Evaluation of land-atmosphere coupling and the Budyko relationship in CMIP6 models

Abstract:

Interactions between land and atmosphere are an essential determinant for extremes like drought and heatwaves, climate responses to land use change, and other potential feedbacks in a changing climate. We present a two-phase assessment of coupled land-atmosphere behavior in CMIP6 models. First, coupling indices between pairs of variables linking soil moisture to surface fluxes and surface fluxes to boundary layer properties are calculated for piControl, historical, and periods within 1pctCO2 and abrupt-4xCO2 simulations. Spatial and seasonal patterns reveal how each model performs for current climate where observational estimates of indices can be made. Multi-model consensus and uncertainty estimates of both past and future changes reveal where and when changing magnitudes of land-atmosphere coupling increase the sensitivity of weather and climate to land use management and land cover changes. Second, the Budyko relationship linking annual mean precipitation, evaporation and net radiation is examined for each model. Spatial patterns reveal hydrological climate regimes, and changes among experiments reveal how changes in terrestrial water storage and partitioning of rainfall between evaporation and runoff may evolve over time.