TOWARDS DEVELOPING A WATERSHED ASSESSMENT FRAMEWORK USING GEOMORPHIC-BASED HYDROLOGIC AND HYDRAULIC MODELING

Urban hydromodification has created a need for an urban watershed assessment framework involving environmental rehabilitation practices to prioritize long-term and cost-effective enhancement strategies. This study sought to develop a watershed-scale model using Personal Computer Storm Water Management Model (PCSWMM) and field reconnaissance to be included in the planning phase of future watershed assessment frameworks. Typical modeling efforts for urban watershed stormwater management plans do not consider fluvial geomorphology during model development. To create an effective stormwater management and stream restoration watershed plan, this study sought to determine the level of model development necessary to prioritize watershed enhancement activities and explore the correlation among stream field data, desktop data, and modeled stream power and bed shear stress. A case study was completed on the Turkey Creek watershed in Knox County, Tennessee. Within the study scope, the effectiveness of geomorphic node placement and inclusion of stormwater control measures (SCMs) into models were assessed. A watershed assessment plan was developed including rapid geomorphic assessments (RGAs), a hydraulic grade control (HGC) inventory, geomorphic-based analysis where reaches were determined, three watershed-scale models with increasing detail, and stream power and bed shear stress metrics calculated from model outputs. W_30,50,80 [percentile stream power] and t_30,50,80 [percentile bed shear stress] were calculated from model outputs and statistically analyzed. The three models were: (I) PCSWMM automated conveyance network, (II) developed conveyance network of geomorphic-based nodes and links, and (III) Model II with SCMs. The most detailed model, Model III, outperformed other simulations indicating that watershed models should incorporate geomorphic node placement, SCMs, desktop data, field verification, and manual adjustment of automatically generated features. Single and stepwise linear multiple regression analyses yielded varying results, but field (RGA score and sub-metrics) and desktop (drainage area, channel slope, land uses, distance between HGCs) data were determined to be predictors of W_30,50,80 and t_30,50,80. RGA sub-metrics, land use data, and distance to HGCs were frequently included in exclusive multiple regressions, suggesting that field and desktop data should be incorporated during model development. Building a geomorphic-based watershed planning model provides stormwater managers key information to locate areas of instability and test rehabilitation techniques.