How Robust are our Hydrologic Models in Simulating Streamflow Alterations in a Changing Climate?

Abstract:

Hydrologic models are powerful tools for assessing streamflow alterations from natural and anthropogenically driven changes such as land use, water use, and climate change. However, the ability of the models to replicate the magnitude and direction of changes of different components of the hydrograph is not clear. Hence, we initiated a study to evaluate the replicability of the different streamflow components and their changes by employing the widely used Variable Infiltration Capacity (VIC) hydrologic model for the Fraser River Basin, Canada. We analyzed the replicability of a range of indicators of hydrologic alterations (IHA) by using climate inputs derived from gridded observations and statistically downscaled global climate models. Given that the hydrologic regime of the region is known to be influenced by teleconnections to the Pacific Decadal Oscillation (PDO) and El Niño–Southern Oscillation (ENSO), we used hydrologic responses to the PDO and ENSO states as analogues for evaluating the model’s ability to simulate climate-induced changes to streamflow. The results show a generally good skill of the model in replicating the annual and peak flow related IHAs, but more limited replicability of the minimum flow and flow pulse related IHAs. Additionally, while the results revealed a generally good replication of the qualitative patterns and direction of changes between the different climate states, the magnitude of change for some of the indicators showed considerable differences. Hence, there is a need to exercise caution in the use of model-simulated indicators, and the replicability of both magnitude and direction of change need to be carefully examined before using the simulated indicators for projecting future changes to the flow regime.