SMART: Soil Moisture and Runoff Toolkit for Semi-distributed Hydrologic Modeling

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Hoori Ajami, University of New South Wales, School of Civil and Environmental Engineering, Sydney, NSW, Australia, Urooj Khan, Bureau of Meteorology, Melbourne, VIC, Australia, Narendra Kumar Tuteja, Bureau of Meteorology, Woodbridge, Australia and Ashish Sharma, University of New South Wales, Sydney, NSW, Australia
A new GIS based semi-distributed hydrologic modeling framework is developed for simulating runoff, evapotranspiration and soil moisture at large catchment scale. The framework is based upon the delineation of contiguous and topologically connected Hydrologic Response Units (HRUs). The HRU delineation methodology consists of delineating first order sub-basins and landforms. To reduce the number of computational elements, simulations are performed across a series of cross sections or equivalent cross sections (ECS) in each first order sub-basin using a 2-dimensional, Richards’ equation based distributed hydrological model. Delineation of ECSs is performed by weighting the topographic and physiographic properties of the part or entire first-order sub-basins and has the advantage of reducing the computational time while maintaining reasonable accuracy in simulated hydrologic fluxes. Simulated fluxes from every cross section or ECS are weighted by the respective area from which the cross sections or ECSs were formulated and then aggregated to obtain the catchment scale fluxes.

SMART pre- and post-processing scripts are written in MATLAB to automate the cross section delineation, model simulations across multiple cross sections, and post-processing of outputs for visualization. The MATLAB Parallel Processing Toolbox is used for simultaneous simulations of cross sections and is further reduced computational time. SMART pre-processing workflow consists of the following steps: 1) delineation of first order sub-basins using a digital elevation model, 2) hillslope delineation, 3) landform delineation in every first order sub-basin based on topographic and geomorphic properties of a group of sub-basins or entire catchment, 4) formulation of cross sections as well as ECSs and 5) extraction of vegetation and soil parameters using spatially distributed land cover and soil information for the 2-d distributed hydrological model. The post-processing tools generate streamflow at the outlet and spatially distributed evapotranspiration and soil moisture across the catchment. SMART modeling framework and the automation procedures improve the toolkit application for regional scale simulations and significantly reduce the model setup and computational times for large catchment scale simulations.