Spatiotemporal patterns of agricultural and meteorological droughts using SPI and MODIS-based estimates over a Brazilian semiarid region: study case of Upper Paraíba River basin

Abstract

Drought is a very complex natural phenomenon that affects the water supply and consequently impacts agricultural and socioeconomic activities, causing social and political crises. The objective of this study is to spatiotemporally analyze meteorological and agricultural droughts based on multiple indices in the Upper Paraíba River basin for the 2001–2017 period. The agricultural droughts were analyzed using four drought indices (NDVI, LST, VSWI, and VHI) based on data from the MODIS sensor on the TERRA and AQUA satellites, while the meteorological droughts were characterized using the standardized precipitation index (SPI) at three different time scales based on TRMM rainfall data. The values of LST, NDVI, VHI, VSWI, and SPI revealed how much vegetation is influenced by rainfall variability in this semiarid region of Brazil. The estimated results of VSWI and VHI show the years with extremely wet and dry conditions from spatial and temporal perspectives. The relationship between estimated drought indices and observed relative humidity showed that VSWI (R2 = 0.78) was best fitted when compared to VHI (R2 = 0.46). Thus, it can be concluded that the VSWI index is better adapted because it presented more consistent results for the behavior of agricultural droughts when compared to the VHI. The results of the SPI analysis were used for the identification of the main drought events that affected the study basin. This combined technique of different drought indices easily produced an alternative method for comprehensive drought assessment.