The use of CMIP5 data to simulate climate change impacts on flow regime within the Lake Champlain Basin

•High uncertainty in regional hydrologic modeling when using CMIP5 climate data.•Lake Champlain streamflow regime implies adaptation to increased flood risks.•Streamflow regime projections suggest an increase in low flows.

Study regionLake Champlain Basin, northwestern New England, USA.Study focusOur study uses regional hydrologic analyses and modeling to examine alternative possibilities that might emerge in the Lake Champlain Basin streamflow regime for various climate scenarios. Climate data as well as spatial data were processed to calibrate the Regional Hydro-Ecological Simulation System (RHESSys) model runoff simulations. The 21st century runoff simulations were obtained by driving the RHESSys model with climate data from the Coupled Model Intercomparison Project phase 5 (CMIP5) for representative concentration pathways RCP 4.5 and 8.5.New hydrological insights for the regionOur analyses suggest that most of CMIP5 ensembles fail to capture both the trends and variability observed in historical precipitation when run in hindcast. This raises concerns of using such products in driving hydrologic models for the purpose of obtaining reliable runoff projections that can aid researchers in regional planning. A subset of five climate models among the CMIP5 ensembles have shown statistically significant trends in precipitation, but the magnitude of these trends is not adequately representative of those seen in observed annual precipitation. Adjusted precipitation forecasts project a streamflow regime described by an increase of about 30% in seven-day maximum flow, a four days increase in flooded days, a three orders of magnitude increase in base flow index, and a 60% increase in runoff predictability (Colwell index).