Journal of Hydrologyhuardda/articles/boyer10.pdf · to 6 of latitude. The rivers differ in their...

Journal of Hydrology 384 (2010) 65–83

Contents lists available at ScienceDirect

Journal of Hydrology

journal homepage: www.elsevier .com/ locate / jhydrol

Impact of climate change on the hydrology of St. Lawrence tributaries

Claudine Boyer a,*, Diane Chaumont b, Isabelle Chartier c, André G. Roy a

a Département de géographie, Université de Montréal, C.P. 6128, Succursale Centre-ville, Montréal (Québec), Canada H3C 3J7b Ouranos, 550, Sherbrooke Ouest, Tour Ouest, 19e étage, Montréal (Québec), Canada H3A 1B9c Institut de recherche d’Hydro-Québec (IREQ), 1800, boul. Lionel-Boulet, Varennes (Québec), Canada J3X 1S1

a r t i c l e i n f o

Article history:Received 17 April 2009Received in revised form 11 January 2010Accepted 20 January 2010

This manuscript was handled by K.Georgakakos, Editor-in-Chief, with theassistance of Ashish Sharma, AssociateEditor

Keywords:RiverHydrologyClimate changeStream flowVariabilityModeling

0022-1694/$ - see front matter � 2010 Elsevier B.V. Adoi:10.1016/j.jhydrol.2010.01.011

* Corresponding author. Tel.: +1 514 343 8035; faxE-mail address: [email protected] (C. B

s u m m a r y

Changes in temperature and precipitation projected for the next century will induce important modifica-tions into the hydrological regimes of the St. Lawrence tributaries (Quebec, Canada). The temperatureincrease anticipated during the winter and spring seasons will affect precipitation phase and conse-quently the snow/precipitation ratio and the water volume stored into snow cover. The impact on north-ern river hydrology and geomorphology will be significant. In this study we aim to assess the magnitudeof the hydrological alteration associated with climate change; to model the projected temporal shift inthe occurrence of winter/spring center-volume date; to assess the sensitivity of the winter/spring cen-ter-volume date to changes in climatic variables and to examine the latitudinal component of the pro-jected changes through the use of five watersheds on both shores of the St. Lawrence. The studyemphasizes changes in the winter and spring seasons. Projected river discharges for the next centurywere generated with the hydrological model HSAMI run with six climate series projections. Three Gen-eral Circulation Models (HadCM3, CSIRO-Mk2 and ECHAM4) and two greenhouse gas emissions scenarios(A2 and B2) were used to create a range of plausible scenarios. The projected daily climate series wereproduced using the historical data of a reference period (1961–1990) with a perturbation factor equiva-lent to the monthly mean difference (temperature and precipitation) between a GCM in the future forthree 30 year horizons (2010–2039, 2040–2069; 2070–2099) and the reference period. These climateprojections represent an uncertainty envelope for the projected hydrologic data. Despite the differencesdue mainly to the GCM used, most of the hydrological simulations projected an increase in winter dis-charges and a decrease in spring discharges. The center-volume date is expected to be in advance by22–34 days depending on the latitude of the watershed. The increase in mean temperature with thesimultaneous decrease of the snow/precipitation ratio during the winter and spring period explain a largepart of the projected hydrological changes. The latitude of the river governed the timing of occurrence ofthe maximum change (sooner for tributaries located south) and the duration of the period affected bymarked changes in the temporal distribution of discharge (longer time scale for rivers located at higherlatitudes). Higher winter discharges are expected to have an important geomorphological impact mostlybecause they may occur under ice-cover conditions. Lower spring discharges may promote sedimentationinto the tributary and at their confluence with the St. Lawrence River. The combined effects of modifica-tions in river hydrology and geomorphological processes will likely impact riparian ecosystems.

� 2010 Elsevier B.V. All rights reserved.

Introduction

The hydrological regime of northern rivers could be severelymodified in response to the anticipated changes in temperatureand precipitation during the present century. For the Great-Lakes-St. Lawrence watershed (USA and Canada), a reduction rang-ing between 4% and 24% of the mean annual discharge is projectedfor the next 90 years as a consequence of current scenarios of cli-mate change (Croley, 2003). For this large watershed, increased

ll rights reserved.

: +1 514 343 8008.oyer).

evaporation from lakes due to a rise in temperature explains alarge part of the projected reduction (Croley, 2003). Shifts in thetiming and amount of input runoff are also expected to occur(Mortsch and Quinn, 1996). Downstream of Lake Ontario, seasonalchanges in the discharge of the St. Lawrence River may be accentu-ated or attenuated by the water regulation plan in order to modu-late the temporal variation in the Great Lakes levels.

