Comparison of three remotely sensed drought indices for assessing the impact of drought on winter wheat yield

Authors: HUANG, JIANXI - China Agricultural University; ZHUO, WEN - China Agricultural University; DONG, JINWEI - Chinese Academy Of Sciences; TIAN, LIYAN - Texas A&M University; Huang, Yanbo; ZHANG, XIAODONG - China Agricultural University

Submitted to: International Journal of Digital Earth
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/25/2018
Publication Date: 3/1/2020
Citation: Huang, J., Zhuo, W., Dong, J., Tian, L., Huang, Y., Zhang, X. 2020. Comparison of three remotely sensed drought indices for assessing the impact of drought on winter wheat yield. International Journal of Digital Earth. 13(4): 504-526.
DOI: https://doi.org/10.1080/17538947.2018.1542040

Interpretive Summary: Drought is a serious threat to crop production. For regional study and management of agricultural areas data from individual meteorological stations are difficult to be used to cover the entire region. For drought detection and analysis, a number of indices have been developed from remotely sensed data. With these drought indices, the scientists from China Agricultural University, Chinese Academy of Sciences, Texas A&M University and USDA ARS Crop Production Systems Research Unit at Stoneville, Mississippi investigated the sensitivity of these indices to agricultural drought and their effects on winter wheat yield in Shaanxi and Henan, China. The results suggested that one of the drought indices correlated well with the precipitation index to be able to better identify typical drought events. The results further indicated that drought at different growth stages had varying impacts on winter wheat yield and the drought detection accuracy is higher in rainfed area than in irrigated area. This study provides data and tools for drought detection and analysis for studies and management of regional agricultural areas for crop production.

Technical Abstract: Agricultural drought is a serious threat to food security. While meteorological observations have been widely used for drought monitoring, site-based observations are difficult to be extended to regional scale. Therefore, numerous remote sensing drought indices have been developed, such as the Vegetation Health Index (VHI), the Temperature Vegetation Dryness Index (TVDI) and the Drought Severity Index (DSI). However, the sensitivity of these indices to agricultural drought and their bio-indicating functions toward drought effects on winter wheat yield are still unclear. In this study we analyzed the performance of three typical remote sensing based drought indices (VHI, TVDI and DSI) for drought monitoring and explored the impacts of agricultural droughts on winter wheat yield in two typical agricultural production regions in Shaanxi and Henan, China, respectively. Then, these indices were correlated with the Standardized Precipitation Index (SPI) at a number of meteorological observation sites. The results suggested that DSI outperformed among the three indices in correlation with SPI, and also DSI better identified typical drought events according to the National Meteorological Drought Monitoring. Moreover, we analyzed the correlation between these three indices in different phenological stages and winter wheat yield. The results indicated that drought at different stages showed varying impacts on winter wheat yield; specifically, there were two peak correlations during the jointing and grain filling stages, and the correlation coefficient could reach 0.66, which indicated that winter wheat required more water during these phenological stages. Further we found TVDI showed better correlation with winter wheat yield at the jointing stage and the grain filling stage than VHI and DSI. The correlation between drought indices and SPI in rainfed-dominated Shaanxi province were generally stronger than that in irrigation-dominated Henan province.