Fig. 4: Process-level evaluation of boreal warm-season surface soil moisture–precipitation (SSM–P) causality based on CMIP6 simulations and reanalysis.
From: Causal pathways underlying global soil moisture–precipitation coupling
a ERA5-estimated and model-simulated proportions (%) of grid cells with significant SSM→P causality (p < 0.1) over the eight SSM–P coupling hotspots. b Comparison of the proportions of grid cells with significant SSM→evapotranspiration (ET) causality simulated by ERA5-similar models (AWI-ESM-1-1-LR, MPI-ESM1-2-LR, MPI-ESM1-2-HR and MPI-ESM-1-2-HAM) and ERA5-dissimilar models (BCC-CSM2-MR, NorESM2-LM, MRI-ESM2-0 and BCC-ESM1) with those estimated by ERA5. c–e Same as (b), but for SSM→sensible heat (SH), ET→P and SH→P causality, respectively. f Comparison of the relationship between SSM variability and probability of significant SSM→P causality simulated by ERA5-similar models with that derived by ERA5. g Same as (f), but for ERA5-dissimilar models.
