Assessment of the CLASIC urban hydrology model, in the Spring Creek Watershed, northern Colorado
Date
2021
Authors
Mohammad Zadeh, Mahshid, author
Arabi, Mazdak, advisor
Bhaskar, Aditi, committee member
Ronayne, Michael, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Urban development influences the quantity and quality of water at local to watershed scales. Urban hydrology models are commonly used to plan, design, and implement stormwater infrastructure systems to minimize water quality and flooding consequences of urban development. However, the applicability of existing models at municipal scales is hampered by extensive data and computational requirements. The Community-enabled Life-cycle Analysis of Stormwater Infrastructure Costs (CLASIC) tool is a cloud-computing web application that facilitates the simulation of hydrological and water quality responses at municipal scales. The tool also provides modules to assess the lifecycle costs of green stormwater infrastructure systems. CLASIC is a modified version of the EPA's SWMM model with direct linkages to disperse land use, climate, soils, and other data resources. This study aims to assess the performance validity of the CLASIC tool for the characterization of urban hydrologic processes and responses. Specifically, the objectives of the study are to: i) evaluate the performance of the model compared to the SWMM model and observed stream discharge at various spatial and temporal scales; and ii) identify the most influential model parameters to inform model parameterization. The study is conducted in the Spring Creek catchment within the Cache la Poudre River watershed in Colorado. Streamflow in Spring Creek is influenced by urban activities in the City of Fort Collins. Model evaluation is conducted at hourly, daily, and monthly time steps at two USGS gaging stations along the stream. Comparison of observed and simulated flow duration curves along with several goodness-of-fit measures, including Nash-Sutcliff coefficient of efficiency and percent bias are used to evaluate the model performance. The Sobol' Global Sensitivity Analysis method is used to assess the importance of model parameters for different model responses, including mean and peak stream discharge. The first and total order sensitivity indices are computed to evaluate the effects of parameters individually and in combination. Overall, hydrological budgets are simulated similarly between CLASIC and SWMM. The results indicate the performance validity of CLASIC stream discharge simulations at computational time steps greater than the time of concentration of the catchment. However, SWMM peak discharge simulations at smaller time steps are closer to the observed behavior of the system. Sensitivity analysis results underline the importance of the Horton infiltration parameters and the percent of imperviousness of the catchment.
Description
Rights Access
Subject
global sensitivity analysis
SWMM
CLASIC
urban hydrology model
method of Sobol