Integrating MODIS-EVI and Gridded Rainfall/Temperature Fields for Assessing Land Degradation Dynamics in Horqin Sandy Lands, Inner Mongolia (China)

Horqin Sandy Lands in NE-China represent a widely discussed case of ongoing land degradation which receives much attention in China and the outside world. Located in an area of strongly continental climate and extended sandy substrates, the region is undergoing substantial land use change dynamics, largely determined by political decisions and regulations issued by the central and regional government. A simple analysis of trends derived from existing satellite time series is felt not fully adequate for characterizing the changes in biological productivity during the past decades. As climate data suggest for some stations an increase in annual precipitation during the past 10 years it appears mandatory to relate satellite measurements of biological productivity to corresponding changes in plant-available water over time.

In the course of EU-funded research projects (LADAMER; DeSurvey) an approach based on ecological responses of vegetation to physical drivers (e.g. 2dRUE ) has been developed to more adequately characterize areas affected by desertification and transitional land use change processes. The concept relates to earlier work (Le Houérou 1984; Prince et al., 1998) and departs from the assumption that Rain Use Efficiency, i.e. the ratio of net primary productivity to precipitation, decreases with land degradation. The approach is based on combining satellite-derived indicators of biological productivity (e.g. EVI, FAPAR) and spatially interpolated climate data, and it is climatically de-trended for aridity.

The concept is currently applied to different case studies of the DeSurvey Project, one of which covers the Horqin Sandy Lands in Inner Mongolia (China). Results for this site are produced by integrating gridded climate station data and EVI time series acquired by MODIS for the time span 2001-2002, and will be compared to a calibrated 20-year-time-series of Landsat-TM/ETM+ data (1987-2007).