Climatic driving mechanisms of the propagation from meteorological drought to agricultural and ecological droughts.

Droughts significantly impact terrestrial vegetation ecosystems. Understanding the mechanisms by which drought affects ecosystems under different hydrogeological conditions is crucial for ecosystem protection. The aim of this study was to investigate the characteristics and mechanisms of propagation from meteorological drought (MD) to agricultural drought (AD) and ecological drought (ED) in the Jinsha River Basin from 2000 to 2014. The monthly standardized precipitation evapotranspiration index (SPEI), soil moisture index (SSMI), normalized difference vegetation index (SNDVI), and solar-induced chlorophyll fluorescence (SSIF) data were used to investigate the responses of AD and ED to MD. On the basis of the maximum correlation coefficients (MCCs), the differences in the drought propagation times of MD to AD and ED were explored in positively and negatively correlated areas. A random forest algorithm was used to identify the impacts of climatic factors driving drought propagation. The results revealed that AD was mainly positively correlated with MD, whereas the correlation coefficients between ED and MD ranged from negative to positive. The propagation time from MD to AD was relatively short in summer and autumn. In positively correlated areas, the propagation time from MD to ecological drought indicated by NDVI (EDndvi) was longer than that indicated by SIF (EDsif), and the opposite was true in negatively correlated areas. The random forest algorithm results indicated that temperature (T), solar radiation (S) and precipitation (P) were key factors influencing ED in positively correlated areas and that T was an important factor in controlling the occurrence of ED in negatively correlated areas. Solar-induced chlorophyll fluorescence (SIF) was more sensitive to MD and had a shorter response time in positively correlated areas, suggesting its potential for monitoring vegetation growth responses to drought. We found that MD was not the main factor influencing vegetation growth in negatively correlated areas. The findings of this study had significant implications for understanding the mechanisms of the response of vegetation growth to MD and offered scientific guidance for maintaining terrestrial ecosystem health.