Global vegetation and associated ecosystem services critically depend on soil moisture availability which has decreased in many regions during the last three decades. While spatial patterns of vegetation sensitivity to global soil water have been recently investigated, long-term changes in vegetation sensitivity to soil water availability are still unclear. Here we assess global vegetation sensitivity to soil moisture during 1982-2017 by applying explainable machine learning with observation-based leaf area index (LAI) and hydro-climate anomaly data. We show that LAI sensitivity to soil moisture significantly increases in many semi-arid and arid regions. LAI sensitivity trends are associated with multiple hydro-climate and ecological variables, and strongest increasing trends occur in the most water-sensitive regions which additionally experience declining precipitation. State-of-the-art land surface models do not reproduce this increasing sensitivity as they misrepresent water-sensitive regions and sensitivity strength. Our sensitivity results imply an increasing ecosystem vulnerability to water availability which can lead to exacerbated reductions in vegetation carbon uptake under future intensified drought, consequently amplifying climate change.

 Vertically resolved soil moisture improves the understanding of large-scale vegetation productivity. 11  Extended water-related control on vegetation productivity emerges when considering multi-layer soil moisture versus total soil moisture.  Sub-surface soil moisture is particularly important for vegetation productivity in semiarid regions.

Abstract. Vegetation plays a vital role in the Earth system by sequestering carbon, producing food and oxygen, and providing evaporative cooling. Vegetation productivity extremes have multi-faceted implications, for example on crop yields or the atmospheric CO2 concentration. Here, we focus on productivity extremes as possible impacts of coinciding, potentially extreme hydrometeorological anomalies. Using monthly global satellite-based Sun-induced chlorophyll fluorescence data as a proxy for vegetation productivity from 2007–2015, we show that vegetation productivity extremes are related to hydrometeorological hazards as characterized through ERA5-Land reanalysis data in approximately 50 % of our global study area. For the latter, we are considering sufficiently vegetated and cloud-free regions, and we refer to hydrometeorological hazards as water- or energy-related extremes inducing productivity extremes. The relevance of the different hazard types varies in space; temperature-related hazards dominate at higher latitudes with cold spells contributing to productivity minima and heat waves supporting productivity maxima, while water-related hazards are relevant in the (sub-)tropics with droughts being associated with productivity minima and wet spells with the maxima. Alongside single hazards compound events such as joint droughts and heat waves or joint wet and cold spells also play a role, particularly in dry and hot regions. Further, we detect regions where energy control transitions to water control between maxima and minima of vegetation productivity. Therefore, these areas represent hotspots of land–atmosphere coupling where vegetation efficiently translates soil moisture dynamics into surface fluxes such that the land affects near-surface weather. Overall, our results contribute to pinpointing how potential future changes in temperature and precipitation could propagate to shifting vegetation productivity extremes and related ecosystem services.

The response of vegetation physiology to drought at large spatial scales is poorly understood due to a lack of direct observations. Here, we study vegetation drought responses related to photosynthesis, evaporation, and vegetation water content using remotely sensed data, and we isolate physiological responses using a machine learning technique. We find that vegetation functional decreases are largely driven by the downregulation of vegetation physiology such as stomatal conductance and light use efficiency, with the strongest downregulation in water-limited regions. Vegetation physiological decreases in wet regions also result in a discrepancy between functional and structural changes under severe drought. We find similar patterns of physiological drought response using simulations from a soil–plant–atmosphere continuum model coupled with a radiative transfer model. Observation-derived vegetation physiological responses to drought across space are mainly controlled by aridity and additionally modulated by abnormal hydro-meteorological conditions and vegetation types. Hence, isolating and quantifying vegetation physiological responses to drought enables a better understanding of ecosystem biogeochemical and biophysical feedback in modulating climate change.

Rising anthropogenic greenhouse gas concentrations and the consequent global warming have increased the intensity and duration of drought in many regions (Canadell et al., 2021;Dai et al., 2018). Drought affects a variety of environmental and socio-economical sectors, and has been identified as one of the most impactful natural hazards (