g e
Prepared By: Kalid Hassen
2. Go to your output location and right click on Topology Roads then click New
Network Dataset.
3. Tick No
82 | P a g e
Prepared By: Kalid Hassen
5. Click Next
83 | P a g e
Prepared By: Kalid Hassen
4. We have to choose facilities around which we are going to create service area polygon
5. Right click on facility and load locations
6. Select Hotels from Load From Draw dawn arrow
7. Click Ok
84 | P a g e
Prepared By: Kalid Hassen
85 | P a g e
Prepared By: Kalid Hassen
86 | P a g e
Prepared By: Kalid Hassen
87 | P a g e
Prepared By: Kalid Hassen
5. Click Solve
88 | P a g e
Prepared By: Kalid Hassen
Geostatistical Analyst
89 | P a g e
Prepared By: Kalid Hassen
90 | P a g e
Prepared By: Kalid Hassen
• Start ArcMap and enable Geostatistical Analyst To begin, start ArcMap and enable the
Geostatistical Analyst extension.
Steps:
1. Start ArcMap.
2. Click Cancel on the ArcMap - Getting Started dialog box.
• This dialog box may not open if you've previously opted not to show it.
3. On the main menu, click Customize > Extensions.
4. Check the Geostatistical Analyst check box.
5. Click Close.
I. Add the Geostatistical Analyst toolbar
Steps:
1. On the main menu, click Customize → Toolbars → Geostatistical Analyst.
• The Geostatistical Analyst toolbar is added to your ArcMap session. The extension
and toolbar only need to be enabled and added once; they will be active and present
the next time you open ArcMap.
2. We will add our data to ArcMap and alter its symbology.
Steps:
1. Click the Add Data button on the Standard toolbar.
2. Navigate to the folder where datas are found.
3. Double-click the Rain_fall.gdb geodatabase to see its contents.
4. Press the Ctrl key and choose the oromia_Town and oromia_Zone datasets.
5. Click Add.
6. Right-click the oromia_Zone layer legend (the box below the layer's name) in the table
of contents and click No Color.
7. Double-click the Oromia_Town layer's name in the table of contents.
8. In the Layer Properties dialog, click the Symbology tab.
9. In the Show box, click Quantities and click Graduated colors.
10. In the Fields box, set the Value to Rain_Fall. Choose the White to Black color ramp.
11. Click ok
12. On the main menu, click File > Save.
91 | P a g e
Prepared By: Kalid Hassen
5. Click Next.
6. Click Ordinary Kriging; notice that Prediction Map is selected as the output type.
• Since the method to map the Rainfall surface is selected, we could click Finish to
create a surface using the default parameters. However, steps 7 to 11 will expose
us to other dialog boxes. In each step of the wizard, the interior panels (windows)
can be resized by dragging the dividers between them.
92 | P a g e
Prepared By: Kalid Hassen
7. Click Next.
• The semivariogram/covariance model is displayed, allowing you to examine
spatial relationships between measured points. You can assume that things that are
closer together are more alike than things that are farther apart. The semivariogram
allows you to explore this assumption. The process of fitting a semivariogram
model to capture the spatial relationships in the data is known as variography.
8. Click Next.
• The crosshairs show a location that has no measured value. To predict a value at
the crosshairs, we can use the values at the measured locations. We know that the
values of the closest measured locations are most alike to the value of the
unmeasured location that you are trying to predict. The red points in the image
below are going to be weighted (or influence the unknown value) more than the
green points since they are closer to the location you are predicting. Using the
93 | P a g e
Prepared By: Kalid Hassen
9. Click Next.
• The cross-validation diagram gives you an idea of how well the model predicts the
values at the unknown locations.
94 | P a g e
Prepared By: Kalid Hassen
14. Click the Save button on the Standard toolbar to save your work.
• Notice that the interpolation continues into the other region because the extent of
the layer is the same as the extent of the input data (oromia_Town).
15. To restrict the prediction surface to within Oromia region, right-click the Default
Kriging layer and click Properties.
95 | P a g e
Prepared By: Kalid Hassen
22. Drag the oromia_town layer to the top of the table of contents.
23. Right-click the Default Kriging layer in the table of contents and click
Validation/Prediction.
• This opens the GA Layer to Points geoprocessing tool with the Default Kriging
layer specified as the input geostatistical layer.
96 | P a g e
Prepared By: Kalid Hassen
24. Input geostatistical layer should be automatically set to Default Kriging. For Point
observation locations, navigate to the geodatabase that contains the data, and click the
oromia_Town dataset. Leave the Field to validate on empty as we just want to generate
Rainfall predictions for the Oromia towns, not validate the predicted values against
measured values. For Output statistics at point locations, navigate to the folder we
created for the output, and name the output file Oromia_Town_Rainfall.shp. Leave the
Append all fields from input features checked as we want to be able to see the names
of the cities in the output feature class.
