Quantitative identification of drought dominant periods and driving factors in China: integrating from TVDI and pixel-wise EMD

Under global change, drought disasters have become increasingly frequent in China, but current research lacks the quantitative driving mechanism of drought by dominant driving factors at multiply spatio-temporal scales. Therefore, this research portrays the spatial pattern of drought in China from 2000 to 2022 based on the temperature-vegetation drought index (TVDI), combined with the Mann–Kendall test and Hurst trend analysis and we introduced the pixel-wise empirical modal decomposition (EMD) method to quantitatively identify the dominant periods and driving factors of drought in China. Integrating the TVDI with a pixel-wise EMD approach reveals multi-scale drought characteristics across China and quantitatively identifies the dominant drivers of drought over multi-temporal scales. The results indicate that precipitation (PRE) is the primary driver of seasonal drought. NDVI shows high sensitivity to drought in ecologically vulnerable regions, while potential evapotranspiration dominates interannual drought dynamics in the arid northwest. Moreover, maximum temperature (Tmax) significantly drives interdecadal drought patterns in northern China. Additionally, as the dominant drought period lengthens, the influence of PRE gradually diminishes, whereas the role of Tmax becomes increasingly prominent. The pixel-wise EMD method can provide a more accurate scientific basis for quantifying multi-scale drought risk.