Spatiotemporal drought forecasting in Xinjiang’s irrigated agriculture: Model comparison and multi-source data integration

Irrigated agricultural lands are essential for global food production, particularly in arid and semi-arid regions. As climate change intensifies pressure on water resources, assessing their drought risks is increasingly important. This study introduces a novel methodology that integrates remote sensing and meteorological observation networks with non-spatial, spatial, and spatiotemporal models to monitor and forecast irrigated agricultural drought in Xinjiang, China. Vegetation Optical Depth (VOD) was used as a drought indicator where its relationship to meteorological, environmental, and anthropogenic factors during spring, summer, and autumn from 2000 to 2019 was assessed. Three models, Multiple Linear Regression (MLR), Geographically Weighted Regression (GWR), and Geographically and Temporally Weighted Regression (GTWR) were evaluated for their predictive accuracy. Our result revealed a strong positive correlation between VOD and precipitation, particularly in spring (r = 0.79, p < 0.001) and autumn (r = 0.79, p < 0.001), with irrigation becoming more significant in spring (r =0.53, p < 0.001) and summer (r = 0.51, p < 0.001). The GTWR model outperformed MLR and GWR for the insample analysis (2000–2016), as evidenced by the higher R2 values (GTWR > GWR > MLR) and the lower AICc values (GTWR < GWR < MLR). Findings highlight the combined impact of meteorological and irrigation factors on vegetation health. Out-of-sample prediction for 2017 to 2019, confirmed GTWR as the most accurate model, which was subsequently used for forecasting VOD under different Shared Socioeconomic Pathways (SSP) scenarios (SSP2-4.5, SSP5-3.4-OS, SSP5-8.5) for 2020 to 2050. Forecasts indicated lower VOD values under SSP2-4.5 compared to other scenarios before 2040, but these values surpassed others after 2040. Forecasts for central and eastern Xinjiang indicated generally lower VOD compared to historical values. This study offers insights for monitoring drought and managing irrigation in regions with similar climates globally