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

https://doi.org/10.1016/j.compag.2020.105756Get rights and content

Highlights

  • 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.

Abstract

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.

Introduction

It is widely acknowledged that the earth is warming due to increasing greenhouse gas emissions generated by anthropogenic activities (López-Ballesteros et al., 2020). Over the period of 1901–2012, the mean global temperature has risen by 0.89 °C (IPCC, 2013), and the warming is expected to continue (Zhang et al., 2019). In a recent governmental report released by Environment and Climate Change Canada, the mean annual temperature in Canada is expected to increase by 2.0–6.0 °C by 2100 (ECCC, 2019). With warmer temperatures, and the atmosphere’s resulting greater capacity to hold water vapor, precipitation regimes will be altered (Rouhani and Leconte, 2018), and, accordingly the watersheds’ hydrologic cycle. For snowmelt-dominated regions, changes in precipitation will affect snow accumulation and thereby streamflow, while changes in temperature will most likely influence the timing of snowmelt (Barnett et al., 2005). According to the Emergency Preparedness Canada (EPC) electronic disaster database, over 65% of the nation’s flood disasters were attributed to snowmelt runoff, storm water, and storm rainfall runoff (Brooks et al., 2001). In the mountainous areas of the western United States, over 70% of runoff have been attributed to snowmelt (Li et al., 2017). Under a changing climate, higher latitude areas are more likely to experience increasingly extreme spring floods (Rouhani and Leconte, 2018). An analysis of Canadian hydrologic trends over the last 30 years has shown an increase in winter streamflow, a decrease in summer streamflow, along with earlier peak flows (Whitfield and Cannon, 2000, Zhang et al., 2000, Jacques and Sauchyn, 2009). A global analysis of climate change impacts on river flow regimes in a number of countries, including Canada, predicted a significant increase in the magnitude of peak flows, but with their occurrence shifted at least one month earlier (Arnell and Gosling, 2013).

In Quebec, high flow events and spring floods are primarily the result of snowmelt, which accounts for up to 40% of annual streamflow (Coulibaly et al., 2000, Ferguson, 1999). Between 1900 and 1997, approximately 14% of the national flood disasters occurred in Quebec. The highly populated areas of southern Quebec have experienced a number of major flood events in the last few decades, and the recent one in 2017 has flooded thousands of houses and caused great damage to the riverside communities (Rouhani and Leconte, 2018). Under changing climate conditions, the occurrence of extreme spring floods in Quebec is predicted to increase, largely as the result of an earlier snowmelt (Beauchamp et al., 2015, Rouhani and Leconte, 2018). It is therefore essential to assess the potential impact of climate change on streamflow characteristics, especially peak flows, and thereby better monitor floods and offer watershed management adaptations to mitigate these disasters.

Hydrological models with snowmelt modules have been recently applied to simulate streamflow in snowmelt-dominated areas (Ficklin and Barnhart, 2014, Lachance-Cloutier et al., 2017, Tang et al., 2019). Accurate modeling and simulation of streamflow under snowmelt conditions typical of snow-dominated regionss are critical to capture watershed’s physical and hydrological characteristics, and can provide guidance for local flood forecasting and control. In addition, the calibrated and validated models provide an option to forecast the long-term shifts in streamflow under various climate conditions. The ArcSWAT model is a semi-distributed hydrological model at a watershed-scale with different modules targeting various user demands (Arnold et al., 1998). Laying a solid foundation for further model application, the model provides accurate predictions of snowmelt-influenced streamflow at both daily and monthly time steps worldwide including West Seti River Basin (Bhatta et al., 2020), Himalayan River Basin (Bhatta et al., 2019), Heihe River Basin (Wu et al., 2015), Taleghan mountainous watershed (Noor et al., 2014), Outardes basin (Troin and Caya, 2014), Blue River watershed (Lemonds and McCray, 2007), Garonne watershed (Grusson et al., 2015), and Ontonagon River basin (Wu and Johnston, 2007).

Given its high accuracy and applicability towards long-term prediction of winter hydrological events in snow-dominated regions, several previous studies undertaken in southern Quebec have employed SWAT to investigate the hydrological response to climate change. These studies have suggested significant greater annual runoff, along with earlier snowmelt and discharge peaks (Gombault et al., 2015, Minville et al., 2008, Shrestha et al., 2012). Gombault et al. (2015) found that under future scenarios (2041–2070), spring floods would begin earlier, annual streamflow would increase by 9–19% and winter flow would increase by 2- to 3- fold as compared to the baseline period (1971–2000). While these studies were only concerned with the combined effects of temperature and precipitation on the peak flow of watersheds, the individual effects of temperature and precipitation changes on snowfall and snowmelt were not well addressed, an important omission since snowmelt is the major contributor to peak flows for the snow-dominated watersheds.

Located in the St Lawrence lowlands, the Nicolet River watershed has faced a growing risk of spring flooding in the recent years. In early April 2014, the watershed was added to a spring flood warning register, due to higher than normal water levels from spring thaw along and warming temperatures (CBC News, 2014). Road accesses to towns in this region were cut off because of high water levels in the spring of 2017. While no study of future streamflow in this watershed has been conducted, it is necessary to assess the impact of future climate change on the Nicolet River watershed hydrology, especially with respect to spring streamflow and snowmelt. Although Gombault et al. (2015) assessed the impact of climate change on the hydrology of the nearby Pike River watershed using a calibrated SWATqc model, they only applied four projected future climate datasets and there was a large variation among the four climate models in terms of spatial and seasonal changes. It is therefore necessary to apply multiple climate change models for the assessment of climate change impacts in this region.