• The GA Layer To Points geoprocessing tool dialog box should look like this:
97 | P a g e
Prepared By: Kalid Hassen
98 | P a g e
Prepared By: Kalid Hassen
Steps:
1. Click the Oromia_zone layer; drag it under the Oromia_Town layer in the TOC.
2. Click the Oromia_Town layer to select it.
3. On the Geostatistical Analyst toolbar, click Geostatistical Analyst > Explore Data >
Histogram.
4. On the Histogram dialog box, click the Attribute arrow and choose OZONE.
5. Select one histogram bars by clicking and dragging the pointer over them. The sample
points within this range are selected on the map.
6. Click the Clear Selected Features button on the Tools toolbar to clear the selected
points on the map and histogram.
7. Click the Close button located in the upper corner of the Histogram dialog box.
II. Create a normal QQ plot
The quantile-quantile (QQ) plot is used to compare the distribution of the data to a standard
normal distribution, providing another measure of the normality of the data. The closer the
points are to the straight (45-degree) line in the graph, the closer the sample data follows
a normal distribution.
99 | P a g e
Prepared By: Kalid Hassen
Steps:
1. On the Geostatistical Analyst toolbar, click Geostatistical Analyst > Explore Data >
Normal QQPlot.
• A general QQ plot is a graph on which the quantiles from two distributions are plotted
versus each other. For two identical distributions, the QQ plot will be a straight line.
Therefore, it is possible to check the normality of the ozone data by plotting the
quantiles of that data versus the quantiles of a standard normal distribution. From the
normal QQ plot above, you can see that the plot is not very close to being a straight
line. The main departure from this line occurs at low values of rainfall distribution if
the data does not exhibit a normal distribution in either the histogram or normal QQ
plot, it may be necessary to transform the data to make it conform to a normal
distribution before using certain kriging interpolation techniques.
3. Click the Close button located in the upper corner of the Normal QQPlot dialog box.
III. Identify global trends in our data
If a trend exists in your data, it is a nonrandom (deterministic) component of a surface that
can be represented by a mathematical formula. For instance, a gently sloping hillside can
be represented by a plane. A valley would be represented by a more complex formula (a
100 | P a g e
Prepared By: Kalid Hassen
second-order polynomial) that creates a U shape. This formula may produce the
representation of the surface you want. However, many times the formula neither produces
a surface that is too smooth to accurately depict the phenomenon because no hillside is a
perfect plane nor is any valley a perfect U shape. Local variation can be added to the
surface by modeling the trend using one of these smooth functions, removing it from the
data and continuing your analysis by modeling the residuals, which is what remains after
the trend is removed. When modeling the residuals, you will be analyzing the short-range
(local) variation in the surface. The Trend Analysis tool enables you to identify the
presence/absence of trends in the input dataset and identify which order of polynomial fits
the trend best.
Steps:
1. On the Geostatistical Analyst toolbar, click Geostatistical Analyst > Explore Data >
Trend Analysis.
101 | P a g e
Prepared By: Kalid Hassen
3. Click the Rotate Locations scroll bar and scroll left until the rotation angle is 90
degrees.
102 | P a g e
Prepared By: Kalid Hassen
indicating that the values of the pairs of points separated by more than this distance
are no longer correlated.
• Looking at the semivariogram, if it appears that some data locations that are close
together have a higher semivariogram value than we would expect, we should
investigate these pairs of locations to see if there is a possibility that the data is
inaccurate.
3. Click the Select Features by Rectangle button on the Tools toolbar, then click and drag
the pointer over some points with large semivariogram (y-axis) values on the
Semivariogram/Covariance Cloud dialog box to select them.
103 | P a g e
Prepared By: Kalid Hassen
• The pairs of sample locations that are selected in the semivariogram are highlighted
on the map, and lines link the locations, indicating the pairing. As it might be expected
from the default kriging prediction map, the lines with high semivariogram values for
a particular distance between the points in a pair are those that correspond to the largest
gradient in the Rainfall values.
4. Check Show search direction.
5. Click and move the directional pointer to any angle.
• The direction the pointer is facing determines which pairs of data locations are plotted
on the semivariogram.
104 | P a g e
Prepared By: Kalid Hassen
Equation (RUSLE)
105 | P a g e
Prepared By: Kalid Hassen
8.1.2 Description
In order to execute RUSLE model, we have to drive six parameters of the RUSLE model
which are directly derived or estimated from the following data:
✓ Rainfalls – R factor;
✓ Digital elevation model (DEM) – LS factor;
✓ Satellite Images or Land use and land cover- C and P factors.
✓ soil databases – K,
The implementation of the RUSLE model consists of three main steps. The preprocessing
of the data consists of data formatting. This step includes
• reprojection of data, clipping of the data according to the area of interest, some
attribute correction and rasterizing;
• indicators’ estimates are then realized according to theoretical formulae or according
to attribute assignment rules;
• the soil loss estimate is calculated using formula.
