The assessment of the effectiveness of the unimodal and bimodal models to evaluate the water flow through nature-based solutions substrates

Nature-based solutions (NBS) are popular techniques for managing stormwater. Most NBS allow porous media as their main layer. The description of the soil water retention curve (SWRC) as the unsaturated hydraulic conductivity curve (UHCC) is often required to run the hydrological simulations with the physically based models. Using the unimodal and bimodal models to assess the SWRC and UHCC of soils is a widespread technique but their evaluation is often present in literature only in terms of curve fitting. Based on these assumptions, this work presents the performance assessment of the van Genuchten unimodal and bimodal models by functional evaluation of them based on the runoff from several NBS substrates. Four soil substrates were investigated to define the structure, the SWRC and the UHCC. Results showed that all substrates had a bimodal behaviour with the lowest values of RMSE (RMSE_T = 0.0023/0.0037, RMSE_K = 0.0636/0.1284). Finally, a numerical simulation using the HYDRUS-1D model was performed for a 3-month data set to check the effectiveness of the unimodal model instead of the bimodal one. The findings have shown that the unimodal model must be preferred instead of the bimodal because it has fewer parameters and assured low discrepancies in runoff volume (e = 0.00%/6.25%).