Sensitivity Analysis and Insights into Hydrological Processes and Uncertainty at Different Scales

Abstract:

Sensitivity analysis (SA) is an essential tool for providing insight into model behavior, and conducting model calibration and uncertainty assessment. Numerous techniques have been used in environmental modelling studies for sensitivity analysis. However, it is often overlooked that the scale of modelling study, and the metric choice can significantly change the assessment of model sensitivity and uncertainty. In order to identify important hydrological processes across various scales, we conducted a multi-criteria sensitivity analysis using a novel and efficient technique, Variogram Analysis of Response Surfaces (VARS). The analysis was conducted using three different hydrological models, HydroGeoSphere (HGS), Soil and Water Assessment Tool (SWAT), and Modélisation Environmentale–Surface et Hydrologie (MESH). Models were applied at various scales ranging from small (hillslope) to large (watershed) scales. In each case, the sensitivity of simulated streamflow to model processes (represented through parameters) were measured using different metrics selected based on various hydrograph characteristics such as high flows, low flows, and volume. We demonstrate how the scale of the case study and the choice of sensitivity metric(s) can change our assessment of sensitivity and uncertainty. We present some guidelines to better align the metric choice with the objective and scale of a modelling study.