106 | P a g e
Prepared By: Kalid Hassen
Pre processing
Reprojection, Clipping, attribute correction
Formula (1)
Soil Losses A
Erosion processes include all the drafting and transportation mechanisms of sediments at
the Earth’s surface. They are either physical or chemical. Erosion agents can be
meteorological, hydric, gravity-related or human. Knowledge, estimation and
management of erosion are key factors in the management of economic (agronomic quality
of soils, economic vulnerability), environmental (river discharge, biodiversity loss) and
social (water drinking supply, risks to lives and goods) issues.
There are many erosion models, which may or may not include the spatial dimension of
processes, to qualify and quantify erosion. One of the most commonly used spatial models
is the empirical model relying on the Universal Soil Loss Equation. This model has the
advantages:
✓ Of exploiting accessible geographical parameters (climate, land cover, topography);
✓ Of being operated using GIS software;
✓ Of being adaptable to the degree of knowledge of the area studied: it can thus be
exploited quantitatively or qualitatively depending on the validation given to each
parameter;
✓ Of being able to provide quantitative estimates of soil loss.
107 | P a g e
Prepared By: Kalid Hassen
The model had been revised since its first release. The Revised Universal Soil Loss
Equation (RUSLE) relies on the same equation but the methods to obtain the different
parameters have been improved.
The RUSLE model, initially used to quantify soil losses over agricultural parcels, provides
estimates of sediment displacement within each cell of the model due to hydric processes
(rainfall and surface runoff). The soil losses are expressed in [Mass] · [Surface Unit] -
1
· [Time period]–1 such as ton/ha/year.
The RUSLE equation is the following:
A= R × K × LS × C × P
✓ A is the soil loss expressed in the same unit as the K factor [Mass] · [Surface unit] –1
and for the period selected for the R period. Theoretically, the soil loss varies between
0 and infinity; in practice, splash and sheet erosion correspond to A values between 0
and 5 ton/ha/year and erosion inside lavakas (hole) or gully erosion can easily exceed
1,000 ton/ha/year;
✓ R is the rainfall aggressiveness factor. R varies between 0 and infinity; under
tropical climates, R factor mean values reach 1500-–1900 MJ·mm·(ha·h·year)–1, and
can exceed 3,000 MJ·mm·(ha·h·year)–1 [4]
✓ K factor characterizes soil erodibility. This dimensionless factor is experimentally
obtained and provides soil loss capacity in ton per surface per R unit for each soil.
Usually, R varies from 0 to 0.7 (t·acre·year)/(ha·MJ·mm);
✓ L is the slope length expressed in length unit;
✓ S is the slope gradient. For convenience, S factor is estimated jointly with L factor in
a unique LS factor. LS is dimensionless and varies between 0 and infinity;
✓ C is the vegetation cover factor. It is also a dimensionless factor that varies between 0
and 1;
✓ P factor qualifies agricultural and soil protection practices. P is dimensionless
and varies between 0 and 1.
The use of the RUSLE equation to model erosion must take into account some limits
inherent to the methodology and to each parameter:
• the choice of the RUSLE equation is often driven by the availability of the spatial
parameters required over a selected region;
108 | P a g e
Prepared By: Kalid Hassen
• erosion modeling using RUSLE provides soil losses estimates in t/ha/year. If erosion
modeling relies on scientifically uncalibrated parameters or if the model results are
not validated by comparison with field survey measurements, it is better to use the
results as a relative or qualitative assessment of erosion risk (low to high erosion);
• access to certain modeling parameters, such as the P factor, may turn out to be
difficult;
• the parameters used (soil map, topography, land use) are often available at different
scales and precisions that could lead to inaccuracies related to the least precise spatial
parameters;
• RUSLE equation considers each parameter with the same weight, whereas some
parameters may have greater influence than others according to local conditions
However, RUSLE modeling shows a good correlation between the location of erosion
zones compared to field observations. The modeling allows also for comparison of
different sites as well as temporal monitoring.
8.3 Climatic factor: rainfall aggressiveness R Factor
Rainfall aggressiveness is a factor that determines the capacity of rainfall events to
displace soil. The R factor therefore takes into account total rainfall and the occurrence of
rainfall peaks capable of causing soil displacement.
Based on 30 years of rainfall records, [5] consider that rain aggressiveness is not limited
to intense rainfall events. It must include the cumulative effect of moderate rainfalls. The
authors therefore consider that soil losses during a rainfall event are proportional to the
total energy of the precipitations (E) expressed in MJ·mm·(ha·h)–1 multiplied by the
maximal intensity of the rain event measured over 30 min (I30) expressed in mm·h–1.
Rain aggressiveness R expressed in MJ·mm·(ha·h·period)–1 for a given period of time is
modelled as follows [6]:
with:
• (EI 30) i= EI30 for the rain event i;
• j is the number of rain events;
• N is the period considered (a month, a year, several decades, etc.).
