W ATER O BSERVATION & I NFORMATION S YSTEM

Transcription

1 W ATER O BSERVATION & I NFORMATION S YSTEM AND P RODUCT P ORTFOLIO Version 2, September 2014 Technical Team of the TIGER-Net project:

2 As part of the TIGER initiative supported by ESA, the TIGER-NET project aims to support the assessment and monitoring of water resources from watershed to cross-border basin level delivering indispensable information for Integrated Water Resource Management in Africa through: 1. Development of an open-source Water Observation and Information Systems (WOIS) for monitoring, assessing and inventorying water resources in a cost-effective manner; 2. Capacity building and training of African water authorities and technical centers to fully exploit the increasing observation capacity offered by current and upcoming generations of satellites, including the Sentinel missions. The project was officially kicked off at the World Water Forum in Marseilles in March 2012 and provides ongoing development of the WOIS towards Sentinel processing capacity as well as uptake of new users with training and capacity building. For more information on the TIGER-Net project please visit 2

3 Table of Content The Water Observation and Information System... 1 Earth Observation Products... 3 Base Map Data... 3 High-resolution basin characterization... 5 Land degradation mapping (incl. vegetation indices)... 7 Medium resolution full basin characterization... 9 Water body and shoreline mapping Wetlands mapping Water quality monitoring (lake surface temperature, chlorophyll and sediment load) Hydrological monitoring (precipitation, evapotranspiration, soil moisture, water level) Hydrological modelling (scenario analysis & operational forecasting) Dynamic and historical flood monitoring Erosion potential mapping Urban sanitation planning support

4 WO IS The Water Observation and Information System The WOIS is a multipurpose system consisting of a storage container for the geodata, extraction and processing of the EO data through customized processing facilities, and integrative tools and models aimed at decision support e.g. Hydrological modelling and GIS embedded visualisation and analysis tools. The WOIS is designed around Quantum GIS (QGIS), which acts as the front-end Graphical-User-Interface (GUI). The other components of WOIS are: GRASS GIS (large toolbox of raster and vector analysis algorithms), BEAM and NEST (processing of optical, thermal and radar ESA data products), Orfeo Toolbox (high resolution image processing), Soil Water Assessment Tool SWAT (hydrological modeling), R scripts (statistical and graphical tools) and PostGIS (spatial database). The functionality from the different software components are integrated using the SEXTANTE spatial data analysis library Figure 1. Open- source software packages which provides a framework for incorporating algorithms from the integrated as part of the Water Observation and various providers (such as GRASS GIS, BEAM, etc.) into QGIS Information System (WOIS) plugin for QGIS. (Figure 1). A key advantage of SEXTANTE is the ability to seamlessly use functionalities from different providers for data processing and analysis, and via a unique QGIS plugin it is possible to sequentially combine algorithms from the different providers into wizard-based processing i.e. standardised workflows of complex tasks with instructions (Figure 2). The plugin has been used to generate a workflow library with step-by-step guidance for the users to extract specific water information products. The workflows are first and foremost a help for the novice and intermediate users, while the more advanced users may choose to explore the full suite of Figure 2. WOIS graphical user interface and the embedded workflow library. 2 1

5 WO IS algorithms and tools available from SEXTANTE in order to create their own workflows. Moreover, preparameterized SEXTANTE models for certain products are provided to enable an automatized production for operational usage. As part of the WOIS framework a PostGIS database is provided, enabling de-centralized operation of multiple user identities. A library of import/export function further ensures the integrate ability and/or connection to existing IT infrastructures and databases. SYSTEM KEY ADVANTAGES Cost and license free Open Source Easily transferable Easy to operate Capable of retrieving, storing and processing EO satellite data as well as integrating in-situ data producing EO-based water related information products integrating hydrological modeling functions supporting decisions based on full GIS framework mapping and reporting functionality integrating and linking to existing user systems scaling up for future applications and demands supporting the full observational capacity of the upcoming Sentinels Figure 3. Screenshot of WOIS d erived water stress anomaly product and single step workflow example for generating report on affected areas / population. 2 2