Winter and spring seasons are particularly vulnerable tochanges in air temperature. Warmer temperature during the win-ter season can increase the number of days with air temperatureabove zero Celsius resulting in more frequent rain events. Theseevents will contribute to an increase of winter runoff and not to

http://dx.doi.org/10.1016/j.jhydrol.2010.01.011

mailto:[email protected]

http://www.sciencedirect.com/science/journal/00221694

http://www.elsevier.com/locate/jhydrol

66 C. Boyer et al. / Journal of Hydrology 384 (2010) 65–83

the accumulation of a snowpack (Whitfield et al., 2003). The rela-tive amount of precipitation falling as rain or snow will directly af-fect water supplied from the snowpack in the spring and theamplitude and timing of river flows during the winter and spring(Hodgkins et al., 2003). Cooley (1990) has suggested that changingmean air temperature by 2–4 �C may have a significant impact onthe accumulation and melt of a snowpack depending on the origi-nal temperature regime of the site. These changes will alter thehydrological regime.

Modification of winter and spring streamflows has been ob-served in southeastern Canada and northeastern USA during thecourse of 20th century (Hodgkins et al., 2003; Hodgkins and Dud-ley, 2006; Whitfield and Cannon, 2000; Zhang et al., 2001).Changes toward earlier freshet (spring thaw resulting from snowand ice melt in rivers located in the northern latitudes) were foundto be significant for most of these areas. Hodgkins and Dudley(2006) also found that river flows in January, February and Marchshow a tendency to increase from 1953 to 2002 in northeasternUSA. Conversely, river flows in April and May show a relative de-crease during the same period. These changes are attributed tolong term changes in temperature and their impact on the phase(snow or rain) of precipitation. Decline in the annual ratio of snowto total precipitation has been reported from many climatologicstations in New-England from 1948 to 2000 with the most impor-tant decrease occurring after 1975 (Burakowski et al., 2008; Hun-tington et al., 2004). This negative trend has been partly linkedwith positive North Atlantic Oscillation (NAO) anomalies indexwhich is associated with mild winters (in eastern USA) and lowsnowfall to rain ratio. Large-scale atmospheric and oceanic oscilla-tions (e.g. North Atlantic Oscillation, Pacific Decadal Oscillation)account for most of the climate natural variability by modulatingprecipitation and temperature regimes through the regulation ofthe number and intensity of significant weather events particularlyduring the winter and early spring. These oscillations influence thesnowpack variability and the timing and magnitude of flood peaksat a decade scale (Cayan, 1996; Hartley and Keables, 1998; Jain andLall, 2000, 2001; Thompson and Wallace, 2001). Global warmingeffects will be superimposed on NAO or other large-scale oscilla-tions and it remains uncertain whether global climate warmingwill influence the variability of those oscillations (Hurrell et al.,2006; Visbeck et al., 2001).

Changes in temperature and precipitation and the shift in win-ter precipitation from snow to rain will be crucial for the hydrolog-ical regime of St. Lawrence tributaries. For the future periods(2010–2039, 2040–2069 and 2070–2099), a few studies in Québechave already suggested an increase in winter flow and decrease inspring flow compared to a historical reference period (Fortin et al.,2007; Minville et al., 2008; Quilbé et al., 2008). Higher winterflows, lower spring flows and changes in the timing of spring run-off can also have important impacts on fluvial processes, watermanagement and on riparian and aquatic ecosystems. The ampli-tude, duration and timing of spring floods play a critical role onthe structure and diversity of aquatic ecosystems (Toner and Ked-dy, 1997). Stronger winter floods can also substantially modify thephysical characteristics of habitats, enhance river channel erosionand alter stream conditions for winter spawning fish species. Theseeffects could be enhanced as winter floods could occur under ice-cover conditions.

This study is part of a larger project that aims to model the mor-phological and sedimentological response of St. Lawrence tributar-ies to the anticipated environmental changes (hydrological andbase level drop changes). The paper focuses on future changes inthe hydrological regimes of the St. Lawrence tributaries as a resultof projected climate changes. The specific objectives of the studyare to: (1) analyze changes in the simulated river discharges atthe annual to monthly scales; (2) quantify the projected temporal

shift in the occurrence of spring peak discharges (represented bythe winter/spring center-volume date); (3) assess the sensitivityof the winter/spring center-volume date to changes in climaticvariables and (4) identify similarities and differences among theSt. Lawrence tributaries. The paper presents historical (1932–2005) and projected (2020s, 2050s and 2080s) climatic and hydro-logical data for five tributaries with a particular focus on the winterand spring period. This research significantly adds to other recentstudies in Québec and Eastern Canada because the selected water-sheds cover a latitudinal range from 43.2�N to 49.1�N on both thesouth and north shores of the St. Lawrence. The selection of riversspans a current seasonal mean air temperature gradient for thewinter and spring period from �4 �C to 0 �C.

