Deliverable

2.3 Climatic, hydrologic and remote

sensing data ready for upload

Summary: The IMDROFLOOD geoportal will provide data discovery and access
interfaces for a number of relevant repositories of satellite, modeled and in-situ data,
making use of information and knowledge generated by the consortium team for the
selected case studies. The georeferenced repositories are built in the framework
defined by river basin scale. Input station observations for data products ready for
upload are quality controlled and homogenized. In some cases, interpolation
procedures are applied and relevant indices are built for selected river basins. Input
satellite measurements are used to derive spectral vegetation indices for monitoring
droughts in the selected river basins. The IMDROFLOOD geoportal content will be
continuously updated with flood and drought related indices during the life time of the
project.

Completed: June/2017
1. Introduction
The IMDROFLOOD geoportal will provide data discovery and access interfaces for a
number of relevant repositories of satellite, modeled and in-situ data, making use of
information and knowledge generated by the consortium team for the selected case
studies. The georeferenced repositories are built in the framework defined by river basin
scale.

2. Data quality control and homogenization

procedures
2.1. Data quality control

Input data are quality controlled by national meteorological and hydrological services.
In some cases (e.g. data covering the regions from Ukraine and Republic of Moldova
situated in the river Prut basin), basic quality control and data filling of the missing
records were realized with the software Climate Tools for Series Homogenization and
Derived Products (CLIMATOL). The CLIMATOL is a contributed R package holding
functions for climatological series homogenization and filling missing values, which
allows to use of other synchronous data from nearby stations, using a form of
orthogonal regression known as Reduced Major Axis (Guijarro 2016).

2.2. Homogenization

Before any climate analyses based on observed time series, one has to perform
inhomogeneity corrections. In general, monthly temperature and precipitation data
have been homogenized using HOMogenizaton softwarE in R (HOMER). HOMER
method includes the best features of some other state-of-the-art methods, namely
PRODIGE (Caussinus and Mestre, 2004), ACMANT (Domonkos, 2011), and cghseg a joint
segmentation method that was developed originally by bio-statisticians in the context
of DNA segmentation (Picard et al., 2011). Daily data from the Prut basin were
homogenized using the CLIMATOL (Guijarro, 2011).

2.3. Interpolation

Several gridded data sets have been built from daily station data to cover selected river
basins (e.g. the Prut River Basin) at high spatial resolution (e.g. 1km x 1km and 10 km x
10 km). For instance, in the Prut River basin we used several auxiliary variables (with the
continuous spatial distribution) as ancillary data to build the maps of the gridded daily
air temperature and precipitation at spatial resolution of 1km x 1km covering the
interval 2000-2016. For temperature, two categories of ancillary data were selected:
DEM-derived predictors based on the 1-km average of topography from the USGS
SRTM30 DEM (Jarvis et al. 2008) and 2-meter air temperature (T2M) extracted from
MERRA-2 reanalysis (Molod et al. 2015). The purpose of the DEM-derived predictors

Completed: June/2017
(altitude, latitude, distance to the Black Sea) was to take into account both the
latitudinal and altitudinal distribution of air temperature, as well as the direct influence
of the major water body which moderate temperatures in inland areas. In order to
improve the predictive power of the regression models, a PCA was applied on DEM-
derived predictors, for removing the possible effect of multicollinearity between them.
In addition to the three "static" predictors, T2M variable extracted from the MERRA-2
was used as auxiliary field which can provide valuable meteorological information for
regions with scarce observations. MERRA-2 integrates a variety of observing systems
with numerical models to produce a high-quality atmospheric and surface weather
history dataset. It has a native resolution of 0.5° lat. x 0.625° long. x 72 hybrid
sigma/pressure levels. The MERRA-2 collection contains hourly air temperature data
(T2M) beginning in 1980 (Bosilovich, Lucchesi, and Suarez 2015).

Figure 1. Multiannual mean of air temperature in the Prut basin (2000 – 2016) from the
daily gridded data set with spatial resolution of 1 km x 1 km.

By using a quasi–physically based meteorological model (MicroMet), we adjusted the

MERRA-2 meteorological variable beyond its initial spatial interpolation to 1000 × 1000
m spatial resolution. The MicroMet model quantify the strong temperature–elevation
relationships that are known to exist (Liston and Elder 2006).

Completed: June/2017
In the Prut basin, gridded precipitation fields were estimated by combining precipitation
measured at the ground and weather radar reflectivity fields, hence radar information
represent other auxiliary variable which have been used to achieve the daily
precipitation maps. The continuous precipitation information was provided by the
Birnova WSR-98D radar from the Romanian National Radar Network, which currently
estimate the rainfall rate using the NEXRAD Z–R (reflectivity-rainfall rate) relation
(Burcea et al. 2012).