Nevertheless, the use of this formula to assess the spatial repartition of erosion requires
109 | P a g e
Prepared By: Kalid Hassen
accurate rainfall measurements covering the entire area of interest, which is often difficult
to obtain. Furthermore, the authors propose several approximations methods of the R
factor according to the parameters available. [7] developed an approximation based on the
following equation:
R= K × P × 0.5
with:
• P is mean annual rainfalls in MJ·mm·(ha·h·year)–1;
• K is an adjustment coefficient. We use a k value of 1.73 in this Manual, proposed by
[4].
However, if we access metrological data from local metrological station the above stated
equation might not work for Ethiopia case the reason that of rainfall kinetic energy and
intensity data are not available in most cases. Therefore, the erosivity factor R was
calculated according to the equation given by Hurni [8], derived from a spatial regression
analysis for Ethiopian conditions, the model adapted by Hurni for Ethiopian conditions is
based on the available mean annual rainfall data (P) where
R = - 8. 1 2 + (0. 562 x P)
8.3.1 Calculation of the average annual Precipitation
The average annual rainfall corresponds to the amount of precipitation in mm received on
average in each mesh of the model. In this implementation, the average annual rainfall is
estimated by summing the mean monthly precipitation according to the formula.
110 | P a g e
Prepared By: Kalid Hassen
Therefore, average annual rainfall should be reprojected and clipped on the area covered
by the other datasets. The geographic intersection of all the datasets corresponds to the
shapefile layer that covers the most restricted area.
Steps:
1. With ArcMap open, and Open up the Arc Toolbox, Data
management→sampling→Create random point
111 | P a g e
Prepared By: Kalid Hassen
5. Click OK.
II. Extracting Value to Point
Step:
1. Click on Geoprocessing ArcToolbox
2. Click Spatial Analyst Tool → Extraction → Extract multi values to point
112 | P a g e
Prepared By: Kalid Hassen
5. Click Ok
III. Spatial Interpolation
Steps:
1. With ArcMap open, and New created point (rainfall) data added from your working
directory.
2. Open up the ArcToolbox spatial Analyst Tools→Interpolation→Kriging
113 | P a g e
Prepared By: Kalid Hassen
Where
λ: is the actual slope length and m is the slope length exponent that is the ratio of rill to
interill erosion.
114 | P a g e
Prepared By: Kalid Hassen
The S factor (slope steepness factor) is the ratio of soil loss relative to a 9% slope, which
is the standard slope that experiment plots use. The slope steepness factor is calculated as
a function of slope as shown below:
Where
𝑆 is the slope factor, and 𝜃 is the slope angle.
Depending on the measured slope gradient, a different equation for 𝑆 must be used.
Choosing 𝑆 allows the RUSLE to be more finely tuned for different terrains. This is
important because the topographic factor (and the RUSLE entirely) is very sensitive to the
slope factor 𝑆.
In this Manual, LS is calculated by the USPED (Unit Stream Power Erosion and
Deposition) method, which is using the raster calculation between flow accumulation and
slope of watershed, because it is obvious that it can be done with the tools included in a
normal GIS ArcMap installation.
In comparison to the RUSLE, the USPED is a physically based model that incorporates
a spatial component. In the RUSLE, 𝐿 is dependent on linear distance 𝜆𝑖, which is the
horizontal length from the start of sediment transport to point 𝑖 on the slope. Thus, they
are inherently a single dimensional function. The USPED instead uses the area of upland
contributing flow at distance 𝑖. In the USPED model, the area is substituted in place of the
former slope length. The L calculation for point 𝑖 on a slope is shown in Equation:
where,
𝐿 is the slope length factor at some point on the landscape,
𝜆 is the area of upland flow,
m is an adjustable value depending on the soil’s susceptibility to erosion, and is the unit
plot length.
The calculation of S value is shown in Equation
115 | P a g e
Prepared By: Kalid Hassen
where,
S is the slope in degrees, is the slope gradient constant, and is an adjustable
value depending on the soil’s susceptibility to erosion. Designations for exponents and
values can be found in the literature.
Using USPED method, the LS factor is calculated in the GIS program according to the
following steps:
Step:
1. Calculate Flow Direction from clipped Watershed DEM layer Using Flow
Direction Tool,
2. Calculate Flow Accumulation with Flow Accumulation Tool using flow direction
data as the input raster,
3. Calculate slope of watershed in degrees using Slope Tool using clipped
watershed DEM as the input layer, and
4. Copy and paste the LS-factor formula below into Raster Calculator:
Power (“flow accumulation” *[cell resolution]/22.1,0.4) *Power (Sin (“slope of
degree” *0.01745))/0.09, 1.4) *1.4.
116 | P a g e
Prepared By: Kalid Hassen
To take into account natural environments, the C factor is likened to the degree of soil
cover by vegetation. The variation of the soil loss potential of a defined soil under different
vegetation covers is then better adjusted (from bare soil to soil under dense vegetation).
