Effects of Explicit Convection on Land Surface Air Temperature and Land‐Atmosphere Coupling in the Thermal Feedback Pathway

Abstract

Abstract Simulating and understanding continental temperature extremes is a critical issue in Earth System Modeling. Conventional general circulation models are impaired by imperfect cloud and boundary layer parameterization schemes with implications for potentially unrealistic distributions of atmospheric variables and land‐atmosphere coupling signals. In this study we examine a modern version of Superparameterization (SP) in the Community Atmosphere Model v.5 to examine impacts of SP on the characteristics of land surface air temperature, thermal land‐atmosphere coupling, and the relationship between them. The results show that SP in the Community Atmosphere Model (SPCAM) improves mean land surface air temperature at daily time scales regionally compared to Community Atmosphere Model 5 and better simulates thermal land‐atmosphere coupling (soil moisture‐surface air temperature coupling) in several well‐known hot spots (e.g., the Great Plains, Sahel, and India). Detailed analysis of regional probability distribution functions reveals how the intrinsic relationships of atmosphere variables, land‐atmosphere coupling, and temperature extremes are modified by SP. Regression‐type metrics are also used to examine global land‐atmosphere coupling, with some expected limitations where multiple coupling regimes coexist temporally. Stratifying results by soil moisture percentile proves illuminating in this regard and reveals that SP frequently amplifies land‐atmosphere thermal coupling at the dry end of regional coupling regimes.