Lake–river and lake–atmosphere interactions in a changing climate over Northeast Canada
Lakes influence the regional climate and hydrology in a number of ways and therefore they should be represented in climate models in a realistic manner. Lack of representation of lakes in models can lead to errors in simulated energy and water fluxes, for lake-rich regions. This study focuses on the assessment of the impact of climate change on lakes and hydrology as well as on the influence of lakes on projected changes to regional climate and surface hydrology, particularly streamflows, for Northeast Canada. To this end, transient climate change simulations spanning the 1950–2100 period are performed, with and without lakes, with the fifth generation of the Canadian Regional Climate Model (CRCM5), driven by the Canadian Earth System Model (CanESM2) at the lateral boundaries for Representative Concentration Pathway 8.5. An additional CRCM5 simulation, driven by European Centre for Medium-Range Weather Forecasts Re-Analysis Interim (ERA-Interim) for the 1980–2010 period, is performed in order to assess performance and boundary forcing errors. Performance errors are assessed by comparing the ERA-Interim-driven simulation with available observation datasets, for the 1980–2010 period, for selected variables: 2-m air temperature, total precipitation, snow water equivalent and streamflow. The validation results indicate reasonable model performance over the study region. Boundary forcing errors are studied by comparing ERA-Interim-driven simulation with the one driven by CanESM2 for the current 1980–2010 period, to identify regions and seasons for which projected changes should be interpreted with extra caution. Comparison of projected changes from the CRCM5 simulations with and without lakes suggest that the presence of lakes results in a dampening of projected increases to 2-m air temperature for all seasons almost everywhere in the study domain, with maximum dampening of the order of 2 °C occurring during winter, mostly in the vicinity of the lakes. As for streamflows, projected increases to spring streamflows, based on the simulation with lakes, are found to be smaller than that without lakes and this is due to the storage effect of lakes. Similarly, lower decreases in summer streamflows in future climate are noted in the simulation with lakes due to the gradual release of snowmelt water stored in lakes. An additional CRCM5 transient climate change simulation with lakes and interflow, i.e. lateral flow in the soil layers, is compared with the simulation with lakes, but without interflow, to assess the impact of interflow on projected changes to regional climate and hydrology. Maximum interflow is projected to shift earlier in spring and the maximum interflow rate is expected to decrease by around 25 % in future. Results suggest that the impact of interflow on projected changes to precipitation, soil moisture and humidity are modest, even though the interflow intensity is changing noticeably in future climate. The impact of the interflow on projected changes to streamflows is in the range of ±50 m3/s. This study thus for the first time demonstrates the impact of lakes and interflow on projected changes to the regional climate and hydrology for the study region using a single regional modelling system.