[11].
Several methods exist to evaluate the C factor exploiting remote sensing imagery and
calculation of a vegetation index (NDVI, vegetation fraction) or through the estimate of
the mean vegetation cover according to a land cover map. This last method is presented
here. It consists of associating an estimated value of vegetation cover for each class of land
use.
The C factor can be calculated from satellite images through NDVI (Normalized
difference vegetation Index). Since the C factor ranges from 0 (full cover) to 1 (bare land)
and the NDVI values range from 0 (full cover) to 1 (bare land). The calculated NDVI value
inversed using Van der Kniff, 1999 equation:
C= EXP ((-2 * NDVI / (1-NDVI))
Steps:
1. Open the NDVI created file
2. In the Spatial Analysis tool Under ArcTool box: Click on: Map Algebra→ Raster
Calculator…
3. In the Raster Calculator: Enter the following expression:
C= EXP ((-2 * NDVI)/ (1-NDVI))
C factor can be also estimated by the association of a C value to each class of land use.
The chosen values are those proposed and presented below. [12].
117 | P a g e
Prepared By: Kalid Hassen
Though in reality, especially at local level, these data are often difficult to find and may
not be suitable for extrapolation from one area to another. For these reasons Hurni (1985)
derived K values based on easily observable soil colors.
Soil erodibility value estimated based on soil color.
Soil Color Black Brown Red Yellow
K-factor 0.15 0.2 0.25 0.3
118 | P a g e
Prepared By: Kalid Hassen
II. Rasterizing
Steps:
1. Open the Arc toolbox → Conversion Tools → To raster → Polygon to Raster.
2. In the window of Polygon to Raster
• Input raster: Allocation to Soil
• Value: Allocate to the new created and filled column
• Output raster features: on selected file destination
• Cell size: Set Appropriate cell size compared to others.
119 | P a g e
Prepared By: Kalid Hassen
P factor estimates rely on the knowledge level and the availability of sufficient databases
concerning cultivation practices of the area of interest. Under natural conditions, the P
factor is set to 1 [4].
In this manual, the ratio is kept at 1, indicating straight-row farming.
8.8 soil loss A calculation from the RUSLE equation
The calculation of soil losses A according to the RUSLE formula corresponds to the
multiplication of each previously prepared factor (R, LS, K and C).
120 | P a g e
Prepared By: Kalid Hassen
Steps:
1. Open the previously created file
2. In the Spatial Analysis tool Under ArcTool box: Click on: Map Algebra→ Raster
Calculator…
3. In the Raster Calculator: Enter the previous created files (R, LS, K and K)
4. Save the file
The result is a raster file with values range between 0 and 1,496 t/ha/year.
121 | P a g e
Prepared By: Kalid Hassen
122 | P a g e
Prepared By: Kalid Hassen
Assume you have hired at federal ministry of agriculture (MOA). You are assigned to
develop a Land Surface Temperature for Dire Dawa region using a Landsat image to be
an input factor for climate change, urban land use/land cover, heat balance studies
9.1.2 Description
In this Exercise you will shows how to calculate Land Surface Temperature (LST) using
the Landsat 8 bands. In particular, band 10 as the thermal band, and bands 4 and 5 to
calculate the Normal Difference Vegetation Index (NDVI). The first step is to download
Landsat 8 (OLI-TIRS) image from a particular location, unzip it, and check certain
information needed (within the metadata or Landsat Tool software) to execute this
procedure. To calculate the LST, you will use the USGS formulas this example simply
performs the calculations without going into detail about the fundamentals. The Land
Surface Temperature can be estimated or calculated using the Landsat 8 thermal bands. It
simply requires applying a set of equations through a raster image calculator (ArcMap,
ArcGIS Pro, QGIS). You can get necessary information’s from Landsat meta data, you can
also open your meta data using the different software (Lansat Tool, SCP Plugin in QGIS..).
Landsat tools version 1.0.0 (http://WWW.ukscience.org/_Media/LSTools_setup.zip) were
used to explore detail meta data.
Your input data set are;
The first step of the algorithm is the input of Band 10. After inputting band 10, in
the background, the tool uses formulas taken from the USGS web page for retrieving the
top of atmospheric (TOA) spectral radiance (𝐿𝜆), Using the radiance rescaling factor,
Thermal Infra-Red Digital Numbers can be converted to TOA spectral radiance.
123 | P a g e
Prepared By: Kalid Hassen
Lλ = ML * Qcal + AL
Where:
Lλ = TOA spectral radiance (Watts/ (m2 * sr * μm))
ML = Band-specific multiplicative rescaling factor from the metadata
(RADIANCE_MULT_BAND_x, where x is the band number).
AL = Band-specific additive rescaling factor from the metadata (RADIANCE_ADD_BAND_x,
where x is the band number).
