From near surface to root zone soil water losses: a new model validated with field TDR and remotely sensed data

Understanding soil moisture drydowns, especially within the root zone, is critical for predicting water availability and crop stress. Remote sensing provides continuous observations of surface soil moisture, but their practical value for agricultural application depends on linking surface dynamics to root zone layers. With this aim, we developed the muSEC model, based on the recently proposed surface evaporative capacitor (SEC). We compared it with three models of increasing complexity: the Laio bucket scheme, a desorptivity-based model, and a Richards solver. The analysis combined multiyear in-situ TDR measurements (depth-averaged soil moisture at 0–15, 0–30, and 0–60 cm) from a sandy grassland with NASA SMAP retrievals from a loamy maize field, thus covering both accurate local root zone monitoring and satellite-scale surface observations, while testing the models across two contrasting soils. Across both sites, all models reproduced drydowns with good accuracy, with a root mean square error consistently lower than a benchmark of 0.06 . Performance systematically improved with depth compared to the surface, in line with expectations for tilled agricultural soils. Calibration on 0–15 cm and validation on 0–60 cm with the in-situ data showed that, under the relatively homogeneous soil conditions considered here, parameters calibrated on surface layers can predict root zone dynamics, an important result for agricultural applications. While model complexity did not automatically translate into superior performance at a single depth, the proposed muSEC showed the best depth transferability among the simplified models, combining structural flexibility with a substantially lighter numerical framework than fully physically based approaches. These results support its potential use for integrating satellite surface soil moisture into agricultural water management, while highlighting the need for further assessment in more heterogeneous soil profiles. • A new multilayer model (muSEC) based on the surface evaporative capacitor (SEC). • Links surface and root zone soil moisture drydown dynamics. • Integrates satellite soil moisture with minimal physical parameters. • Hourly simulations capture daily drydowns across two contrasting soils. • Calibration shows transferability (RMSE 0.06 cm 3 cm -3 ) and physical realism.