Modeling climate change impact on streamflow as affected by snowmelt in Nicolet River Watershed, Quebec

• Published in: Computers and electronics in agriculture 2020-11, Vol.178, p.105756, Article 105756
• Main Authors: Jiang, Qianjing, Qi, Zhiming, Tang, Fei, Xue, Lulin, Bukovsky, Melissa
• Format: Article
• Language: English
• Subjects: ArcSWAT; Climate change; Climate models; Evapotranspiration; Flooding; Floods; Hydrology; Peak flows; Precipitation; Regional analysis; Runoff; Snow; Snowfall; Snowmelt; Spring (season); Stream flow; Streamflow; Temperature; Watersheds; Weather forecasting; Winter
• ISSN: 0168-1699; 1872-7107
• DOI: 10.1016/j.compag.2020.105756

•Higher risk in winter flooding was forecasted in Nicolet River watershed.•Eleven sets of future climate data were projected and applied to ArcSWAT model.•Predicted future peak flow amount increased due to increased precipitation.•Predicted future peak flows tended to occur earlier due to warmer temperature. Frequent spring flooding in Southern Quebec’s Nicolet River watershed has a history of causing severe damage, which is likely to worsen as climate change progresses. Employing the ArcSWAT model, an attempt was made to assess the potential impacts of climate change on the Nicolet River watershed’s seasonal and annual streamflow, particularly that portion affected by snowmelt. Calibrated and validated against observed streamflow data for the periods of 1986–1990 and 1991–2000, respectively, the model reliably predicted daily streamflow (e.g., percent bias within ±15%, Nash-Sutcliffe model efficiency >0.50, and the ratio of root mean square error to the standard deviation ≤0.70). In an effort to investigate the impacts of climate change on streamflow, future climate datasets were generated for 2053–2067 by implementing the eleven sets of existing Regional Climate Model (RCM) simulations produced for the North American Regional Climate Change Assessment Program (NARCCAP) in the ArcSWAT model. The ArcSWAT model’s hydrological responses were closely tied to changes of climate variables: a strong correlation existed between simulated runoff and precipitation, and between temperature and predicted evapotranspiration, snowfall, and winter snowmelt. Projected future climate data showed increases in both average temperature (+2.5 °C) and precipitation (+21%). Significant greater total precipitation was forecasted for the winter season, while total snowfall was projected to decrease by 6%. However, the snowmelt showed an increasing trend for the late winter and earlier spring period. Streamflow was expected to increase annually and in most seasons except spring. Annual peak flows volumes would increase by 13% in the future and the occurrence of peak flows would shift to the winter (vs. the spring), indicating a greater risk of winter flooding in the future. The individual impact of temperature and precipitation on peak flows showed that increases in peak flows were mainly tied to increased precipitation, while the shift in their timing was mostly tied to warming temperatures.