Qcal = Quantized and calibrated standard product pixel values (DN) corresponds to band
10.
Therefore, the equation must be solved using the Raster Calculator tool in ArcMap.
124 | P a g e
Prepared By: Kalid Hassen
After the digital numbers (DNs) are converted to reflection, the TIRS band data should be
converted from spectral radiance to brightness temperature (BT) using the thermal
constants provided in the metadata file [14]. The following equation is
where:
125 | P a g e
Prepared By: Kalid Hassen
L = TOA
Therefore, to obtain the results in Celsius, the radiant temperature is adjusted by adding
the absolute zero (approx. -273.15°C) [15].
126 | P a g e
Prepared By: Kalid Hassen
Landsat visible and near-infrared bands were used for calculating the Normal Difference
Vegetation Index (NDVI). The importance of estimating the NDVI is essential since the
amount of vegetation present is an important factor and NDVI can be used to infer general
vegetation condition [16]. The calculation of the NDVI is important because, afterward,
the proportion of the vegetation (𝑃V) should be calculated, and they are highly related
with the NDVI, and emissivity (𝜀) should be calculated, which is related to the 𝑃V
where:
𝑃v is calculated Using
A method for calculating 𝑃v [17]. suggests using the NDVI values for vegetation and soil
(NDVImax = 0.5 and NDVImin= 0.2) to apply in global conditions [18].
127 | P a g e
Prepared By: Kalid Hassen
However, since the NDVI values differ for every area, the value for vegetated surfaces,
0.5, may be too low. Global values from NDVI can be calculated from at-surface
reflectivity, but it would not be possible to establish global values in the case of an NDVI
computed from TOA reflectivity, since NDVImin and NDVImax will depend on the
atmospheric conditions [18].
Usually the minimum and maximum values of the NDVI image can be displayed directly
in the image (both in ArcGIS, QGIS, ENVI, Erdas Imagine), otherwise you must open the
properties of the raster to get those values.
128 | P a g e
Prepared By: Kalid Hassen
The land surface emissivity (LSE (𝜀)) must be known in order to estimate LST, since the
LSE is a proportionality factor that scales blackbody radiance (Planck’s law) to predict
emitted radiance, and it is the efficiency of transmitting thermal energy across the surface
into the atmosphere [18]. The determination of the ground emissivity is calculated
conditionally as suggested in [19]:
where
𝐶= the surface roughness (𝐶 = 0 for homogenous and flat surfaces) taken as a constant
value of 0.005 [20].
When the NDVI is less than 0, it is classified as water, and the emissivity value of 0.991
is assigned. For NDVI values between 0 and 0.2, it is considered that the land is covered
with soil, and the emissivity value of 0.986 is assigned. Values between 0.2 and 0.5 are
considered mixtures of soil and vegetation cover is applied to retrieve the emissivity. In
the last case, when the NDVI value is greater than 0.5, it is considered to be covered with
vegetation, and the value of 0.973 is assigned.
ε = 0.004 * Pv + 0.986
Simply apply the formula in the raster calculator, the value of 0.986 corresponds to a
correction value of the equation.
129 | P a g e
Prepared By: Kalid Hassen
The last step of retrieving the LST or the emissivity corrected land surface temperature
𝑇𝑠 is computed as follows [21]:
where
BT is at-sensor BT (∘C),
𝜆 is the wavelength of emitted radiance (for which the peak response and the average of
the limiting wavelength (𝜆 = 10.895) will be used), [22]:
[23]
Where
130 | P a g e
Prepared By: Kalid Hassen
Finally apply the LST equation to obtain the surface temperature map.
As a result of the process developed, there is a map of the Land Surface Temperature, it
should be noted that it is not equal to the air temperature.
131 | P a g e
Prepared By: Kalid Hassen
Supervised Image
132 | P a g e
Prepared By: Kalid Hassen
Assume you have hired at Harar municipality. You are assigned to develop a Land use
Land Cover with a high accuracy using Deep learning Algorithm. the developed land use
will be as an input for different urban planning issues. Due this you prefer to use Open CV
(Machine learning) SVM classification to result a high accuracy classified image.
10.1.2 Description
In this Exercise you will learn how to create a training sample, classifier definition file and
finally resulting a high accurate classified LULC map of Harar. In this analysis you will
use high resolute Aerial photograph to classify LULC using support vector machine. In
SVM any Esri-supported raster is accepted as input, including raster products, segmented
rasters, mosaics, image services, or generic raster datasets. Segmented raster’s must be 8-
bit rasters at least with 3 bands. To create the training sample file, use the Training Sample
Manager from the Image Classification toolbar. The Segment Attributes parameter is
enabled only if one of the raster layer inputs is a segmented image.
