Southwest China, characterized by its climate sensitivity and ecological fragility, is experiencing heightened vulnerability to recurrent extreme drought due to climate change. However, not all drought events impart identical damage effects on terrestrial ecosystems. The ecosystem's response to drought becomes intricately diverse and is remain poorly understood in this region. Here, we comprehensively distinguish the inhibiting and promoting effects of cumulative drought on vegetation productivity and the modulation of hydrothermal conditions using various remote-sensing data and meteorological observations. Our results show that cumulative drought exerts pronounced inhibiting and promoting effects on vegetation productivity in 57.3% and 25.0% of vegetated area in southwest China, respectively, and shows large discrepancies in different geomorphic settings and different vegetation types. Particularly, vegetation productivity is more easily inhibited by cumulative drought in the karst landform with inadequate water-holding capacity. The croplands suffer the most inhibited effects and hardly benefit from the cumulative drought because that most of croplands are mainly distributed on the karst landform. Productivity for much of the grasslands is most strongly promoted by cumulative drought with relatively low temperature over western Sichuan and northwestern Yunnan, where with rich solar radiation can compensate energy for vegetation growth and less precipitation prevents prolonged waterlogging of plant roots owing to soil water saturation. Forests have well-developed deep root systems and can draw deep groundwater to compensate for water shortages during prolonged droughts, making them the least inhibited by cumulative drought. Savannas are the second weakly inhibited by cumulative drought, ranking below forests owing to their intermediate ecological character, straddling the boundaries between grasslands and forests. Overall, this study significantly advances our knowledge of the effects of cumulative drought on vegetation productivity and the role of hydrothermal conditions, providing valuable insights for efforts to mitigate cumulative drought risk under changing climate conditions.

Vegetation growth is adversely impacted by multiple climate extremes related to the water and thermal stress over the Tibetan Plateau (TP). However, it remains unknown at which stress level these climate extremes can trigger the abrupt shifts of vegetation response to climate extremes and result in the maximum vegetation response across TP. To fill this knowledge gap, we combined the hydrometeorological data and the satellite-derived vegetation index to detect two critical thresholds that determine the response of vegetation productivity to droughts, high-temperature extremes, and low-temperature extremes, respectively, during 2001-2018. Our results show that the response of vegetation productivity to droughts rapidly increases once crossing -1.41 +/- 0.6 standard deviation (sigma) below the normal conditions of soil moisture. When crossing -2.98 sigma +/- 0.9 sigma, vegetation productivity is maximum damaged by droughts. High-temperature extremes, which have the two thresholds of 1.34 sigma +/- 0.4 sigma and 2.31 sigma +/- 0.4 sigma over TP, are suggested to trigger the strong response of vegetation productivity at a milder stress level than low-temperature extremes (two thresholds: -1.44 sigma +/- 0.5 sigma and -2.53 sigma +/- 0.8 sigma). Moreover, we found the compounded effects of soil moisture deficit in reducing the threshold values of both high- and low-temperature extremes. Based on the derived thresholds of climate extremes that impact vegetation productivity, Earth System Models project that southwestern TP and part of the northeastern TP will become the hotspots with a high exposure risk to climate extremes by 2100. This study deciphers the high-impact extreme climates using two important thresholds across TP, which advances the understanding of the vegetation response to different climate extremes and provides a paradigm for assessing the impacts of climate extremes on regional ecosystems.