6 EO Products Earth Observation Products Base Map Data Product Description The WOIS basin characterizations components include an array of key data layers along with a ready-to-use GIS map template and automated query scripts (enabling non-gis staff) to provide water related information / summary reports in map and tabular for Integrated Water Resource Management. It builds on already existing national and/or international and/or global cartographic information and user-supplied geodata bases. Content Product Example Hydrological network Administrative boundaries Basin and sub-basin delineation Roads Villages, cities Main water relevant infrastructures Main service infrastructures Terrain elevation data Input Data Hydrological network: Aquastat, Global Lakes and Wetlands Database (GLWD), Global River Discharge Database, Vector Map (VMap), user supplied GIS data Administrative boundaries: GADM Global Administrative Database, Country Files (GNS), user supplied GIS data Basin and sub-basin delineation: HydroSHEDS, user supplied GIS data Roads: Vector Map (VMap), user supplied GIS data Villages, cities: Vector Map (VMap), user supplied GIS data Main water relevant infrastructures: Aquastat, user supplied GIS data Main service infrastructures: Vector Map (VMap), user supplied GIS data Terrain elevation data: SRTM Digital Elevation Data, ASTER GDEM, HYDRO1k Elevation Derivative Database, user supplied GIS data Additional Earth Observation and GIS data can be added if required Methodology Collection of input data from user-side or other data collections Geometric correction, adaptation of attributes, format conversion, metadata generation Assembling of data for base mapping Establishment of QGIS basemap templates and standardized legend files WOIS Workflow See Training Documentation for Product Group #2: Base mapping Spatial Coverage Volta Basin, Nile Basin, Lake Chad Basin, South Africa basins (selected), North Namibia Spatial resolution Depending on data layers for mapping scales 1: or smaller Temporal resolution (time period and frequency) various Coordinate Reference System 2 3

7 Geographic Lat/Lon (WGS84) or user defined Geometric accuracy (positioning scale) Depending on data layers Thematic accuracy (in %) Depending on data layers Data type Raster and Vector 2 4

8 EO Products High-resolution basin characterization Product Description The WOIS high-resolution land cover component provides a monitoring service at sub-basin level, including a recent high-resolution land cover inventory as well as long-term land cover change processes, such as effects of climate change or overexploitation (e.g. disappearance of vegetation, change of cropland area). Content Product Example Detailed land cover mapping of main areas of interest within the basin at high resolution (20m pixel size) describing the major features relevant for water management (irrigated and non-irrigated cropland, forests, urban areas, water bodies, etc.) Long-term change analysis providing a longterm assessment of the major changes occurred in the basin EO input data Optical, multispectral Earth Observation data in high spatial resolution (5-30m) Sentinel-2 type (e.g. SPOT, RapidEye, Landsat) Ideally one homogeneous, cloud-free data take in either dry- or wet-season Other Input Data Existing land cover information for calibration and validation (hydrographic features, irrigation areas, road infrastructure etc.) Methodology Radiometric and geometric co-registration of satellite imagery to establish consistent time-series imagery Automated calculation of biophysical parameters (vegetation index data) Hybrid pixel- and object-based image classification and change detection Visual image interpretation Metadata generation and internal product validation WOIS Workflow See Training Documentation for Product Group #5: High-resolution basin characterization Spatial Coverage Volta Basin, Nile Basin, Lake Chad Basin, South Africa basins (selected), North Namibia Spatial resolution 5-30 m pixel size, target map scale: 1: Minimum mapping unit: 0.5 hectares Temporal resolution (time period and frequency) Long-term changes (2 years or longer intervals) Coordinate Reference System Geographic Lat/Lon (WGS84) or user defined Geometric accuracy (positioning scale) Better than 1 Pixel Thematic accuracy (in %) Overall thematic accuracy >= 80% 2 5