For segmented raster’s that have their key property set to Segmented, the tool computes
the index image and associated segment attributes from the RGB segmented raster. The
attributes are computed to generate the classifier definition file to be used in a separate
133 | P a g e
Prepared By: Kalid Hassen
classification tool. The attributes for each segment can be computed from any Esri-
supported image.
There are several advantages with the SVM classifier tool, as opposed to the maximum
likelihood classification method:
• The SVM classifier needs much fewer samples and does not require the samples
to be normally distributed.
• It is less susceptible to noise, correlated bands, and an unbalanced number or size
of training sites within each class.
Supervised classification uses the spectral signatures obtained from training samples to
classify an image. With the assistance of the Image Classification toolbar, you can easily
create training samples to represent the classes you want to extract. You can also easily
create a signature file from the training samples, which is then used by the Spatial Analyst
multivariate classification tools to classify the image.
After enabling the Spatial Analyst extension, you can add the Image Classification toolbar
and begin to use its tools to work with your data.
Steps:
The Image Classification toolbar allows you to classify a multiband raster. A multiband
Aerial photograph is an example of a multiband raster. To specify a source image for the
classification analysis, add the image to ArcMap. Then choose the image layer in the Layer
drop-down list on the Image Classification toolbar.
134 | P a g e
Prepared By: Kalid Hassen
Steps:
1. Add the Image Classification toolbar to ArcMap. Refer to the topic Adding the
Image
2. Classification toolbar for how to add it.
3. On the Standard toolbar, click the Add Data button.
4. On the Add Data dialog box, browse to the image on the disk and click Add.
5. The image is added as a layer in the table of contents. The layer points to all the bands
on the disk.
6. On the Image Classification toolbar, click the Layer arrow and click the image layer
you just added.
7. This specifies the source of images for all subsequent image classification tasks. all
the bands associated with the image layer are used in the classification analysis.
To create training samples, use the training sample drawing tools on the Image
Classification toolbar.
Steps:
The steps below show how to create training samples with the Image Classification
toolbar:
1. On the Image Classification toolbar, choose an appropriate image layer in the Layer
list.
2. Click one of the drawing tools on the Image Classification toolbar. Notice that there
are three drawing tools available, for drawing polygons, circles, and rectangles.
135 | P a g e
Prepared By: Kalid Hassen
4. In ArcMap, identify an area that belongs to a known class. Draw a training sample to
enclose it.
• The following image shows a polygon training sample in ArcMap
• Once you finish drawing the training sample, a new class is created in Training
Sample Manager with a default name, value, and color.
3. Open Training Sample Manager and change the class name, value, and color if desired.
4. Repeat steps 2 to 4 to create a few more training samples to represent the rest of the
classes on the image. The following is what Training Sample Manager looks like after
ten samples are created:
During the process of creating and evaluating training samples, you will need to use the
tools on the Training Sample Manager to maintain the list of classes. For example, you
may need to change the class name, value, and color. Also, classes may need to be merged,
deleted, and reorganized, and so on. See the following on how to perform some common
training sample management tasks.
136 | P a g e
Prepared By: Kalid Hassen
Steps:
1. click the first row and hold shift button to select the same class type with different
sample collected area.
2 . In order to merge the selected class, click on merge training sample button
3 . In the Training Sample Manager, click the Save button. A file browser dialog box
appears.
4 . In the file browser dialog box, navigate to the desired location and specify a name
for the feature class or shapefile, then click Save.
Steps:
1. open ArcTool box in ArcMap →Click on spatial Analysis tool→ navigate to train
support vector machine classifier Toolset
137 | P a g e
Prepared By: Kalid Hassen
3. click OK
138 | P a g e
Prepared By: Kalid Hassen
Steps:
1. Create training samples if no training samples have been created. Refer to the topic
creating training samples on how to create training samples.
2. Create a signature file for the training samples if it has not been created already.
3. Create a classifier definition file for the training samples if it has not been created
already.
4. On the spatial analysis toolbar under ArcToolbox, click on Classify raster.
5. In the tool dialog box, specify values for the three required parameters.
• Input raster: Allocate to the Arial image to be classified
• Input classifier definition file: Allocate to the created classifier definition you
saved earlier
• Output raster features: on selected file destination create your classified file.
(don’t forget to write .tif) after labeling your output file.
The output classified raster will be automatically added to ArcMap when the tool
finishes.
139 | P a g e
Prepared By: Kalid Hassen
140 | P a g e
Prepared By: Kalid Hassen
REFERENCES
[1] T. Saaty, "A scaling method for priorities in hierarchical structures," Mathematical
Psychology, vol. 15, pp. 231-281 pp, 1977.
[2] Saaty and Vargas, Prediction, Projection and Forecasting, Dordrecht: Kluwer
Academic Publishers, 1991, p. 251 pp.
[3] ArcGISresources, "ArcGIS Help 10.1," 2013. [Online]. Available:
http://resources.arcgis.com/en/help/.. [Accessed 7 August 2021 ].