Figure 2. Multiannual mean of air temperature in the Prut basin (1961 – 2016) from the
daily gridded data set with spatial resolution of 10 km x 10 km.

Gridded daily air temperature and precipitation covering the interval 1961-2016 in the
Prut basin at spatial resolution of 10 km x 10 km were built using as ancillary data only
DEM-derived predictors based on the 1-km average of topography from the USGS
SRTM30 DEM (Jarvis et al. 2008).

3. Drought and flood-related variables

The IMDROFLOOD geospatial system will be developed following a distributed
architecture (Figure 1). The main components of the system are listed below:

Completed: June/2017
 3.1. Hydroclimate data

The IMDROFLOOD geospatial system integrates climate-derived induces starting from

input data from several sources: (1) national meteorological services from the countries
of the IMDROFLOOD consortium; (2) the Global Historical Climatology Network-Daily
(GHCND-D) archive. GHCN-D is an integrated database of daily climate summaries from
land surface stations across the globe which are subjected to a common suite of quality
assurance reviews (Menne et al. 2012). The IMDROFLOOD geospatial system integrates
hydrological data from national hydrological services from the countries of the
IMDROFLOOD consortium.

Input climate data for monthly drought index ready for upload for the Emajõgi river
basin are presented in Table 1. In this case, the drought index covers the time interval
1951-2015.

Table 1. Input climate data for the climate-derived drought and flood indices for the
Emajõgi river basin.
No. Variable name and Time step Spatial Time interval Quality Homogenization Data
unit resolution control File
format

1 Temperature (°C) daily 2 stations 1951-2015 y y EXCEL

2 Precipitation (mm) daily 4 stations 1951-2015 Y Y EXCEL

3 Sea-level pressure monthly 1 station 1951-2015 Y Y EXCEL

(hPa)

4 Relative humidity monthly 9 stations 1951-2015 Y Y EXCEL

(%)

5 Mean maximum monthly 9 stations 1951-2015 Y Y EXCEL

temperature (°C)

7 Mean minimum monthly 9 stations 1951-2015 Y Y EXCEL

temperature (°C)

8 Wind speed (m/s) monthly 9 stations 1951-2015 Y Y EXCEL

9 Sunshine duration monthly 9 stations 1951-2015 Y Y EXCEL

(hours)

10 Precipitation (mm) monthly 9 stations 1951-2015 Y Y EXCEL

Input climate data for monthly drought indices such as the Standardized Precipitation
Evapotranspiration Index (SPEI) (Vicente-Serrano et al 2010), ready for upload for the
Prut river basin are presented in figure 3. In this case the drought index covers the time
interval 1961-2016. For the gridded product of SPEI drought index ready for upload we
used daily temperature from 35 stations and daily precipitation from 51 stations, all of
them for the time interval 1961-2016. An example from the SPEI gridded data set

Completed: June/2017
covering the Prut basin with the spatial resolution of 10 km x 10 km is presented in Figure
4.

Figure 3. Location of stations used for the gridded products of temperature,

precipitation and SPEI covering the Prut river basin at the spatial resolution of 10
km x 10 km for the time interval 1961-2016.

Figure 4 SPEI indices in the Prut river basin for August 2010.

Completed: June/2017
 N Variable Time step Spatial Time interval Quality Homogeniz Data File format
r. name and resolution & control ation (e.g ASCII, raster
unit location (yes/no (yes/no) etc)
)

1. ET0 daily 532 stations 1989-2016 Y Y .CSV

data
Iberian basins

2. Temperature daily 1 km x 1 km 2000-2016 Y Y NetCDF

(ºC) Prut basin

3 Temperature daily 10 km x 10 1961-2016 Y Y NetCDF

(ºC) km
Prut basin

4 Evapotranspi monthly 10 km x 10 1961-2016 Y Y NetCDF

ration (mm) km
Prut basin

5 SPEI monthly 10 km x 10 1961-2016 Y Y NetCDF

km
Prut basin

6 ETO daily 9 stations 1951-2015 Y Y .CSV

Emajõgi basin

7 River runoff Monthly 186 stations 1911-2013 Y Y .CSV

(m3/s) data Iberian
basins

8 River runoff Daily 1 stations 1951-2015 Y Y .CSV

(m3/s) data Emajõgi
basin

Table 2. Climate-related data ready for upload on the IMDROFLOOD geospatial

portal.