Methodology

All located along the St. Lawrence fluvial corridor betweenMontreal and Quebec City, the studied tributaries are the Richelieuand St-François rivers on the south shore and the Yamachiche, St-Maurice and Batiscan rivers on the north shore (Fig. 1). Thesewatersheds are mostly located in low relief areas and cover closeto 6� of latitude. The rivers differ in their hydrology, sedimentologyand dynamics and they are representative of the diversity of tribu-taries along the St. Lawrence. Except for the Yamachiche River, allwatersheds are exploited for hydro electricity or influenced bymultiple dams used for flood control, water intake or recreationalactivities. Data concerning the management plan of these struc-tures are, however, not available and cannot be accounted for inthe hydrological simulations. Exploitation of the rivers for hydroelectricity and flood control began before 1950 and around 1964.The impact of these structures on the natural regime of the riveris low for the Batiscan and Richelieu and moderate for the St-François. The natural hydrological regime of the St-Maurice Riverhas been substantially modified by water management for hydroelectricity. To reduce the impact of this hydrological control onthe St-Maurice, we have elected to study the response of a smallerwatershed, LaGabelle, instead of the whole basin. The LaGabellewatershed is used by Hydro-Quebec (Québec national hydro elec-tricity company) to study the natural response of the drainage ba-sin to meteorological variations.

Hydrological modeling

Hydrological simulations were performed with the HSAMImodel (Bisson and Roberge, 1983; Chaumont and Chartier,2005; Fortin, 2000). This model is a lump rainfall (rain-and-snow)runoff model. It is a discrete time conceptual model containingthree linear reservoirs (snow cover, surface water (surface runoffand base discharge), unsaturated and saturated zones) in cascadewhich generate impulses filtered by two hydrograph units. Snowaccumulation (following a degree-day approach), snow melt, soilfreezing and thawing, evapotranspiration (estimated from dailymaximum and minimum temperatures) and vertical and horizon-tal transit of water are simulated by the model with a system ofequations and empirical parameters which are adjusted duringthe calibration of the model. Simulations are carried out with atime step of one day. HSAMI model is simple and easy to usefor the estimation of potential impacts of climate change on waterresources. This model has been used for more than twenty yearsby Hydro-Quebec over the Québec province to predict runoff fortheir reservoirs. It has been largely tested and successfully appliedover the same region and more northern watersheds (Chartier,2006; St-Hilaire et al., 2003). It requires a small amount of inputdata and optimization of the parameters is done automaticallyusing the shuffled complex evolution method (Duan et al.,

Fig. 1. Location of the studied river basins. The GCM grids are represented HadCM3 (black solid lines), ECHAM4 (black dot lines) and CSIRO-Mk2 (black dash lines). Thereference region, for which monthly projected changes were calculated, is represented by the red dash line.

C. Boyer et al. / Journal of Hydrology 384 (2010) 65–83 67

1992). The model uses input data (minimum and maximum tem-perature, rain and snow) that are averaged over the basin. Themodel distinguishes the phase of precipitation (rain or snow). Itwas thus necessary to carry out a partition between snow andrain according to the average temperature. A linear transfer wasapplied when the daily mean temperature was between �2 �Cand +2 �C. All precipitation was converted into snow at �2 �Cand kept as rain at 2 �C.

Calibration and validation of HSAMI was carried on for each wa-tershed over the reference period, 1961–1990 (Table 1). For thisanalysis, we have used discharges recorded by the Hydat network(Environment Canada and provincial partnership) at the mostdownstream gauging station. The period used for model calibrationon each river was dictated by the availability of data. For theYamachiche River, where no gauging station exists, we have useddata recorded on a larger neighboring watershed with a drainagearea ratio to estimate historical mean daily discharges.

For each studied watershed, a unique mean time series for tem-perature (maximum and minimum daily temperatures) and for pre-cipitation (daily amounts of rain and snow) were used in HSAMI to

calibrate the model and to build the simulated daily discharge datafor the reference period. These time series were created by spatiallyaveraging the daily observed data taken from several meteorologicalstations. We have selected meteorological stations located within aradius distance of 50 km from the tributary. Observed meteorologi-cal data were extracted from the Environment Canada network andthe American NOAA network (for the Richelieu). Thiessen polygonswere used to calculate area-weighted average temperature and pre-cipitation for each watershed (Heywood et al., 2006).

We have evaluated the quality of the simulated discharges withthe Nash–Sutcliffe coefficient, E (Nash and Sutcliffe, 1970). Thiscoefficient sums up the daily square differences between observedand simulated data over a year (Eq. (1)).

E ¼ 1�PT

t¼1ðQtobs � Q t

simÞ2

PTt¼1ðQ

tobs � Q obsÞ2

ð1Þ

where t is the day, Qobs and Qsim are observed and simulated dis-charge, respectively. The main priority during the calibration pro-cesses was to minimize the annual and spring peak volume

Table 1Drainage area of the studied rivers and period of calibration and validation of the hydrological data.