[4] Dumas p., Printemps j., Mangeas m., "Developing Erosion Models for Integrated
Coastal Zone Management: A Case Study of The New Caledonia West Coast,"
Marine Pollution Bulletin,, Vols. vol. 61, nos 7–12, p. pp. 519–529, 2010.
[5] wischmeier and smith, "a guide to conservation planning," in Predicting rainfall
erosion losses , 1978.
[6] Renard and Freimund, "Using Monthly Precipitation Data to Estimate the R Factor
in the Revised USLE," Journal of Hydrology, Vols. vol. 157, no. 1, pp. pp. 287–306,,
1994.
[7] Roose E.J., Use of the Universal Soil Loss Equation to Predict Erosion in West Africa
in, Soil Erosion: Prediction and Control, Soil Conservation Society of America,
1977.
[8] Hurni, "Erosion-Productivity-Conservation Systems in Ethiopia," in Proceedings of
4th International Conference on Soil Conservation, Maracay, Venezuela, 1985.
[9] Wischmeier w.h., Johnson c.b., Cross b.v.,, "Soil Erodibility Nomograph for
Farmland and Construction Sites," Journal of Soil and Water Conservation, vol. vol.
26, p. pp. 189–193, 1971.
[10] Renard K.G., Freimund J.R, "Using Monthly Precipitation Data to Estimate the R
Factor in the Revised USLE," Journal of Hydrology, Vols. vol. 157,, no. no. 1,, p.
pp. 287–306, 1994..
[11] Renard K.G., Foster G.R., Weesies G.A, "Predicting Soil Erosion by Water," in
Guide to Conservation Planning With the Revised Universal Soil Loss Equation
(RUSLE), USA: Agricultural Research Service,, 1997.
[12] Dumas P., Printemps J.B, ""Assessment of Soil Erosion Using USLE Model and GIS
141 | P a g e
Prepared By: Kalid Hassen
for Integrated Watershed and Coastal Zone Management in the South Pacific
Islands”," in Proceedings Interparent, International Symposium in Pacific Rim,
Taipei, Taiwan,, 2010, pp. pp. 856–866,.
[13] Morgan RPC, Soil Erosion and Conservation., Silsoe College, 1994.
[14] USGS, 2013. [Online]. Available: http://landsat.usgs.gov/Landsat8 Using Product.
[Accessed 02 08 2021].
[15] H.-Q. Xu and B.-Q. Chen,, "Remote sensing of the urban heat island and its changes
in Xiamen City of SE China," Journal of Environmental Sciences,, Vols. vol. 16, ,
no. no. 2, pp. pp. 276–281,, 2004.
[16] Q. H. Weng, D. S. Lu, and J. Schubring,, "Estimation of land surface temperature-
vegetation abundance relationship for urban heat island studies," Remote Sensing of
Environment, Vols. Vol, 89,, no. no. 4, pp. pp. 467–483,, 2014.
[17] F. Wang, Z. Qin, C. Song, L. Tu, A. Karnieli, and S. Zhao,, "improved mono-window
algorithm for land surface temperature retrieval from landsat 8 thermal infrared
sensor data," Remote sensing, Vols. vol. 7, , no. no. 4,, p. pp. 4268–4289, 2015.
[18] J. C. Jimenez-Munoz, J. A. Sobrino, A. Gillespie, D. Sabol, and W. T. Gustafson,,
"“Improved land surface emissivities over agricultural areas using ASTER NDVI,"
Remote Sensing of Environment, Vols. vol. 103,, no. no. 4, pp. pp. 474–487, , 2006..
[19] J.A. Sobrino, J.C. Jim´enez-Mu˜noz, and L. Paolini,, "Land surface temperature
retrieval from LANDSAT TM5,," Remote Sensing of Environment, Vols. vol. 90,,
no. no. 4, , pp. pp. 434–440,, 2004.
[20] J. A. Sobrino and N. Raissouni,, "Toward remote sensing methods for land cover
dynamic monitoring: application to Morocco,," International Journal of Remote
Sensing, vol. 21, no. 2, p. 353–366, 2000.
[21] M. Stathopoulou and C. Cartalis,, "Daytime urban heat islands fromLandsat ETM+
and Corine land cover data: an application to major cities in Greece," Solar Energy,
vol. vol. 81, no. no. 3,, p. pp. 358–, 2007.
[22] B. L. Markham and J. L. Barker,, "Spectral characterization of the Landsat
ThematicMapper sensors," International Journal of Remote sensing, Vols. vol. 6, ,
no. no. 5,, p. pp. 697–716, 1985.
[23] Q. H. Weng, D. S. Lu, and J. Schubring, , "Estimation of land surface temperature-
142 | P a g e
Prepared By: Kalid Hassen
vegetation abundance relationship for urban heat island studies," Remote Sensing of
Environment, Vols. vol. 89,, no. no. 4,, pp. pp.467–483,., 2004.
143 | P a g e