A comprehensive agricultural drought stress monitoring method integrating MODIS and weather data (A case study of Iowa)

摘要

This article aims to propose a new method that detects the occurrence and severity of agricultural drought by evaluating and investigating vegetation performance of and drought stress upon croplands using remote sensing techniques combined with station based observations. It is based on the assumption that crops suffer from agricultural drought stress only when vegetation and water stresses exist at the same time. Merely using NDVI or VCI can hardly determine whether a vegetation stress is drought related, because a lower NDVI/VCI could mean that the crop land is suffering from vegetation stress due to many reasons - drought, flood, extreme temperature, wild fire, pesticides or lack of fertilizers, etc. In conclusion, this research is designed to combine visibleear-infrared/shortwave-infrared remote sensing approaches (namely, from MODIS products) for vegetation drought stress estimation, based on a pre-knowledge of the crop type per pixel, and to validate the estimated result with the observed temperature, precipitation, and soil moisture levels. The new approach is to provide more accurate result than a single vegetation index used alone.