Further improvement of the approach to monitoring drought using vegetation and temperature condition indexes from multi-years' remotely sensed data

This study is focused on developing appropriate methods for determining the "warm edge" and "cold edge" of vegetation temperature condition index(VTCI) drought monitoring approach. By integrating AVHRR's land surface temperature(LST) and normalized difference vegetation index(NDVI) products, the edges' determining methods were studied by using the multi-years' composited LST and NDVI products in the given period of the first ten days of May from 1999 to 2003. The results show the "warm edge" can be determined by using multi-years' maximum value composited LST and NDVI products in the given period. While, the "cold edge" can be determined by applying the combinations with the multi-years' maximum value composited NDVI product and a minimum composited LST product. The minimum LST product was acquired by using the minimum value compositing technique to process each year's maximum value composited LST products. Ground-measured precipitation data in Guanzhong Plain, Shaanxi Province of China were employed to validate the edges' determining methods and VTCI approach. A linear correlation analysis was applied to study the correlation between precipitation and VTCI. There are positive correlations between precipitation and VTCI, the correlation coefficients between VTCI and cumulative precipitation in one or two periods of ten days interval are the highest. These results indicate that the edges' determining methods are feasible, and VTCI is a close real-time drought monitoring approach.