Enhanced Land‐Surface Processes Sustain Evapotranspiration and Improve Tropical Climate Variability in a Coupled Earth System Model

Land surface hydrological processes influence climate predictability through soil moisture persistence and coupling efficiency, the capacity of surface flux anomalies to generate spatially coherent atmospheric forcing at large scales. We examine coupling efficiency using century‐long free‐running simulations of a fully coupled Earth system model, comparing two configurations that differ only in land surface parameterizations. The Noah‐MP configuration differs from the Noah 2.7.1 control in subsurface hydrology, vegetation physiology, and surface‐layer exchange, and produces larger‐scale and more coherent atmospheric responses despite nearly identical soil moisture memory. Because every other model component is shared, the contrast isolates the effect of the land surface formulation. The Amazon–Pacific teleconnection is absent in the control but is reproduced closely with Noah‐MP. Drought composites with nearly equal rainfall show higher latent heat flux over tropical evergreen forests with accessible subsurface water with Noah‐MP than control. Over the Amazon, sustained evapotranspiration anomalies are accompanied by a Matsuno–Gill type response and a reduction in the Pacific Walker Circulation bias. The El Niño–Southern Oscillation (ENSO) spatial structure is also more realistic, with a stronger east–west sea surface temperature gradient. Atlantic Niño variability becomes less tied to ENSO, closer to the observed relationship. These results suggest that coupling efficiency, rather than soil moisture persistence alone, governs how strongly land surface processes influence large‐scale variability, and that a more realistic representation of land surface processes can strengthen that influence with potential implications for long‐term predictability.