River Total area of thedrainage basin (km2)

Area ofcalibration (km2)

Hydrological data(observed and estimated)period

Calibration perioda Validation period Nash coefficient

Yamachiche 269 269 1932–2005 1960–1972 1980–1990 0.842Batiscan 4,700 4,580 1932–2005 1950–1975 1976–1990 0.849St-Maurice (LaGabelle) 43,250 716 1932–2005 1974–1987 1988–1998 0.827St-François 10,180 9,610 1932–2005 1960–1975 1976–1990 0.831Richelieu 23,720 22,000 1932–2005 1960–1975 1976–1990 0.789

a It is important to note the calibration and validation period may differ from one basin to the other depending of the availability of the data.


differences. We have considered the model to be well calibratedwhen a threshold value of 0.75 was obtained or exceeded (Table 1).Hydrological simulations for the future periods were generatedwith the calibrated model and the projected climate variable series.

Modeling climate variables

Climate models and GHG scenariosThe output of three GCMs driven by two greenhouse gas emission

(GHG) scenarios (A2 and B2) were used to construct climate variabletime series over three future periods (2010–2039, 2040–2069 and2070–2099). The three chosen climate models are: ECHAM4 (Roeck-ner et al., 1992), CSIRO-Mk2 (Hirst et al., 1996, 1999) and HadCM3(Gordon et al., 2000; Pope et al., 2000) (Table 2). From the six climatemodels published by the IPCC Data Distribution Centre associatedwith the Third Assessment Report (Giorgi et al., 2001), these threemodels are the only ones that include a multilayer surface schemewhich allows for the minimization of the bias in the reconstructionand simulation of the surface processes (Crossley et al., 2000; Verse-ghy, 1996). This is of paramount importance for simulating climatechange in regions subjected to rigorous winters.

The selection of models covers a wide spectrum of temperatureand precipitation anomalies. In southern Québec, the ECHAM4model is usually projecting very little changes in precipitationand a moderate increase in mean temperature compared to theother models. The HadCM3 model projects the largest increase inprecipitation and the lowest increase in mean temperature. Thevalues obtained from the CSIRO-Mk2 model are usually in betweenthose two models for precipitation and show the highest tempera-ture increase. The use of an intermediate model gives less weightto the two models that represents the extreme of the studied spec-trum for one or the other climate variables. However, the use ofmore models would have given a better estimate of the uncertaintyassociated with the projected climates variables (Johnson andSharma, 2009). The number of models and the variety of modelschosen was partly dictated by the main objective of the largerstudy which is to model river response to hydrological and base le-vel changes as a result of projected climate changes. This third le-vel of modeling has constrained the number of models that couldeffectively be used in the study.

The climate models and scenarios were selected to meet theobjectives of producing a range of plausible climate scenarios interms of the amplitude of change in precipitation and temperatureand of capturing a part of the uncertainties associated with the cli-

Table 2Global climate models and scenarios of greenhouse gas concentration used to generate cl

GCM Research center and country

CSIRO-Mk2 Australia’s Commonwealth Scientific and Industrial Research OECHAM4 Max Planck Institute for Meteorology, GermanyHadCM3 UKMO United Kingdom Meteorological Office, United Kingdom

mate models themselves and the GHG scenarios used (Houghtonet al., 2001).

Projected climate variable time series: selection and justifications ofthe downscaling method

In order to apply the GCMs at a regional scale and create futureclimate variable time series for local hydrological impact assess-ment, three main approaches can be used: (1) the dynamicaldownscaling method (Regional Climate Model) with bias correc-tions; (2) the statistical downscaling method and (3) the perturba-tion (or delta) method. The advantages and disadvantages of eachapproach will not be discussed in details here. However, detailsregarding the justification of our choice of the perturbation methodfor this study are given.

In the perturbation approach, the observed climate variablesduring the reference period are scaled by the monthly anomaliescalculated, for the temperature and precipitation amounts, be-tween the future and reference periods for a given GCM simulation.The perturbation method assumes that: (1) the biases of the GCMare similar during the reference and the future periods and (2)temporal variability (daily to inter annual) of the observed climatevariables during the reference period is maintained for the simu-lated series. This method is simple and can be used to generate awide range of plausible climate scenarios from a group of globalclimate models which is an important aspect of this study. It isused as an early phase method to assess the sensitivity of the riversto changes in the mean values (annual to monthly scale) of climatevariables (Diaz-Nieto and Wilby, 2005). It has the advantage ofbeing stable and robust (Graham et al., 2007) and it has been usedin many studies (Andréasson et al., 2004; Hay et al., 2000; Hewis-ton, 2003; Merritt et al., 2006; Minville et al., 2008; Prudhommeet al., 2002, 2003). This provides a strong basis for comparison ofour results with previous studies. Despite the limitations of theperturbation method (e.g. the analysis of extreme statistics, suchas summer extreme runoff is not appropriate), it is a general meth-od that can account for a wide range of GCMs and consequently al-low the assessment of the uncertainty linked with mean GCMsprojected changes which represent a larger contribution to climateprojection uncertainties compared to the downscaling methods(Boé et al., 2009; Chiew et al., 2009). The hydro-indicators analyzedin this study are mostly linked with snow accumulation. The per-turbation method is suitable to estimate these indices becausethe preservation of the frequency distribution of the precipitationdata has only a small impact for snow accumulation. Also, the sta-tistical analysis of the data will mainly focus on changes in the

imate scenarios.

