Multifractal Description of the Agricultural and Meteorological Drought Propagation Process

This study uses Multifractal Detrended Fluctuation Analysis to examine the scaling behavior and spatiotemporal variability of drought indicators. We employ two hybrid drought indicators—the GRACE Root Zone Soil Moisture Percentile (SMI) and Evaporative Demand Drought Index (EDDI)—to elucidate agricultural drought characteristics, and we quantify meteorological drought using the Standardized Precipitation Index (SPI). Our investigation was carried out in the United States in the California and Mississippi watersheds, each characterized by distinct agricultural climate conditions and planting systems. The changes in drought indicators exhibit multifractality primarily derived from the long-term memory embedded within the time series. Spatially, the increment sequence of SPI displays pronounced dependence in almost all areas. Specifically, the changes in SPI demonstrate persistence, as the Hurst exponents changed from 0.53 to 0.58 in California, revealing the possibility of a degree of trend continuation (albeit, not very strong) in the future. A regular summary cannot be obtained for Mississippi, since the Hurst exponent ranges from 0.46 to 0.51 for the increment sequence of SPI in that watershed. However, the increment sequences for the agricultural drought indicators exhibit obvious short-term persistence (anti-persistence), indicating that future drought trends may be discontinuous. The variations in the strength of anti-persistence among agricultural drought indicators can be attributed to geographical differences, diverse irrigation strategies, and varying cropping systems. No significant evolution pattern is evident for the scaling structure of drought processes due to minor variations in external controlling factors, such as rainfall and temperature, within the drought process.