Impact of soil heterogeneity on evapotranspiration and irrigation modeling using POLARIS and SSURGO-based soil hydraulic properties

Accurate maps of soil hydraulic properties are needed to enable precise, site-specific modeling of crop water use. Most available maps (e.g., the USDA-NRCS Soil Survey Geographic Database (SSURGO)) have significant uncertainties that may limit their usefulness for implementing site-specific irrigation (I_rri). Recently, a probabilistic, high-resolution (30 m) soil property map of the contiguous United States (POLARIS) was developed. In this study, we compare the use of POLARIS and SSURGO for parameterizing a water and energy-balance model of simulated I_rri and simulated evapotranspiration (ET_s). The Backwards Averaged Iterative Two Source Energy Balance Solution (BAITSSS) model was applied to a major irrigated agricultural region (Yuma Valley of Arizona and the adjacent Bard region of California, ∼ 110 km²). The BAITSSS simulation covered a lettuce growing cycle between September and December 2019 (78 days) utilizing Landsat-based vegetation indices, North American Land Data Assimilation System (NLDAS) weather variables, and Parameter-elevation Relationships on Independent Slopes Model (PRISM) precipitation data. SSURGO-based soil hydraulic properties resulted in approximately a 7 % increase in field-averaged simulated seasonal ET_s and a 25 % increase in field-averaged seasonal simulated I_rri. The POLARIS-based simulations had higher I_rri efficiency and less drawdown of soil moisture prior to harvest. Overall, this study highlights uncertainties introduced by soil parameterization in ET_s and I_rri modeling and their role in agricultural water management.