Fig. 6: Spatial distribution of changes in land-atmospheric coupling and feedback under global warming levels.

Multi-model mean of regression coefficients for the period 2014–2100 and for shared socioeconomic pathway (SSP) 2–4.5 and SSP5–8.5 scenarios between global mean temperature anomalies relative to 1985–2014 and; a land-atmospheric coupling computed as the correlation between 90th percentile summer (June, July, August) daily maximum 2 m air temperatures and summer mean soil moisture; b land-atmospheric feedback computed as the linear regression between 90th percentile summer daily maximum 2 m air temperatures and summer mean soil moisture. In the non-dashed regions at least three out of the five models agree on the regression coefficient sign, and at least three out of the five models show significant changes at the 95% confidence level.