Crop Drought Area Extraction Based On Remote Sensing Time Series Spatial-Temporal Fusion Vegetation Index

Shufu Liu,Jingguo Tian,Shudong Wang,Dacheng Wang,Tianhe Chi,Ying Zhang

DOI: https://doi.org/10.1109/IGARSS.2019.8899853

2019-01-01

Abstract:Agricultural drought given the significant impacts on social economy and human living, the research on agricultural drought monitoring has been paid broad attention by many scholars. Development of remote sensing technology provides a potential opportunity for effective and real-time agricultural drought monitoring. Vegetation index, as one of the most important parameters to characterize vegetation growth and monitor vegetation change, can effectively monitor and evaluate the crop drought status. High resolution vegetation index time series can not only monitor the feature of changes of affected time, but also can extract the area of crop drought disaster accurately and exactly. However, due to cloud pollution, satellite revisit cycle, sensor design limitations and other factors, it is difficult for a single remote sensing sensor to simultaneously acquire high spatial-temporal images. So, it is urgent to get high spatial and temporal data which can be built by fusion algorithm. In this paper, the time series curve of vegetation index was constructed, and the drought area of crops was extracted by using the time series curve characteristics of maize vegetation index in drought-affected and non-drought-affected years. The overall classification accuracy was 90.03%.