4.2. Satellite data

The IMDROFLOOD geospatial system integrates satellite data from the MODIS products.
Spectral vegetation indices are among the most commonly used satellite data products
for the evaluation, monitoring, and measurement of vegetation cover, condition,
biophysical processes, and change. They have been used for over last decades in a broad
variety of applications, including monitoring the effects of drought over regional,
national, and even multinational areas (Brown et al., 2008). The architecture of

Completed: June/2017
vegetation cover varies in time, from fractions of seconds and minutes (wind, water
stress, etc.) to seasonal (phenological evolution, environmental constraints) and years
(ecosystem dynamics) (Weiss et al., 2004). Vegetation indices are a subset of spectral
indices and are based on spectral responses of the objects that interact with the incident
solar radiation.

Nr. Variable name Time step Spatial resolution Time interval Data File format
&
location

1. NDWI 8-days 1 km x 1 km 2000-2017 geotiff

Prut basin

2. 8-days 1 km x 1 km 2000-2017 geotiff

NDVI Prut basin

3 NDDI 8-days 1 km x 1 km 2000-2017 geotiff

Prut basin

4 NDII 8-days 1 km x 1 km 2000-2017 geotiff

Prut basin

5 LAI 8-days 1 km x 1 km 2000-2017 geotiff

Prut basin

6 fAPAR 8-days 1 km x 1 km 2000-2017 geotiff

Prut basin

Table 3. Satellite-derived indices for droughts ready for upload on the IMDROFLOOD
geospatial portal.

We used for the IMDROFLOOD project the data collected by Moderate Resolution
Imaging Spectreradiometer (MODIS) instrument, aboard the Terra and Aqua earth
observation system satellites. Terra's orbit around the Earth is timed so that it passes
from north to south across the equator in the morning, while Aqua passes south to north
over the equator in the afternoon. Terra MODIS and Aqua MODIS are viewing the entire
Earth's surface every 1 to 2 days.

The MODIS data products collection 6, h19v04 and h20v04 tiles used to compute the
vegetation indices and biophysical variables have been downloaded from
http://glovis.usgs.gov/. MOD09A1 provides MODIS band 1-7 surface reflectance. It is a
level-3 composite of 500 m resolution. Each product pixel contains the best possible
level-2 observation during an 8-day period as selected on the basis of high observation
coverage, low view angle, absence of clouds or cloud shadow, and aerosol loading

Completed: June/2017
(MOD09 User guide, v1_3). MOD15A2 products provide global LAI and FPAR fields
retrieved from atmospherically corrected Bidirectional Reflectance Factors (MOD09
Surface Reflectance Product), using up to 7 spectral bands. The resolution of the data is
1km and the temporal frequencies are 1 and 8 days (MOD15 User guide). The 500m
MODIS surface reflectance products (MOD9A1) have been used to obtain the
Normalized Difference Vegetation Index - NDVI, Normalized Difference Water Index -
NDWI, Normalized Difference Drought Index - NDDI and Normalized Difference Infrared
Index – NDII. The images were processed in specific remote sensing and GIS software.

Figure 5. Example of NDWI product integrated into IMDROFLOOD geoportal.

Figure 6. Example of NDWI product integrated into IMDROFLOOD geoportal.

Completed: June/2017
NDWI is a good indicator of water content of leaves and it can be used for detecting and
monitoring the vegetation cover humidity. It is known that vegetation is affected by
water stress during dry or very wet periods, which influence plant development and can
cause its damage. In his study Gu et al. (2007) showed that NDWI has a stronger
response to drought conditions than NDVI. The average of NDVI and NDWI were
consistently lower (NDVI<0.5 and NDWI<0.3) under drought conditions than under non-
drought conditions (NDVI>0.6 and NDWI>0.4). 8-day MODIS composite NDWI, covering
the entire Prut basin, was calculated for the time interval February 2000 - February 2017
and integrated into the IMDROFLOOD geoportal (Figure 5).

NDVI is directly related to the: leaf area index (LAI), which is often used in crop growth
models; herbaceous or total green biomass (tons/ha) for given vegetation types,
photosynthetic activity of the vegetation, percent ground cover. Indirectly, NDVI has
been used to estimate the cumulative effective of rainfall on vegetation over a certain
time period, rangeland carrying capacity, crop yields for different crop types, and the
quality of the environment as habitat for various animals, pests and diseases (Kriegler
et al., 1969). 8-day MODIS composite NDVI, covering the entire Prut basin, was
calculated for the time interval February 2000 - February 2017 and integrated into the
IMDROFLOOD geoportal (Figure 6).