Resolution (lat � long) SGHG

rganization, Australia 3.2� � 5.6� A2 and B22.8� � 2.8� A2 and B22.25� � 3.75� A2b and B2b


mean values in order to be consistent with the perturbationmethod.

Regional Climate Models were not chosen for two main reasons:(1) only one RCM simulation was available over the region of inter-est at the time of the study thus severely reducing the span of con-ditions needed for this research; (2) the relatively low relief of thestudied regions and its continental position reduce the orographicand maritime effects on precipitation. The overall added value ofRCMs is small for areas weakly influenced by topographical forcing(Denis et al., 2003; Feser, 2006) and concerns mostly the analysis offrequency distributions and of high-order statistics of climatic vari-ables (Laprise, 2008).

As dynamical downscaling, statistical downscaling allowsexamination of changes in the temporal structure of the future cli-mate (e.g. occurrence of precipitation). The statistical downscalingapproach involves the development of statistical relationships be-tween large-scale atmospheric variables and local variables (Wilbyet al., 1998). Many types of statistical techniques can be used (Fow-ler et al., 2007; Giorgi et al., 2001; Wilby et al., 2004) and differentmethods have been compared over Europe (see for example Hay-lock et al., 2006). The statistical relations have to be carefully de-fined for each specific site and require a large high-qualitydataset for calibration and validation (Diaz-Nieto and Wilby,2005). In view of the objectives of our study and of the recommen-dations from Diaz-Nieto and Wilby (2005), the added value of sta-tistical downscaling techniques for a global assessment (analysis ofindices associated with snow accumulation, first phase impactstudy and analysis of changes in mean values) of climate changeimpacts should be small compared to the perturbation method.

Projected climate variable time series: production of the daily seriesIn order to obtain projected climate variable time series, we

have calculated the monthly differences (or anomalies) betweenthe future and the reference periods for temperature and precipita-tion over a region covering at least four grid points on each GCMgrid. The region is delimited by the latitudes 43�N and 49.5�Nand by the longitudes 69�W and 77�W (Fig. 1). The monthly meanregional differences are calculated using spatial averaging for eachgrid point (area-based weight method). Regional differences ob-tained with at least four grid points give more physically represen-tative results than value calculated with the closest grid point(Wilby et al., 2004). The monthly mean regional differences werecalculated for the six selected GCMs projections between the refer-ence period (1961–1990) and three future horizons: 2010–2039,2040–2069 and 2070–2099 hereafter respectively referred to as2020s, 2050s and 2080s.

We have created the projected climate series in three steps: (1)the mean daily precipitation and the mean daily minimum andmaximum temperatures simulated by each GCM were averagedover the region of interest at a monthly scale for the reference per-iod (30 years) and for each future horizon; (2) the anomalies werecomputed for minimum and maximum temperatures (in �C) andfor precipitation (ratio in %) by comparing the future horizons withthe reference period; (3) for each watershed, the anomalies wereadded to the observed daily minimum and maximum temperatureduring the reference period while the ratio was applied for the dai-ly precipitation. The 18 sets (six for each horizon) of projected cli-mate time series generated were integrated into the calibratedhydrological model to produce hydrological scenarios for the fu-ture at a daily time step.

Quality of the hydrological simulations

The quality of the hydrological simulations for the future is sen-sitive to the selected hydrological model, to the calibration of themodel and most importantly to the climate variables projections.

Effect of the hydrological model: context of climate change studiesChartier (2006) has assessed the effect of the hydrological mod-

el for studies of climate change impacts by comparing the results ofthe HSAMI model (averaging water volume over the watershed)with those of the distributed model Hydrotel (Fortin et al., 1995)both calibrated for a sub-basin (10 000 km2) of the Gatineau river(Quebec). The simulation results indicated that the effect of thehydrological model used is less important than the effect attrib-uted to the choice of the climatic model (Chartier, 2006). These re-sults support other studies showing that the choice of ahydrological model has a relatively minor impact on the resultsof hydrological simulations based on climatic projections (Bateset al., 2008; Kay et al., 2006). However, the hydrological modelmay have a greater impact in contrasted watershed (e.g. alpineforeland watersheds) as shown by Ludwig et al. (2009) for theSouthern Bavaria region in Germany.