9 Kappa coefficient >= 0.7 Data type Raster or Vector 2 6

10 EO Products Land degradation mapping (incl. vegetation indices) Product Description Land degradation and desertification ultimately result in long lasting and observable loss of vegetation cover and biomass productivity over time and in space. Satellite remote sensing is a viable option to monitor vegetation productivity on a long time scale and at a vast spatial scale and the WOIS land degradation mapping tool integrates statistical methods for dynamic analysis of vegetation to assess land degradation (i.e. loss of productivity) or greening at basin scale. Content Product Example Inter-annual vegetation trend analysis providing a long-term assessment of land degradation processes with and without controlling for water availability. Vegetation trends detected in this way can help pin-point areas of change, act as a firstline land degradation assessment tool and initial indicator of where to focus on further investigations and as the basis for possible interventions. EO input data Sentinel-3 type (e.g. Geoland 2 SPOT VGT NDVI) Satellite Rainfall Estimate (RFE) Other Input Data n/a Methodology The WOIS mapping method for land degradation is based on principles put forward in Huber, Fensholt et al. (2011) and Hellden and Tottrup (2008). Time series of NDVI and rainfall estimates for are used to analyse vegetation/rainfall correlation and control NDVI trends for variability in rainfall. NDVI residual time series, originating from regressing NDVI on rainfall is subsequently searched for significant long term trends in vegetation greenness which is not related to rainfall but possibly contributable to humans (e.g. population growth, poverty, unsustainable land use practices and urbanization). WOIS Workflow Product group #3: Medium resolution land degradation index Spatial Coverage Lake Chad Basin, Volta Basin, Nile Basin (Lake Victoria and Kagera subbasins) Spatial resolution 1000 m Temporal resolution (time period and frequency) Coordinate Reference System Geographic Lat/Lon (WGS84) or user defined Geometric accuracy (positioning scale) 0.5 x pixel size Thematic accuracy Qualitative validation performed by local experts has shown that there are physical causes behind the significant 2 7

11 vegetation trends, regardless whether they are positive or negative. Data type Raster 2 8

12 EO Products Medium resolution full basin characterization Product Description Identifying, delineating and mapping land cover is vital for water resource management as it establish the baseline information from which monitoring activities (change detection) can be performed, and provide the ground cover information needed for water resource management and planning activities. The aim of this product group is therefore to provide mapping and monitoring approaches for the characterization and mapping of relatively large benchmark areas using medium resolution optical EO data. Content Product Example This following information products are provided under this product group: Recent land cover and land use mapping over large benchmark areas at medium resolution and describing the major features of relevance for water management. Long-term inter-annual analysis of the major land cover changes within the basin. Intra annual (seasonal) biophysical characterisation of changes and landscape dynamics for monitoring water related events e.g. drought or change in water management. EO input data Sentinel-3 type (eg. MODIS and ENVISAT MERIS) Geoland 2 SPOT VGT FAPAR Other Input Data GlobCover (2006, 2009) MODIS Vegetation Continues Field (MOD44B) Global Map of Irrigation Areas Terrestrial Ecoregions of the world Methodology The WOIS includes several methods to derive medium resolution land cover characterisation products at the basin level. Key features are unsupervised and supervised classification routines as well as simple and advanced techniques for land cover change detection and monitoring of landscape seasonal dynamics. WOIS Workflow Product group #4: Medium resolution full basin characterization Spatial Coverage Volta Basin, Lake Chad (Chari-Logone subbasins), Nile Basin (Lake Victoria and Kagera subbasins), Namibia (Kavango basin) Spatial resolution 500 m Temporal resolution (time period and frequency) Recent land cover: 2012 Historic land cover: Two points in time (2000, 2006 and /or 2009) Seasonal Monitoring: 2011 and 2012 (reference period ) Coordinate Reference System Geographic Lat/Lon (WGS84) 2 9

13 Geometric accuracy (positioning scale) 0.5 x pixel size Thematic accuracy (in %) Products meets target accuracies of 80% overall accuracy (Kappa coefficient > 0.7) Data type Raster 2 10