The NDII was developed to monitor leaf water content (Hardisky et al., 1983), so there
is a direct relation between NDII and root zone moisture deficit. On the other hand the
deficit is a direct function of the amount of moisture stored in the root zone. In this
context, if leaf water thickness and the suction pressure in the root zone are connected,
then the NDII would directly reflect the moisture content of the root zone. So, the NDII
would allow us to see vegetation like a natural manometer, providing information about
how much water is available in the sub-surface for use by vegetation. It can be an
integrated indicator of soil moisture in the root zone, available directly at the scale of
interest (Sriwongsitanon et al., 2016). 8-day MODIS composite NDII, covering the entire
Prut basin, was calculated for the time interval February 2000 - February 2017 and
integrated into the IMDROFLOOD geoportal.

NDDI can offer an appropriate measure of the dryness of a particular area, because it
combines information on both vegetation and water. NDDI takes advantage of the fact
NDVI senses plant matter and NDWI senses plant moisture. NDDI had a stronger
response to summer drought conditions than a simple difference between NDVI and
NDWI, and is therefore a more sensitive indicator of drought than NDVI alone (Gu et al.,
2007). 8-day MODIS composite NDDI, covering the entire Prut basin, was calculated for
the time interval February 2000 - February 2017 and integrated into the IMDROFLOOD
geoportal.

Completed: June/2017
Figure 7. Example of LAI product integrated into IMDROFLOOD geoportal.

Figure 8. Example of fPAR product integrated into IMDROFLOOD geoportal.

The leaf area index (LAI) is an important vegetation parameter, usually defined as the
one sided area of leaves per unit ground area (Chen and Black 1992). LAI appears as a
key variable in many models describing vegetation-atmosphere interactions, particularly
with respect to the carbon and water cycles. The interest in information on LAI
distribution and changes has grown substantially in recent decades, due to its intrinsic
importance and the emerging capability for LAI estimation over large areas using
satellite measurements. The LAI parameter reflects the biochemical and physiological
processes of vegetation, therefore indicating the productivity of vegetation, and it

Completed: June/2017
serves as an important input variable in land surface process models. Fang et al. (2011)
highlight that to understand the LAI for crops and its dynamics is very important for a
wide range of agricultural studies, such as crop growth monitoring and crop yield
estimation. 8-day MODIS composite LAI, covering the entire Prut basin, was calculated
for the time interval February 2000 - February 2017 and integrated into the
IMDROFLOOD geoportal (Figure 7).

The Fraction of Absorbed Photosynthetically Active Radiation (FAPAR) plays a critical

role in the energy balance of ecosystems and in the estimation of the carbon balance
over a range of temporal and spatial resolutions (GTOS, 2008). It is defined like the
fraction of the incoming solar radiation in the Photosynthetically Active Radiation
spectral region that is absorbed by a photosynthetic organism, typically describing the
light absorption across an integrated plant canopy. This biophysical variable is directly
related to the primary productivity of photosynthesis and some models use it to
estimate the assimilation of carbon dioxide in vegetation. Satellite data measured by
optical remote sensing sensors are the most appropriate for monitoring the FAPAR by
vegetated surfaces. Also, 8-day MODIS composite FAPAR, covering the entire Prut basin,
was calculated for the time interval February 2000 - February 2017 and integrated into
the IMDROFLOOD geoportal (Figure 8).

4.3. Drought and flood-related data ready to upload

The hydroclimate-related data ready for upload on the IMDROFLOOD geospatial system
at this stage are presented in table 2. The satellite-derived data ready for upload on the
IMDROFLOOD geospatial system at this stage are presented in table 3.

5. Conclusions
The IMDROFLOOD geoportal content offers advanced mapping and visualization
capabilities in a virtual environment for a well-defined thematic scope: drought and
flood early warning and monitoring using real-time hydroclimate and satellite indicators
at the river basin scale. A user-friendly web-based cartographic client is built to allow
the users to inspect the hydroclimate and satellite indices. The IMDROFLOOD
SDI/geoportal content will be continuously updated with flood and drought related
indices during the life time of the project.

6. References
Bosilovich, MG, R Lucchesi, and M Suarez. 2015. “MERRA-2: File Specification.” 9.
Greenbelt, Maryland 20771: Global Modeling; Assimilation Office, Earth Sciences
Division, NASA Goddard Space Flight Center.

Brown, J., Jenkerson, C., Gu, Y. (2008), Using eMODIS Vegetation Indices for operational
drought monitoring, National Integrated Drought Information System Knowledge

Completed: June/2017
Assessment Workshop: Contributions of Satellite Remote Sensing to Drought
Monitoring, February 6-8.

