Data for: Quantifying the uncertainty created by non-transferable model calibrations across climate and land cover scenarios: A case study with SWMM
THIS REPOSITORY This repository contains data used in support of the manuscript: "Quantifying the uncertainty created by non-transferable model calibrations across climate and land cover scenarios: A case study with SWMM”. This includes: urban image patterns, SWMM simulations, SVE model outputs, and Python scripts. Detailed SVE model documentation is provided separately on GitHub: https://github.com/octavia-crompton/SVE-R. BACKGROUND This research asked: 1. For a given subcatchment, how much do calibrated k_width and f_idc (width and connected impervious fraction) parameters vary across environmental conditions (given by storm intensity, impervious fractions, and saturated hydraulic conductivity of pervious areas)? 2. How transferable are these parameters, calibrated under one set of environmental conditions, to different environmental conditions? 3. What magnitude of error can arise from the uncertainty in transferability of calibration parameter (or `calibration transfer uncertainty')? METHODS To answer these questions, we calibrated a semi-distributed SWMM model to a distributed model based on the SVE. This process involved (1) conducting a Sobol Sensitivity Analysis of SWMM overland flow parameters; (2) generating urban images with different impervious and pervious patterns for simulation in the SVE; (3) Simulating the urban images directly in the SVE across different soil and storm conditions; (4) Simulating analogous urban patterns in SWMM across the same soil and storm conditions, and varying k_width and f_idc; and (5) compiling the SWMM and SVE outputs to calibrate the k_width and f_idc parameters, plot behavioral parameter space, and calculate their transferability between conditions.