14 EO Products Water body and shoreline mapping Product Description The WOIS water body mapping service offers a detailed mapping and monitoring of open water bodies and their seasonal changes. Water bodies include lakes and reservoirs with a minimum size of 1 ha, showing their location, extent and dynamics across dry- and wet-seasons. Content Product Example Dynamic mapping of open water bodies based on high resolution optical data identifying water bodies, reservoirs and lakes with a minimum size of 1 hectare Temporal monitoring of seasonal and longterm changes in water extent and shoreline EO input data Optical, multispectral Earth Observation data in high spatial resolution (5-30m) Sentinel-2 type (e.g. SPOT, RapidEye, Landsat) Future Sentinel-1 SAR foreseen Ideally one homogeneous, cloud-free data take per dry- and wet-season Other Input Data Existing information on water bodies for calibration and validation Methodology Radiometric and geometric co-registration of satellite imagery to establish consistent time-series imagery Automated calculation of biophysical parameters and spectral indices (vegetation index, soil water index) Index thresholding and spatial edge detection filtering Metadata generation and internal product validation WOIS Workflow See Training Documentation for Product Group #6: Small water body mapping Spatial Coverage Volta Basin, Nile Basin, Lake Chad Basin, South Africa basins (selected), North Namibia Spatial resolution 5-30 m pixel size, target map scale: 1: Minimum mapping unit: 1 hectare Temporal resolution (time period and frequency) Seasonal (dry- vs. wet-season) or long-term changes (2 years or longer intervals) Coordinate Reference System Geographic Lat/Lon (WGS84) or user defined Geometric accuracy (positioning scale) Better than 1 Pixel Thematic accuracy (in %) Overall thematic accuracy >= 90% Kappa coefficient >=

15 Data type Raster or Vector 2 12

16 EO Products Wetlands mapping Product Description The WOIS wetlands mapping and monitoring service provides a highly automated capacity for wetland identification from medium resolution optical data, covering both long-term development as well as seasonal changes of wetlands and permanent water bodies. Content Product Example Dynamic mapping of open water bodies based on high resolution optical data identifying water bodies, reservoirs and lakes with a minimum size of 1 hectare Temporal monitoring of seasonal and longterm changes in water extent. EO input data Optical, multispectral Earth Observation data in medium spatial resolution ( m) Sentinel 3 type (e.g. ENVISAT-MERIS, TERRA-MODIS) Future Sentinel-1 SAR foreseen The identification of wetland requires a temporally dense observation cycle with at least 3 cloud-free image acquisitions per season, in order to perform a reliably identification of wetlands that may be covered by water only for a limited time of the year. Other Input Data Existing information on wetlands for calibration and validation Methodology Radiometric and geometric co-registration of satellite imagery to establish consistent time-series imagery Automated calculation of biophysical parameters and spectral indices (vegetation index, soil water index, soiladjusted vegetation index, water index) Index thresholding and filtering Metadata generation and internal product validation WOIS Workflow See Training Documentation for Product Group #11: Wetlands mapping Spatial Coverage Volta Basin, Nile Basin, Lake Chad Basin, South Africa basins (selected), North Namibia Spatial resolution m pixel size, target map scale: 1: Minimum mapping unit: 25 hectare Temporal resolution (time period and frequency) Seasonal (dry- vs. wet-season) or long-term changes (2 years or longer intervals) Coordinate Reference System Geographic Lat/Lon (WGS84) or user defined Geometric accuracy (positioning scale) Better than 1 Pixel Thematic accuracy (in %) 2 13

17 Overall thematic accuracy >= 80% Kappa coefficient >= 0.7 Data type Raster or Vector 2 14

18 EO Products Water quality monitoring (lake surface temperature, chlorophyll and sediment load) Product Description Clean fresh water is a serious environmental challenge, and optical remote sensing has become an increasingly important to tool to monitor water quality on a regular basis. This product group provides operational and historical satellite derived water quality monitoring products for major lakes in Africa need for e.g. potential identification of point sources, establishment of possible correlation with regular cholera outbreaks, better understanding of eutrophication processes and regular reporting obligations. Content Product Example This product group encompass the following products: Chlorophyll concentration (CHL), Total Suspended Matter (TSM) Diffuse attenuation coefficient of the downwelling spectral irradiance at 490 nm (Kd490) Surface Water Temperature (SWT) EO input data Envisat MERIS (CHL,TSM, Kd490) Envisat AATSR (SWT) MODIS Aqua (CHL, Kd490, SWT) Other Input Data n.a. Methodology Information about water quality and temperature is derived from ENVISAT MERIS and AATSR (historical information) and MODIS AQUA (current information).the fundamental methodologies behind water quality and temperature retrievals are included in open source software packages from both ESA (BEAM) and NASA (SeaDAS). ENVISAT data is processed using WOIS embedded BEAM functionalities including the Eutrophic Lakes processor to derive water quality parameters from MERIS, and the SST processor to obtain surface water temperature from AATSR data. MODIS data on the other hand is processed and delivered using the L2 data processors available in SeaDAS. WOIS Workflow Product group #1: Large Lakes Water Quality and Temperature Spatial Coverage Lake Victoria and Lake Chad Spatial resolution Envisat MERIS: 300 m Envisat AATSR: 1000 m MODIS Aqua: 1000 m 2 15