Effect of calibration: observed vs. simulated hydrographs at a seasonaltime scale

Using the optimal combination of parameters, the calibratedHSAMI model generally performs well to reproduce the observedaverage annual runoff and the spring peak for the reference period(Fig. 2). However, the quality of the simulation is variable at amonthly and seasonal scale. The comparison between the simu-lated mean seasonal discharges with the observed data show thatthe simulated mean winter and spring discharges are lower thanthe observed discharges for most rivers except for the Richelieuand LaGabelle (winter only) (Table 3). The correlation between ob-served and simulated discharges are generally lower for January(r = 0.4–0.8). The simulated mean summer discharges are higherthan the observed values while the simulated mean fall dischargesare slightly higher than the observed ones for the Richelieu, Bati-scan and LaGabelle rivers and lower for the St-François andYamachiche rivers. The difference between mean seasonal simu-lated and observed discharges are generally lower than 35% exceptfor the mean winter discharge on the Yamachiche river, wheresimulated discharges are much lower than the observed discharges(77%), and for the mean summer discharge of the Richelieu wheresimulated discharges are higher than the observed discharges(54%). For the Yamachiche, the larger differences during wintermay arise from the overestimation of observed winter dischargeusing the drainage area-ratio method. The over- or underestima-tion in seasonal (and monthly) discharges will also be present inthe projected hydrological simulations for the next century. Thesepercentages represent uncertainty zones under which interpreta-tion of future changes will be hazardous.

Effect of uncertainty in climate variable data setsUncertainty in modeling precipitation, evapotranspiration and

climate variability represents a large part of the error in hydrolog-ical projections in climate change studies (Bates et al., 2008).Therefore, changes in modeled river hydrology using projected cli-mate data sets have to be carefully appraised (Kay et al., 2006).

Compared to the observations during the reference period, Had-CM3 and ECHAM4 offer a better potential for the simulation of thevariables of interest in Southern Quebec. CSIRO-Mk2 shows anoverestimation of precipitation especially during summer (datanot shown). Although the quality in reproducing the present dayclimate does not necessarily imply a more accurate simulation offuture climate change, it is expected from these results that theadequacy of the models to response to climate forcing is higher(Giorgi et al., 2001).

The trend in precipitation change for the next century whencompared to the reference period shows a great seasonal variabil-ity (Fig. 3). For the winter, all simulations project an increase inprecipitation; for the spring, HadCM3 and CSIRO-Mk2 models

0.050.0

100.0150.0200.0250.0300.0350.0400.0450.0

0.0

100.0

200.0

300.0

400.0

500.0

600.0

700.0

800.0

1 51 101 151 201 251 301 351

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

Dis

char

ge (m

3 /s)

0.010.020.030.040.050.060.070.080.090.0

0.0100.0200.0300.0400.0500.0600.0700.0800.0900.0

Day of the year

(a) (b)

(d)(c)

(e)

Dis

char

ge (m

3 /s)

Dis

char

ge (m

3 /s)

Dis

char

ge (m

3 /s)

Dis

char

ge (m

3 /s)

1 51 101 151 201 251 301 351Day of the year

1 51 101 151 201 251 301 351Day of the year

1 51 101 151 201 251 301 351Day of the year

1 51 101 151 201 251 301 351Day of the year

Fig. 2. Mean hydrograph for observed and simulated data for the reference period. (a) St-François; (b) Richelieu; (c) Batiscan; (d) Yamachiche; (e) St-Maurice. The blue linesrepresent the simulated results and the red lines the observed data. (For interpretation of the references to colour in this figure legend, the reader is referred to the webversion of this article.)

Table 3Comparison of seasonal mean discharges observed and simulated for the reference period (1961–1990).

River Qobs Qsim Qsim�QobsQobs

� �� 100 Qobs Qsim Qsim�Qobs

Qobs

� �� 100

m3/s m3/s % m3/s m3/s %

Winter SpringRichelieu 304.34 363.53 19.45 680.09 695.39 2.25Batiscan 38.96 36.73 �5.72 178.85 171.43 �4.15St-Françoisa 164.72 126.92 �22.95 357.94 344.83 �3.66Yamachiche 3.26 0.76 �76.66 15.41 13.40 �13.05St-Maurice (LaGabelle sub-basin) 4.33 4.75 9.81 29.11 28.79 �1.10

Summer AutumnRichelieu 237.14 364.43 53.68 283.62 321.57 13.38Batiscan 70.72 91.96 30.03 80.44 84.50 5.05St-Françoisa 120.84 144.50 19.57 190.43 166.00 �12.83Yamachiche 3.51 3.86 10.05 5.84 4.59 �21.47St-Maurice (LaGabelle sub-basin) 6.79 9.79 44.23 8.98 9.34 4.01

a Period: 1974–1990.


anticipate an increase in precipitation while ECHAM4 (A2 and B2)curves project only small changes; for the summer and fall, smallprecipitation changes are projected.

