Fig. 3: Relationship between surface soil moisture (SSM) characteristics (variability and regime) and probabilities of strong causality of SSM–precipitation (P) and its sub-processes in boreal warm seasons based on reanalysis.
From: Causal pathways underlying global soil moisture–precipitation coupling
a Spatial distribution of SSM variability (m3 m−3) over the period of 1979–2021. The magnitude of SSM variability is defined as the standard deviation of daily SSM. The royal blue lines denote the 2500-m isoline of terrain height. b The probability (%) of significant SSM→P causality (p < 0.1) in each bin of SSM variability and SSM regime across the globe. The number of ERA5 grid cells (Ni,j) in the bin with the ith SSM variability range (i = 1, 2, …, 6) and the jth SSM regime range (j = 1, 2, …, 6) is calculated, and then these grid cells are classified into significant and non-significant parts according to the SSM→P causality. Finally, the proportion of the grid cells (ni,j) with significant SSM→P causality in the total grid cells (Ni,j) is calculated, which is the probability (ni,j/Ni,j). c, d Same as (b), but for evapotranspiration (ET)→P and sensible heat (SH)→P causality. e The probability (%) of significant SSM→P, ET→P and SH→P causality in different ranges of SSM variability.