19 Temporal resolution (time period and frequency) Envisat MERIS: monthly, ca Envisat AATSR: monthly, ca MODIS Aqua: monthly, 2012 to current Coordinate Reference System Geographic Lat/Lon (WGS84) Geometric accuracy (positioning scale) 0.5 x pixel size Thematic accuracy Surface water temperature: Assumed better than 1 degree Celsius Water quality: Depends on the performance of the processor for the specific lake, but generally spatio-temporal variations are in accordance with expected patterns and absolute values largely reside within the range of published numbers. Data type Raster 2 16

20 EO Products Hydrological monitoring (precipitation, evapotranspiration, soil moisture, water level) Product Description Precipitation, evapotranspiration and soil moisture climatologies are important for basin hydrology (e.g. by impacting runoff and streamflow) and for the current and future utilization potential of the land. Information on these parameters is therefore important for the basic basin characterization. Moreover they represent important parameters for calibration and validation of hydrological models. Content Product Example This product group encompass the following products: Precipitation Evapotranspiration Soil moisture Water levels EO input data Precipitation (Various multispectral sources) Evapotranspiration (Meteosat Second Generation SEVIRI) Soil Moisture (ENVISAT ASAR, Meteosat, ASCAT, Future Sentinel-1 foreseen) Water Levels (ENVISAT, Jason) Other Input Data n.a. Methodology The precipitation data is the satellite Rainfall Estimate (RFE) which is a blended gauge-satellite rainfall product produced by NOAA's Climate Prediction Center. Evapotranspiration (ET) is derived from the LSA-SAF instantaneous ET products produced every 30 minute and based on data from the Meteosat Second Generation (MSG) satellites and a simplified version of the TESSEL-SVAT method. The 1 km SAR Surface Soil Moisture (SSM) product is derived from the ENVISAT ASAR sensor operating in Global Monitoring (GM) mode (Pathe et al. 2009) based on the approach of Wagner et al. (1999). The 25km soil moisture water index data (SWI) are derived from surface soil moisture data (SSM), which are retrieved from the radar backscattering coefficients measured by the ASCAT (Advanced scatterometer) onboard the MetOp satellite using a change detection method, developed at the Institute of Photogrammetry and Remote Sensing (IPF), Vienna University of Technology. The lake water levels product provides the longest possible historical record of data based on the altimeter 2 17

21 measurements and lake gauges. For this product the ESA Rivers and Lakes [earth.esa.int/riverandlake] dataset is made available through the WOIS. WOIS Workflow PG #08: Hydrological Characterization Spatial Coverage Pan-African Spatial resolution Precipitation: 10 km Evapotranspiration: 4 km SoilMoisture: 1km and 25km Soil Water Index: 25km Water Level: Point data Temporal resolution (time period and frequency) Precipitation: Daily/10-day, present/operational Evapotranspiration: Daily, since March 2013 SoilMoisture: Decadal, 1km: 2004 Apr. 2012; 25km: Jan.2007 present/operational Water Level: Daily, Point data Coordinate Reference System Geographic Lat/Lon (WGS84) Geometric accuracy (positioning scale) 0.5 x pixel size Thematic accuracy All products are from 3rd parties and delivered as is i.e.in accordance with the official documentation and validation of the respective products. Data type Raster 2 18

