The relationship between precipitation anomalies and satellite-derived vegetation activity in Central Asia

In Central Asia, water is a particularly scarce and valuable good. In many ecosystems of this region, the vegetation development during the growing season is dependent on water provided by rainfall. With climate change, alterations of the seasonal distribution of precipitation patterns and a higher frequency of extreme events are expected. Vegetation dynamics are likely to respond to these changes and thus ecosystem services will be affected. However, there is still a lack in understanding the response of vegetation to precipitation anomalies, especially for dryland regions such as Central Asia. This study aims to contribute to an improved understanding of vegetation sensitivity to precipitation anomalies and corresponding temporal reaction patterns at regional scale. The presented analyses are based on time-series of Normalized Difference Vegetation Index (NDVI) and gridded precipitation datasets (GPCC Full Data Reanalysis) for the years 1982-2006. Time-series correlation analyses show that vegetation development is sensitive to precipitation anomalies for nearly 80% of the Central Asian land surface. Results indicate a particularly strong sensitivity of vegetation in areas with 100-400 mm of annual rainfall. Temporal rainfall-NDVI response patterns show a temporal lag between precipitation anomalies and vegetation activity of 1-3 months. The reaction of vegetation was found to be strongest for precipitation anomalies integrated over periods of 2-4 months. The observed delayed response of vegetation to precipitation anomalies reveals potential for drought prediction in Central Asia. The spatial patterns of vegetation reactions are discussed with focus on the role of precipitation amount and seasonality, land use and land cover. (C) 2012 Elsevier B.V. All rights reserved.