A Distinct Role for Aerosol and GHG Forcing in Historical CMIP6 Evapotranspiration Trends

Evapotranspiration (ET) is crucial for the global water balance, plant growth, and freshwater availability. It connects the surface water balance with surface energy fluxes, making its accurate representation vital for climate projections. However, global climate models (GCMs) struggle with ET representation due to resolution limitations and simplified depictions of soil, plant, and atmosphere interactions. Simulated future changes in ET are uncertain, and the role of driving processes remain unclear. Here, we explore the utility of a simple and interpretable method to disentangle these varying drivers. We investigate the sensitivity of JJA ET to different atmospheric variables through simple linear models predicting ET from atmospheric variables only. The model consistently yields good results across GCMs or forcing scenarios. We find that GCMs have shown strong decreases and subsequent increases in ET over the historical period, related to changes in net surface radiation. For future climate projections, decreases in water availability compete with higher available surface radiation, making future projections uncertain. Single forcing GCM realizations show that historical ET trends in densely populated regions have been more influenced by aerosol emissions than greenhouse gases. Finally, we investigate which atmospheric variables explain most short-term (year-to-year) and long-term (decadal) changes. While water availability may be the most important driver of short-term variability, for certain regions, radiation trends dominate long-term forcing. This paper leverages a simple approach to provide a comprehensive and understandable view into recent and future changes in ET, reconciling the evidence provided by more complex case studies.