22 EO Products Hydrological modelling (scenario analysis & operational forecasting) Product Description Product group 9 provides hydrological modelling and forecasting facilities. The key outputs are simulated river discharge time series at various locations in the basins. Content Product Example This product group has an offline model development workflow (MDWF) and an operational simulation and forecasting workflow (OSFWF). The MDWF uses the hydrological modelling software package SWAT accessed through the GUI MWSWAT. The user can manually insert requested subcatchment outlet points. The SWAT models are calibrated for the requested Technical Demonstration Cases and technical guidance on the calibration of users own SWAT models is provided. Guidance on the setup of typical scenario runs (e.g. construction of new dams) is also provided. The OSFWF uses historical or forecasted gridded data products for precipitation and reference evapotranspiration forcings. Forcing time series are produced for all subcatchments. The models developed in the MDWF are subsequently run with the forcing time series and simulated discharge time series are produced. The MDWF includes a data assimilation facility, which ingests in-situ discharge data. The end product of the OSFWF is a set of simulated time series of river discharge at various points in the basins (incl. uncertainty intervals). EO input data SRTM ACE2 digital elevation model FEWS-RFE precipitation estimates Other Input Data In-situ discharge data from user organisation or the Global Runoff Data Centre ECMWF ERA interim daily minimum and maximum temperature used for computing reference ET NOAA-GFS precipitation and temperature forecast (the latter used for computing reference ET). Global Land Cover Facility landcover map FAO/UNESCO soil map Methodology The hydrological models are set up using the semi-distributed, physically based hydrological simulation software SWAT (Soil and Water Assessment Tool), developed by the US Department of Agriculture. Required static inputs for SWAT are elevation, land cover and soil type. Required time-variable forcings are precipitation and reference ET. Subcatchment discretization is guided by the available in-situ discharge stations, location of reservoirs and other points of interest. Data assimilation techniques are used to improve the performance of operational discharge forecasting. Insitu discharge data are used to update the model. WOIS Workflow Product group #9: Hydrological modelling Spatial Coverage Chari-Logone River basin, Mokolo River basin and Kavango River basin Spatial resolution Subcatchment level. The size of the subcatchments is different for each Technical Demonstration Case. Subcatchment sizes vary from thousands to tens of thousands of square kilometres. 2 19

23 Temporal resolution (time period and frequency) The hydrological models run at daily time steps. The simulation period is determined by the availability of the RFE rainfall product (1/1/2001 to present). For some Technical Demonstration Cases, the hydrological models will be run in forecasting mode to present + 7 days. Coordinate Reference System UTM WGS84 Geometric accuracy (positioning scale) Not applicable Thematic accuracy (in %) Not applicable Data type Time series of simulated discharge m3/s. 2 20

24 EO Products Dynamic and historical flood monitoring Product Description This product group provides information on dynamic and historical flood events. For flood mapping, the C-band backscatter data from high to medium resolution Synthetic Aperture Radars (SAR) are used. The high spatial resolution and their ability to discriminate land/water boundaries in all weather conditions, day and night, make SARs highly effective tools for mapping and monitoring of flood conditions. Content Product Example The historical flood mapping is based on analysis of the backscatter data acquired from the Advanced Synthetic Aperture Radar (ASAR) aboard Envisat satellite. The dynamic flood mapping component will provide delineations of current events based on RADARSAT-2 SAR data (future Sentinel-1 foreseen) acquisitions in response to real flood events. The flood maps are confined to flood-prone areas, as identified through digital elevation data using the Height Above Nearest Drainage (HAND) index. EO input data Sentinel-1 Envisat ASAR WS RADARSAT-2 Other Input Data SRTM Digital Elevation Model (DEM) data Methodology The flood mapping methodology used in TIGER-Net WOIS system consists of pre-processing, classification and postprocessing steps. In the pre-processing step, precise orbit vectors and Range-Doppler terrain correction are applied to obtain a georeferenced SAR image. The classification module of the WOIS workflow relies on the unique backscattering properties of water surfaces, which appear dark to the human eye in SAR imagery. Within the WOIS, both an automatic and a manual thresholding approach are implemented. Finally, in order to provide a mask of areas which are not prone to flooding, the Height Above Nearest Drainage (HAND) index is used which is a hydrologically conditioned DEM derived from the SRTM DEM. WOIS Workflow Product group #7: Flood Mapping System Spatial Coverage Selected regions in Africa (demo cases exist for Nile Basin, Komadougou-Yobe sub-basin, Caprivi Namibia) Spatial resolution Envisat ASAR Wide Swath (WS) mode: approx.150 m RADARSAT-2 Standard beam mode (SGF): approx. 25m Temporal resolution (time period and frequency) Envisat ASAR WS: The whole available Envisat life-time data acquisitions ( ). RADARSAT-2: Single scenes obtained via SOAR orders triggered by users during the TIGER-Net project period. Future synoptic capacity to be provided through Sentinel-1 Coordinate Reference System 2 21

