Soil moisture simulations for a sustainable irrigation management

Abstract

Accurate estimations of crop water requirements accounting for spatial heterogeneous soil properties are recognized as a major contribution towards a sustainable agricultural irrigation management. Crop-specific irrigation demand estimations may be improved by physics-based soil moisture models, although spatially distributed soil moisture simulations strongly rely on profound assessments of the model accuracy and applicability under open-field conditions. Hence, this study aims to investigate simulated root-zone soil moisture dynamics on a variably irrigated potato field provided by the HYDRUS-1D model and its suitability for irrigation management purposes in terms of input parameter requirements and applicability on larger, heterogeneous sites. All simulations were highly accurate (RMSE = 0.018 m3 m-3), when compared to in-situ measurements, but varied stronger in topsoil than in subsoil layers. A pixel-based approach using aggregated soil properties, phenological characteristics and meteorological conditions enables appropriate trade-offs between simulation accuracy and the parameterization effort and applicability in irrigation management.