Experimental and numerical modeling of waves routing on permeable soils

A study on waves routing on permeable, rough boundaries may be of great interest for several purposes, e.g. the optimal use of water in agriculture, the analysis of the interaction between surface and underground water resources, the optimization of innovative solutions for urban runoff mitigation and treatment. The analysis was organized in two phases; a first experimental study on infiltration was completed by the complete numerical modeling of waves routing on a permeable soil. The experimental setup was located in the Laboratory Giorgio Bidone of Polytechnic University of Turin (DIATI). A horizontal, prismatic (width 0.30m, height 0.20 m), 9m long channel was filled with a dry sandy layer of constant thickness (0.10m). A mattress made of polyurethane foam laid on the sandy layer in order to model vegetation or surface roughness. Lifting a gate, we allowed the water stored in the upstream reservoir to flow into the channel. Water flowed through the mattress while infiltrating into the sand (soil). Several combinations of three sand particle size distributions, three mattress types, and three inflow hydrographs allowed a study on the main parameters leading the propagation of waves on permeable soils. Each test was video-recorded, the detection of the profiles of the surface waves propagating through the mattress allowed to focus only on the infiltration process. Laboratory tests pointed out the hydraulic parameters of the three sands used. The implementation of appropriate numerical codes (one original and one commercial - Hydrus 2D, Simunek et al, 2002) led to semi-quantitative, propaedeutic, preliminary conclusions. The complete analysis of the propagation of waves on permeable soils was then achieved through the numerical modeling of a specific case study, i.e. the biofilter installed at Monash University in Melbourne. A biofilter is a vegetated, permeable retention basin useful for urban runoff treatment. Onsite recorded data of flow rates and soil moisture values allowed the calibration and verification of a numerical model resulting from the coupling of original numerical codes with one open source software (WinSRFR, Strelkoff et al, 1990) and a revised version of a commercial software (Hydrus 2D; Simunek et al, 2002). The resulting numerical model is useful for both diagnostic and planning purposes; it allows the definition of the parameters (geometry, soil type, vegetation type) that may enhance hydrologic and pollutant removal performances of the facility according to local climate conditions (frequency, intensity, volume of precipitation, temperature).