Soil Parameterization in Land Surface Models Drives Large Discrepancies in Soil Moisture Predictions Across Hydrologically Complex regions of the Contiguous United States

Abstract. Land surface models (LSMs) are critical components of Earth system models (ESMs), enabling simulations of energy and water fluxes essential for understanding climate systems. Soil hydraulic parameters, derived using pedotransfer functions (PTFs), are key to modeling soil-plant-water interactions but introduce uncertainties in soil moisture predictions. However, a key knowledge gap exists in understanding how specific soil hydraulic properties contribute to these uncertainties and in identifying the regions most affected by them. This study assesses the influence of soil parameter settings on soil moisture variability in the Community Land Model version 5 (CLM5) over the contiguous United States (CONUS) using Empirical Orthogonal Function (EOF) analysis. EOF analysis identified dominant spatial and temporal soil moisture patterns across multiple experimental configurations and highlighted the impact of soil parameter variability on hydrological processes. The results revealed significant discrepancies in soil moisture simulations, particularly in the central Great Plains, potentially due to the combination of arid climate conditions and limitations in modeling saturated hydraulic conductivity and soil water retention curves. Seasonal soil moisture dynamics aligned broadly with observed patterns but showed biases in magnitude and phase, emphasizing the need for refined parameterization, such as improving the representation of infiltration and drainage processes. Comparisons with ERA5-Land reanalysis data revealed improved alignment in regions with consistent climatic gradients but persistent model deficiencies in hydrologically complex areas, particularly under more arid climates such as the Great Plains where hydrological processes are notoriously harder to reproduce. This research highlights the necessity of refining soil parameter representations, utilizing high-resolution datasets, and considering climatic variability to boost the performance of LSMs. Importantly, these findings also open the door to future efforts that incorporate dynamic soil properties into LSMs. Much of this work demonstrates the dynamism of soil properties, and while this study advances modeling by revealing the importance of their inclusion, the next crucial step will be developing approaches that allow these properties to be dynamic within LSMs. This paper serves as a foundational step toward that goal, paving the way for more complex and integrated modeling frameworks that better capture soil-hydrology-climate interactions.