The global distribution of hydrological regimes determines land-atmosphere feedbacks by regulating the water, energy and carbon fluxes. Strong land-atmosphere feedbacks are observed over the drier surface where evapotranspiration and partitioning of energy fluxes are regulated by soil moisture availability. In this study, we propose an advantageous method compared to conventional methods—soil moisture (from Soil Moisture Active Passive (SMAP) and ERA5) and soil diurnal temperature range (from ERA5 and GLDAS) covariation—to estimate the critical soil moisture ($${\theta }^{\ast }$$), a threshold that differentiates the water-limited regime (WLR) and energy-limited regime (ELR). Results show SMAP and ERA5-based $${\theta }^{\ast }$$ have consistent spatial patterns and distribution across various biomes and climate zones. High $${\theta }^{\ast }$$ (0.25–0.3 m³m³) are noted over tropical regions and low $${\theta }^{\ast }$$ (0.15–0.2 m³m³) are observed over semi-arid and dry sub-humid (savanna, shrubland and grassland) regions, which indicate relatively weaker and stronger land feedbacks and persistence of WLR, respectively. Additionally, we find that $${\theta }^{\ast }$$ variability is strongly dependent on mean soil moisture state, aridity, vapor pressure deficit and bioclimatic conditions. A seasonal analysis indicates that landscapes largely stay in WLR during winter and spring, in ELR during summer and in transitional regime during autumn, especially over the western USA, Brazilian savanna, Sahelian grassland, African shrublands, peninsular India, and southern Australia. Although such patterns are prominently observed in ERA5 regimes, SMAP reports the prevalence of dynamic regime during summer. While seasonally dominant regimes provide the expectation of dry extremes development, the probability of dynamic regime shows the gross expectation of WLR and ELR, and changing land-atmosphere coupling strength.