The temperature trend is clearer. All simulations project an in-crease in seasonal maximum and minimum temperatures com-pared to the reference period (Fig. 3). Discrepancies between

Winter (DJF)

Spring (MAM)

Summer (JJA)

Fall (SON)

Δ Tmax (ºC) ΔTmin (ºC)

-10

0

10

20

30

40

1980 2020 2060

-10

0

10

20

30

40

1980 2020 2060

-10

0

10

20

30

40

1980 2020 2060

-10

0

10

20

30

40

1980 2020 2060

0

2

4

6

8

10

1980 2020 2060

0

2

4

6

8

10

1980 2020 2060

0

2

4

6

8

10

1980 2020 2060

0

2

4

6

8

10

1980 2020 2060

0

2

4

6

8

10

1980 2020 2060

0

2

4

6

8

10

1980 2020 2060

0

2

4

6

8

10

1980 2020 2060

0

2

4

6

8

10

1980 2020 2060

Δ P (%)

HadCM3 A2bHadCM3 B2bCSIRO -Mk2A2CSIRO-Mk2 B2ECHAM4 A2ECHAM4 B2

Fig. 3. Seasonal projected changes in temperatures (minimum and maximum) and precipitations over the studied region for six climate projections. The changes arecalculated compared to the reference 1961–1990. The 30 years moving average is represented.


simulations are larger for the winter and spring seasons mainly forthe last period indicating that uncertainties in the projectedchanges are slightly more important for this period. For the sum-mer and fall, the difference between models stays constantthrough time.

For all seasons, most of the simulations used in this study areincluded within the 95% probability density surface limits definedfor temperature and precipitation changes (Fig. 4). These limitswere obtained by the more recent GCM simulations published inphase with the IPCC Fourth Assessment Report (Meehl et al.,2007). The CSIRO-Mk2 is occasionally outside this limit. Thechange projected during spring obtained from this model is alwayshigher than the 95% density probability because of its high temper-ature change projections.

For northern watersheds like those studied here, the state ofprecipitation during late fall and winter is also a source of uncer-tainty for hydrological projected data. Therefore, the frequency ofrain events during winter is highly sensitive to both errors in tem-perature and precipitation projections.

The downscaling method is also a source of uncertainties.However, some studies have found that the uncertainties relatedto the downscaling and bias correction methods is lower thanthose related to the GCMs (Boé et al., 2009; Seguí et al., 2009).In this study, the uncertainty due to the downscaling methodwas not studied as we have only applied the perturbation meth-od. Despite the limitations of this technique, studies have shown

that the monthly means runoff or discharge given by the pertur-bation method are generally comparable (amplitude and direc-tion of change) with other methods like the quantile mapping,weather typing and the daily scaling (Boé et al., 2009; Chiewet al., 2009; Maurer and Hidalgo, 2008; Segui et al., 2009).

Analysis

From the hydrological data (simulated for the reference andfuture periods), we have calculated the center-volume date forthe winter/spring period (WS) for each year. The WS center-vol-ume date (WS CV date) corresponds to the date at which half ofthe total water volume cumulated over the January 1 to May 31period is reached. This approach is more robust than the identi-fication of peak discharge which can be associated with a singleflood occurring before or after the bulk of high flows (Hodgkinset al., 2003). It is also expected to be less sensitive to the modelingtechnique used to simulate the daily discharges. The center-volumedate is closely linked with the discharge temporal distributionduring the period of interest (Court, 1962). Since this value is fre-quently used in other studies, comparison of the results will befacilitated.

We have analyzed the long term trend of the annual WS CVdate and WS mean temperature (WS Tmean) and WS snow/pre-cipitation ratio (WS S/P) only for historical data. The slope of the

Winter_2010-2039 Winter_2040-2069 Winter_2070-2099

Summer_2010-2039 Summer_2040-2069 Summer_2070-2099

Fall_2010-2039 Fall_2040-2069 Fall_2070-2099

Temperature (°C) Temperature (°C)Temperature (°C)

P/P

(%) 40

200

-20-40

40200

-20-40

40200

-20-40

0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10

P/P

(%) 40

200

-20-40

40200

-20-40

40200

-20-40

0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10

P/P

(%) 40

200

-20-40

40200

-20-40

40200

-20-40

0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10

Spring_2010-2039 Spring_2040-2069 Spring_2070-2099

P/P

(%) 40

200

-20-40

40200

-20-40

40200

-20-40

0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10

Fig. 4. Projected seasonal changes in temperature and precipitation in the region of interest. The results of 130 projections are represented in gray. The inner to the outer blueellipses represent the surface with a probability density of 50%, 75% and 95% respectively; the surface is derived from the covariance matrix. The projected changes forHadCM3 (blue), CSIRO-Mk2 (red) and ECHAM4 (green) used in this study are indicated, circles and squares are respectively for A2 and B2 GHG emissions scenarios.


temporal trend was calculated with the Sen’s slope non-para-metric estimator. The Kendall-Tau non-parametric test was usedto evaluate the significance of the trend. The Sen’s slope estima-tor and the Kendall-Tau test are not sensitive to the presence ofoutliers and do not require any particular assumption regardingthe data distribution (Kendall, 1938,1975; Sen, 1968). We alsoemployed the Kendall-Tau correlation coefficient (non-paramet-ric correlation coefficient) to examine the correlation betweenclimatic variables and the WS CV date. Trend analysis was notperformed for the projected climatic data since the time seriesinside each 30 year periods replicated the temporal variabilityof the reference period.

