Tuesday, 30 January 2024
Hall E (The Baltimore Convention Center)
Recent progress in climate modeling underscores the role of Soil Moisture-Latent Heat Flux (SM-LE) coupling in Land-Atmosphere (L-A) interactions. This relationship provides insights into climatic events, notably heatwaves and regime shifts. The segmented regression method by Hsu & Dirmeyer (2022) is a promising tool for forecasting these shifts in the future. However, varying outputs among numerical models, particularly regarding soil moisture, prompt a deeper probe. The hypothesis central to this paper postulates the offline (uncoupled) Land Model (LM) simulation as a potential source of these variances. A thorough examination of indicators affiliated with segmented regression facilitates a comparison between coupled and offline simulations, illuminating the implications of model decoupling. This research utilizes data derived from CMIP6 for the 110 years from 1903 to 2012 to gauge model responsiveness to precipitation and net radiation, introducing four pivotal properties: wilting point, SM-LE slope, critical SM, and saturated LE. This analysis juxtaposes CSM properties across diverse simulations by employing Lasso regression and similarity indexes.

