Gateway for:

Member Countries

Spatial Data Infrastructure for mountainous terrain

#A-2233


Developing Spatial Data Infrastructure: an effective tool for assessing erosion of mountainous territories.

Tech Area / Field

  • AGR-DIG/Diagnostics/Agriculture
  • ENV-MRA/Modelling and Risk Assessment/Environment
  • INF-COM/High Performance Computing and Networking/Information and Communications

Status
3 Approved without Funding

Registration date
21.12.2015

Leading Institute
Center for Ecological-Noosphere Studies National Academy of Sciences of the Republic of Armenia, Armenia, Yerevan

Supporting institutes

  • Desing and Institute FAZO, Tajikistan, Dushanbe\nGIS and RS Consulting Center GeoGraphic, Georgia, Tbilisi

Collaborators

  • University of California, USA, CA, Santa Barbara\nDigitalGlobe, USA, CO, Longmont\nUniversity College Dublin, National University of Ireland, Ireland, Dublin\nVerico SCE, Germany, Freising-Weihenstephan\nDeutche Gesellschalft fur Zusammenabeit (GIZ) GmbH, Georgia, Tbilisi, Tbilisi\nConsultITPro, Slovakia, Banska Stiavnica\nWest Virginia University, USA, WV, Morgantown

Project summary

The Project aim. The overall objective of this project is to strengthen capacities in environmental monitoring and modelling in South Caucasian (Armenia, Georgia) and Central Asian (Tajikistan) countries by adopting an integrated approach including improved data management, data processing and research.
Scope of activities. The specific objectives can be summarized as follows:
    - Adopting an integrated approach of monitoring the ecological state of mountainous farmlands in South Caucasus and Central Asia using high-resolution Remote Sensing Data (RSD).
    - Improving a state-of-the-art Spatial Data Infrastructure (SDI) at the leading Armenian, Georgian and Tajik organizations able to analyze the data and present the results to stakeholders.
    - Assessing the land degradation/soil erosion risk of mountainous agricultural territories according to the previously stated degradation indices (vegetation cover, productivity, soil texture, humus content, organic carbon etc.)
    - To improve national and international networking with key Institutions and programs, by being active in international Initiatives such as Group of Earth Observation (GEO).
Expected results and their application. The implementation of the project will lead to the development of expected results of the Project will include (i) a time- and cost-efficient method of early identification of land degradation, which will serve as a prime tool for developing a remote monitoring system for agricultural lands, (ii) a joint database of spectral signatures and characteristics of land degradation/soil erosion in South Caucasus and Central Asia, (iii) an Environment Oriented Satellite Image Processing Platform, (iv) a joint virtual research lab between CENS, “GEOGRAPHIC” and “FAZO” in the area of application of SDI geoprocessing capabilities for environmental research.
Technical approach and methodology.The most common methods used to assess land degradation are expert opinions, land users’ opinions, field monitoring, observations and measurement, modeling, estimates of productivity changes and remote sensing. A crucial parameter triggering soil erosion that can be derived from satellite imagery is fractional vegetation cover (FVC). Normalized Difference Vegetation Index (NDVI) and spectral unmixing approaches will be used in order to get vegetation abundance maps for study areas, which later will be converted to the FVC map. According to FVC map, the sampling sites will be defined. The final protocols and quality control assurance plans will then be developed. The mountain environment exhibits a spatially complex and heterogeneous biogeophysical structure, where abundances of bare soil, vegetation, and rock vary at small scales making it difficult to map these key parameters for soil erosion assessments. A possible solution to account for the problem of spatial heterogeneity might be the usage of very high-resolution satellite imagery as high-resolution vegetation cover data is crucial for soil erosion modeling. To account for the small-scale heterogeneity of the mountain landscape high resolved multispectral imagery (QuickBird, WorldView2) and aerial (received via unmanned aerial vehicles - UAV) imagery are needed. The Coordination of Information on the Environment (CORINE) methodology - a standard method used mostly by the countries of the European Community will be used to determine the erosion risk and qualities of the lands. The CORINE method of erosion mapping analyses several factors for the determination of actual erosion risk (vegetative cover, land slope, meteorological condition and soil properties). Soil erosion risk will be assessed by integrating the CORINE model with GIS and RS. Supervised and unsupervised classifications of satellite images to erosion detection will be used combined with visual interpretation of images and vegetation indexes. Web-based RS technologies will be used to provide land cover information by using digital image processing techniques. Therefore, a combination of RS, SDI, and CORINE model will provide the potential to assess soil erosion and its spatial distribution with reasonable costs and better accuracy in larger areas. However, the spectral reflectance of land surface is very complex and many studies showed that image classification should be supported basic soil characteristics. Also the CORINE model provide only qualitative output that is hard to validate, so it should be done carefully, though the patterns of areas of high erosion risk may be confirmed by local soil loss measurements. Spatial Data Infrastructure available at CENS, FAZO and Geographic mainly addresses data retrieval and portrayal, but additional web-based geoprocessing capacities are required to analyze efficiently high-resolution remote sensing imagery. An efficient processing tool of the remote sensing imagery on a distributed computing infrastructure will be developed. For this purpose review of the various possible tools, allowing remote sensing analysis in a distributed environment will be conducted and notably the GRASS GIS tool will be explored.


Back