25 Geographic Lat/Lon (WGS84) Geometric accuracy (positioning scale) 0.5 x pixel size Thematic accuracy (in %) Overall accuracy between the satellite area and ground truth is above 90%. As an alternative measure, the Kappa coefficient is better than 0.6. Data type Raster 2 22

26 EO Products Erosion potential mapping Product Description The WOIS erosion potential mapping service establishes indicators to estimate soil erosion potential over time to identify areas with high erosion potential for planning and prioritizing of watershed restoration activities. Content Product Example Dynamic mapping of open water bodies based on high resolution optical data identifying water bodies, reservoirs and lakes with a minimum size of 1 hectare Temporal monitoring of seasonal and longterm changes in water extent. EO input data EO data required for medium-resolution land cover and land degradation converted into coefficients Mean yearly rainfall data Vegetation index data (mean NDVI) as proxy for vegetation cover Other Input Data Terrain model and derived slope and slope length (30m resolution) Soil type data for erodibility coefficients Methodology Preparation of input data regarding rainfall pattern, soil type, topography, crop system and management practices. Ingestion of all input data and calculation of erodibility coefficients using a GIS model based on the Universal Soil Loss Equation (USLE) Calculation of long term average annual rates of erosion and compare soil losses from a particular field with a specific crop and management system to tolerable soil loss rates. Alternative management and crop systems can be evaluated to determine the adequacy of conservation measures in farm planning. WOIS Workflow See Training Documentation for Product Group #12: Erosion potential mapping Spatial Coverage Volta Basin, Nile Basin, Lake Chad Basin, South Africa basins (selected), North Namibia Spatial resolution m pixel size, target map scale: 1: Temporal resolution (time period and frequency) Yearly changes of soil erosion potential Coordinate Reference System Geographic Lat/Lon (WGS84) or user defined Geometric accuracy (positioning scale) Better than 1 Pixel Thematic accuracy (in %) Depending on input data layers 2 23

27 Data type Raster 2 24

28 Urban sanitation planning support Product Description The WOIS Urban sanitation planning support is a mapping and modelling tool to estimate the water demand for sanitation associated with urban environments in Africa Content The water supply and sanitation planning tool has a model and mapping component i.e.: Model environment for estimating current and future demand for urban water use. Urban land cover map with classes relevant for urban water and sanitation demand. Product Example EO input data Optical, multispectral Earth Observation data in high spatial resolution Other Input Data Census data, public databases and city administration reports Sub-urban delineations Administrative boundaries Street layer data ( French data set ) DGPS points at visible points (e.g. Street corner,) with photos for geo-referencing City maps Digital Terrain Model Methodology Preprocessing of census and EO data Modelling of current and future urban water demand based on census data Mapping of land cover and city blocks Assigning of modelled water demand values to building block and land cover classes by application of a customized GIS tool WOIS Workflow Product group #10: Water supply and sanitation planning support (beta) Spatial Coverage Urban regions of N Djameny, Chad Spatial resolution Better than 2.5 m pixel size, target map scale: 1:10.000; MMU: 0.1 hectares Temporal resolution (time period and frequency) Recent (2012) Coordinate Reference System Geographic Lat/Lon (WGS84) or user defined Geometric accuracy (positioning scale) Better than 0,5 Pixel Thematic accuracy (in %) Urban land cover map: Overall accuracy between the satellite area and ground truth is above 90%. As an alternative measure, the Kappa coefficient is better than 0.8. Water demand model: Estimated to be within 10% of reported total water demand for N Djamena city. 2 25

29 Data type Urban land cover map: Vector or Raster Model: Microsoft Excel 2 26