Quantifying the role of calibration strategies on surface-subsurface hydrologic model performance

Authors: ABBAS, SALEM - Colorado State University; BAILEY, RYAN - Colorado State University; WHITE, JEREMY - Intera, Inc; Arnold, Jeffrey; White, Michael

Submitted to: Hydrological Processes
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/19/2024
Publication Date: 10/1/2024
Citation: Abbas, S.A., Bailey, R.T., White, J.T., Arnold, J.G., White, M.J. 2024. Quantifying the role of calibration strategies on surface-subsurface hydrologic model performance. Hydrological Processes. https://doi.org/10.1002/hyp.15298.
DOI: https://doi.org/10.1002/hyp.15298

Interpretive Summary: This research explores how different variables affect the accuracy and efficiency of hydrologic models, in the prediction of streamflow and groundwater head water. Using the SWAT+ model with the gwflow module, the study tested various numbers of calibration parameters and simulation duration in the Winnebago River watershed (Minnesota, Iowa), and the Nanticoke River watershed (Delaware, Maryland. The research found that using around 15 calibration parameters optimally balances accuracy and computational effort. Both streamflow and groundwater levels were important targets for calibration. The method presented here can be used for any watershed to improve the understanding and prediction of water flow in integrated surface and subsurface environments.

Technical Abstract: Distributed parameter coupled surface-groundwater hydrologic models are often overparameterized due to the inclusion of a large number of parameters that are necessary to reflect the spatially diverse nature of hydrologic processes. The impact of the number and type of calibration parameters and observations on the degree of overparameterization, convergence of parameter values, model performance, and uncertainty remains unclear. The objective of this paper is to quantify the impact of variables within a Sensitivity Analysis, Uncertainty Analysis, and Parameters Estimation (SA-UA-PE) framework on model testing for a holistic hydrologic model. Framework variables include 1) the number (5, 10, 15, 20) and type (soil, aquifer, land surface, channel) of calibration parameters; 2) the type of calibration target (streamflow, groundwater head); and 3) the length of testing period (1 to 14 years). Model testing variables include monthly streamflow, groundwater head, and flow duration curves for streamflow. The method is demonstrated for the Winnebago River watershed (Minnesota, Iowa), with significant tile drainage, and the Nanticoke River watershed (Delaware, Maryland), with significant groundwater-channel interactions. The selected hydrologic model is SWAT+, using the gwflow module for physically based groundwater storage and flow modeling, and simulations are run for the 2000-2015 period. Through this process, we found that increasing the number of parameters from 5 to 15 improves representation of streamflow, principally through an improvement of groundwater storage representation and baseflow generation, but minimal or no improvement when increasing to 20 parameters. Therefore, the SA-UA-PE process can be optimized based on an ideal number of parameters that yields accurate results while maintaining an adequate computational burden. Furthermore, the type of parameters included in the process have a significant impact on generated watershed fluxes (runoff, soil lateral flow, ET, recharge, etc.), and hence both streamflow and groundwater head should be included as calibration targets. The method presented here can be used for any watershed, using integrated surface-subsurface hydrologic models.