Trend analysis of MODIS time-series using different vegetation indices for monitoring of cropland degradation and abandonment in central Asia

Abstract

Dryland cropping systems are particularly vulnerable to land degradation processes due to climate aridity and improper management practices. Spatially explicit information on land's state, which can be derived with remote sensing techniques, should become available to support land rehabilitation decisions. This study aimed at (i) mapping a land degradation trend in the irrigated cropland in Central Asia with a linear trend analysis of the time series of the MODIS images and (ii) comparing the captured trends, based on different vegetation indices. About one-third of the study area (171,563 ha) experienced the land degradation trend during the monitoring period of 2000-2010. The trend coefficients of three vegetation indices (NDVI, EVI, SAVI) were similar with the corresponding R2-values of 0.78 (NDVI & EVI) and 0.82 (NDVI & SAVI). The revealed changes were mainly associated with the abandoned lands, where bare soil patches prevailed.