Assessment of meteorological drought change in the 21st century based on CMIP6 multi-model ensemble projections over mainland China

Understanding drought dynamics spatially and temporally is essential for environmental and socio-economic systems, especially under climate change with expected increasing drought. This study aims to update the knowledge about drought spatiotemporal changes over mainland China in the 21st century by utilizing the latest Coupled Model Intercomparison Project Phase 6 (CMIP6) datasets. Precipitation from the nine CMIP6 models is first bias-corrected using a quantile mapping approach and then used to calculate the 3-month Standardized Precipitation Index (SPI) for the historical period and future three scenarios (SSP1-2.6, SSP2-4.5 and SSP5-8.5). We conduct a clustering algorithm to identify three-dimensional (i.e., latitude, longitude, and time) drought events to access the drought dynamics in both space and time. Results show that the drought events occur more frequently during June to September than the other months for both historical and future periods. The drought frequency is significantly increasing during Sep-Dec in the 21st century (increase by about 8.5%, p-value<0.05). Compared with the historical period, drought centers shift toward south under the future projections, indicating a relatively high drought frequency in south China. Additionally, drought events are projected to have shorter duration, smaller affected area, but higher intensity under all three scenarios. The study contributes to our comprehensive understanding of how future drought characteristics will change due to climate change from the perspective of spatiotemporal dynamics. These results are helpful to drought risk management and provide information for developing drought adaptation strategies.