Burcea, Sorin, Sorin Cheval, Alexandru Dumitrescu, Bogdan Antonescu, Aurora Bell, and
Traian Breza. 2012. “Comparison Between Radar Estimated and Rain Gauge Measured
Precipitation in the Moldavian Plateau.” Environmental Engineering and Management
Journal 11 (4): 723–31.

Caussinus H. and Mestre O. (2004) Detection and correction of artificial shifts in climate
series. Applied Statistics, 53, part 3, 405-425.

Chen, J.M., Black. T. (1992), Defining leaf area index for non-flat leaves, Plant Cell
Environ., 15, pp. 421–429

Fang, H., Liang, S., Hoogenboom, G. (2011), Integration of MODIS LAI and vegetation
index products with the CSM-CERES Maize model for corn yield estimation, Int. J.
Remote Sens., 32, pp. 1039–1065

Domonkos P., R. Poza, and D. Efthymiadis (2011) Newest development of ACMANT. dv.
Sci. Res., 6, 7-11, 2011. doi:10.5194/asr-6-7-2011

GTOS (2008), Terrestrial essential climate variables – for climate change assessment,
mitigation and adaptation, Biennial report supplement GTOS, 44

GTOS (2009), Assessment of the status of the development of the standards for the
Terrestrial Essential Climate Variables: FAPAR - Fraction of Absorbed Photosynthetically
Active Radiation, 65.

Gu, Y., Brown, J., Verdin, J., Wardlow, B. (2007), A five-year analysis of MODIS NDVI and
VDWI for grassland drought assessment over the central Great Plains of the United
States, Geophys. Res. Lett., 34, L06407, doi: 10.1029/2006GL029127

Guijarro J.A. (2011): User's guide to Climatol, 40 pp.

http://www.meteobal.com/climatol/climatol-guide.pdf

Hardinsky, M.A., Klemas, V., Smart, R.M. (1983), The influence of soil salinity, growth
form, and leaf moisture on the spectral radiance of Spartina alterniflora canopies,
Photogramm. Eng. Remote Sens., 49, pp: 77-83

Jarvis, Andy, Hannes Isaak Reuter, Andrew Nelson, and Edward Guevara. 2008. “Hole-
Filled Seamless Srtm Data V4, International Centre for Tropical Agriculture (Ciat).”

Kriegler, F.J., Malila, W.A., Nalepka, R.F., Richardson, W. (1969), Preprocessing

transformations and their effects on multispectral recognition, in: Proceedings of the
Sixth International Symposium on Remote Sensing of Environment, University of
Michigan, Ann Arbor, MI, pp: 97-131

Completed: June/2017
Liston, Glen E, and Kelly Elder. 2006. “A Meteorological Distribution System for High-
Resolution Terrestrial Modeling (Micromet).” Journal of Hydrometeorology 7 (2): 217–
34.

Menne, Matthew J, Imke Durre, Russell S Vose, Byron E Gleason, and Tamara G Houston.
2012. “An Overview of the Global Historical Climatology Network-Daily Database.”
Journal of Atmospheric and Oceanic Technology 29 (7): 897–910.

Mestre, O. et al. (2013). HOMER : HOMogenisation softwarE in R- methods and

applications. Idöjárás, 117, 47–67.

Molod, A., L. Takacs, M. Suarez, and J. Bacmeister. 2015. “Development of the Geos-5
Atmospheric General Circulation Model: Evolution from Merra to Merra2.” Geoscientific
Model Development 8 (5): 1339–56. doi:10.5194/gmd-8-1339-2015.

Picard F., Lebarbier, E., Hoebeke, M., Rigaill, G., Thiam B. and Robin S. (2011). Joint
segmentation, calling, and normalization of multiple CGH profiles. Biostatistics.
doi:10.1093/biostatistics/kxq076

Sriwongsitanon, N., Gao, H., Savenije, H.H.G., Maekan, E., Saengsawang, S.,
Thianpopirug, S. (2016), Comparing the Normalized Difference Infrared Index (NDII) with
root zone storage in a lumped conceptual model, Hydrol. Earth Syst. Sci., 20, pp: 3361-
3377

Vicente-Serrano S.M., Santiago Beguería, Juan I. López-Moreno, (2010) A Multi-scalar

drought index sensitive to global warming: The Standardized Precipitation
Evapotranspiration Index - SPEI. Journal of Climate 23: 1696-1718.

Weiss, M., Baret, F., Smith, G. J., Jonckheere, I., Coppin, P. (2004), Review of methods
for in situ leaf area index determination, Part II: Estimation of LAI, errors and sampling,
Agricult. Forest Meteorol., 121, pp: 37–53

Completed: June/2017