The objectives of this study were to evaluate the ArcSWAT model’s capacity to accurately simulate streamflow in a snow-dominated watershed, and subsequently, employing eleven projected climate scenarios, to quantify the potential impact of climate change on the Nicolet River watershed’s hydrology as affected by snowmelt. This research focused on the hydrologic changes induced by climate change between the baseline period (1983–2000) and projected future climates (2050–2067), especially in the winter and spring seasons. Furthermore, both the combined and individual effects of temperature and precipitation change on future annual and seasonal streamflow as well as peak flow, as affected by snowmelt were assessed.

Section snippets

Site description

The Nicolet River, as a southern tributary of the Saint Lawrence River, drains a watershed approximately 3380 km2 into Lake Saint-Pierre. The research area intersected the downtown portions of the towns of Victoriaville and Warwick (Fig. 1). The stream flow data from hydrological station 02OD003 was downloaded from the Environment Canada’s HYDAT database. Weather data, including precipitation and air temperature, were retrieved from four meteorological stations (Station ID: 701HE63; 7027783;

Model calibration and validation

The simulated streamflow closely matched the observed streamflow in both calibration and validation periods (Fig. 4). After model calibration, the model showed satisfactory model performance for simulating both the daily flow (NSE = 0.55, RSR = 0.67, and PBIAS = 9%) and monthly streamflow (NSE = 0.75, RSR = 0.50 and PBIAS = 9%) with an overall water balance error of 5% (Table 4). Validated over the period of 1996–2000 on both daily and monthly time steps, the ArcSWAT model showed adequate model

Conclusions

In this study, the ArcSWAT model was successfully parameterized for the Nicolet River watershed. The model was calibrated using measured streamflow data from 1986 to 1990 and validated against similar data from 1991 to 2000. Statistical analysis suggested a satisfactory performance of the ArcSWAT model in simulating daily and monthly streamflow in both calibration and validation phases (NSE > 0.5, RSR < 0.7 and PBIAS<15%). Subsequently, eleven sets of projected future climate data were drawn

CRediT authorship contribution statement

Qianjing Jiang: Conceptualization, Writing - review & editing. Zhiming Qi: Conceptualization, Supervision, Writing - review & editing. Fei Tang: Visualization, Writing - original draft. Lulin Xue: Resources, Data curation. Melissa Bukovsky: Resources, Data curation.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

The research was funded by NSERC-Discovery Grant (Effect of management practices on hydrology and nutrient losses from a tile-drained field under freeze-thaw conditions), under Grant Agreement No. RGPIN-2019-05662.

References (53)

  • J.E. Nash et al.

    River flow forecasting through conceptual models part I—A discussion of principles

    J. Hydrol.

    (1970)
  • H. Rouhani et al.

    A methodological framework to assess pmp and pmf in snow-dominated watersheds under changing climate conditions – a case study of three watersheds in Québec (Canada)

    J. Hydrol.

    (2018)
  • R.R. Shrestha et al.

    Modelling of climate-induced hydrologic changes in the Lake Winnipeg watershed

    J. Great Lakes Res.

    (2012)
  • N.K. Shrestha et al.

    Assessing climate change impacts on fresh water resources of the Athabasca River Basin

    Canada. Sci. Total Environ.

    (2017)
  • G. Tang et al.

    Streamflow response to snow regime shift associated with climate variability in four mountain watersheds in the US Great Basin

    J. Hydrol.

    (2019)
  • F. Wu et al.

    Streamflow variation due to glacier melting and climate change in upstream Heihe River Basin, Northwest China

    Phys. Chem. Earth, Parts A/B/C

    (2015)
  • K. Wu et al.

    Hydrologic response to climatic variability in a Great Lakes Watershed: A case study with the SWAT model

    J. Hydrol.

    (2007)
  • L. Zhang et al.

    Using boosted tree regression and artificial neural networks to forecast upland rice yield under climate change in Sahel

    Comput. Electron. Agr.

    (2019)
  • Q. Zhao et al.

    Projecting climate change impacts on hydrological processes on the Tibetan Plateau with model calibration against the glacier inventory data and observed streamflow

    J. Hydrol.

    (2019)
  • J.G. Arnold et al.

    SWAT: Model use, calibration, and validation

    Trans. ASABE.

    (2012)
  • Arnold, J.G., Kiniry, J.R., Srinivasan, R., Williams, J.R., Haney, S., Neitsch, S.L., 2012b. Soil and Water Assessment...
  • J.G. Arnold et al.

    Large area hydrologic modeling and assessment part I: Model development1

    J. Am. Water. Resour. AS.

    (1998)
  • O. Aygün et al.

    Impacts of climate change on the hydrology of northern midlatitude cold regions

    Prog. Phys. Geogr.: Earth Environ.

    (2019)
  • T.P. Barnett et al.

    Potential impacts of a warming climate on water availability in snow-dominated regions

    Nature

    (2005)
  • B. Bhatta et al.

    Modelling the impact of past and future climate scenarios on streamflow in a highly mountainous watershed: A case study in the West Seti River Basin, Nepal

    Sci. Total Environ.

    (2020)
  • Brooks, G., Evans, S., Clague, J., 2001. A synthesis of natural geological hazards in Canada. Geological Survey of...
  • Cited by (12)

    • Changes in floodplain regimes over Canada due to climate change impacts: Observations from CMIP6 models

      2021, Science of the Total Environment
      Citation Excerpt :

      A significant fraction in the southern and central parts of the Northern Quebec drainage basin is also expected to face a similar situation. Jiang et al. (2020) employed the ArcSWAT model to quantify the impacts of climate change on Nicolet's river watershed in Quebec. The authors reported that the peak streamflow annually may not only rise by 13% in the future, but their dominant occurrence in winter (rather than spring) may elevate the risk of future winter flooding.

    View all citing articles on Scopus
    View full text