Reconstructing vegetation temperature condition index based on the Savitzky-Golay filter

Abstract

Vegetation temperature condition index (VTCI) is a near-real-time drought monitoring approach which is derived from normalized difference vegetation index (NDVI) changes in a given region to land surface temperature (LST) changes of pixels with a given NDVI value. It can be physically explained as the ratio of temperature differences among the pixels which have the same NDVI values. Due to the noise in the NDVI and LST, results of VTCI have much deviation. The Savitzky.Golay filter, a weighted moving average filter as a polynomial of a certain degree, is applied to smooth out noise in NDVI and LST time-series. VTCI were taken into Guanzhong Plain of Shaanxi Province of each 10-days from March to May in 2007 and 2008 as the study data. In order to reconstruct VTCI space-time (temporal and spatial) series data, the Savitzky.Golay filter was used to reconstruct the VTCI time-series of each pixel in remote sensing images. Then, the results were extended to the surface from the point. The results show that the Savitzky.Golay filter could improve the quality of VTCI and could get a better drought monitoring result. © 2011 IFIP International Federation for Information Processing.