Equivalent Cross Sections - A computationally efficient alternative for distributed hydrological modelling

One of the ongoing issues with adopting distributed hydrological models for operational hydrological simulation is the significant computational effort and additional data need that is involved. While data has become easier to come by through higher resolution topographic and physiographic information, the issue of computational effort remains intact. This study presents a new basis for distributed hydrological modeling aimed at significantly reducing the computational effort that is involved. The approach involves the formulation of a handful of Equivalent Cross Sections (ECSs) for each first order sub-basin in a catchment. These ECSs are formulated on the basis of different weighting of topographic and physiographic variables on four topologically connected landforms. These landforms are delineated on the basis of similarities in topographic and geomorphologic attributes across the catchment. As modeling is performed over each landform for each ECS, the number of computational units used are significantly lower than the number of pixels required in a fully distributed modeling approach. The study presents the rationale for defining these ECSs and topologically connected landforms in any first order sub-basin, and demonstrates the typical accuracies and reductions in run times that are attained through applications to selected first order basins in the Snowy-Monaro region of eastern New South Wales, Australia. It then goes on to assess model performance over the 313.8 km2 McLaughlin catchment located in south-eastern New South Wales (NSW), Australia, that consists of 822 first order sub-basins. Model simulations are based on the explicit solution of the 2-dimensional Richards' equation over each cross-section, along with a use of a simplistic hydraulic routing approach for the catchment scale simulation. Results are compared to observed runoffs, and also to the simulated runoff from a lumped conceptual hydrologic model specified over the catchment. The spatial patterns of simulated evapotranspiration and soil moisture are also analyzed and found to be consistent with the input forcing. The impact of different climate and land cover on soil moisture is investigated at various soil depths and four landforms. Our results show the mean soil moisture in the surface soil layer is lower than the deeper soil layers but its variability is higher for all climates and land use types. Within landforms, the mean soil moisture in shallow layers has increased as we move from landform-upslope to landform-alluvial-flats, but for deeper soil layers the mean soil moisture is almost constant across all landforms. Climate, land use type and soil hydraulic properties play the key role in controlling soil moisture dynamics.