For the graphical representation of the series, we have drawnthe ‘‘locally-weighted scatterplot smoothing” (lowess technique)curve. It is a robust locally weighted polynomial regression modelfor which outliers have less influence (Cleveland and Devlin, 1988).The smoothing window used in this study is 10 years. This windowsize approximates the temporal frequency of large-scale oscilla-tions (NOA) that have an impact on winter temperature and pre-cipitation and as a result on river hydrology. The central point ofthe window has the largest weighting, and the points towardsthe edge of the window have successively less influence on thefit. The smoothed line follows peaks and troughs in the originaldata series.

In the presentation of the results, we frequently used the meanvalue of the six sets of simulations with an indicator of the variabil-ity between these simulations. This choice simplifies the visualpresentation of the data and is justified by the fact that all simula-tions generally point in the same direction despite the variability inthe amplitude of the predicted changes. It also reduces the uncer-tainty associated with the projection obtained from one specificsimulation.

Results

Impact of climate changes on mean temperature, snow/precipitationratio and total precipitation during the winter/spring period

WS mean temperatureThe temporal variation of the WS mean temperature (WS

Tmean) for each tributary is analyzed to identify the decade atwhich the 0 �C threshold is crossed. Fig. 5 shows that the WSTmean crossed the 0 �C threshold at different decades dependingon the latitude of the river. For the reference period, the 0 �Cthreshold is crossed in the mid-1970 (lowess curve) for the Riche-lieu which is the southernmost watershed. During that period, theSt-François is at ffi�2 �C. For the St. Lawrence north shore tributar-ies, the WS Tmean is under �2 �C for all the reference period. Thetrend in the WS Tmean is significant for the St-François and Riche-lieu rivers (Kendall-Tau coefficient = 0.35 and 0.33; p < 0.01).Theaverage increase in temperatures between 1970 and 1980 is con-sistent with observations made for the North-American East coast-al region (Hayhoe et al., 2007; Huntington et al., 2004).

For the future, all models projected that WS Tmean for theRichelieu will be above 0 �C. For the other tributaries, the time atwhich this threshold will be crossed differs. For the St-François,WS Tmean will cross the threshold of 0 �C during the first horizon.For the Yamachiche, St-Maurice and the Batiscan the threshold willbe crossed during the second horizon. For the last period, WSTmean will be over 2 �C, for all five watersheds. The difference be-tween simulations is larger for this period.

WS snow/precipitation ratioHuntington et al. (2004) have shown that the increase in the

annual Tmean coincides with a reduction of the ratio snow/total

- 10.0- 8.0 - 6.0- 4.0- 2.0

0.02.04.06.08.0

10.0

1950 1980 2010 2040 2070 2100

WS

Tmea

n (°

C)

(c)

- 10.0- 8.0-6.0- 4.0- 2.0

0.02.04.06.08.0

10.0

1950 1980 2010 2040 2070 2100

WS

Tmea

n (°

C)

(a)

-10.0- 8.0

- 4.0- 2.00.02.04.06.08.0

10.0

1950 1980 2010 2040 2070 2100

(d)

-10.0- 8.0- 6.0- 4.0- 2.0

0.02.04.06.08.0

10.0

1950 1980 2010 2040 2070 2100

(b)

- 10.0- 8.0- 6.0- 4.0- 2.0

0.02.04.06.08.0

10.0

1950 1980 2010 2040 2070 2100

WS

Tmea

n (°

C)

(e)

- 6.0

Fig. 5. Variations of the WS mean temperature. (a) St-François; (b) Richelieu; (c) Batiscan; (d) Yamachiche; (e) St-Maurice. For the reference period and the three futureshorizons, the dots represent the mean of all models. The robust lowess curve (with a 10 year smoothing) is represented with the red line. The variability between simulationsis indicated by the error type dash line (2r). WS Tmean of each 30 years period is represented by the black line. (For interpretation of the references to colour in this figurelegend, the reader is referred to the web version of this article.)


annual precipitation (S/Pan) for half of the climatic stations studiedin New-England (USA) for the period from 1948 to 2000. The sta-tions located in the northern part of that region showed the stron-gest trend with a decrease of the S/Pan ratio (�7%). For the St.Lawrence tributaries, Fig. 6 shows that the increase in WS Tmeanfrom 1970 to 1980 was associated with a marked and rapid de-crease in the WS S/P ratio calculated for the WS period only. Before1980, the WS S/P ratio was about 50% except for the Richelieuwhere this ratio was close to 35%. From 1980 to 1999, this ratiowas around 30% for the Richelieu and 40% for the other tributaries.The decrease detected at the beginning of the 1980s is less clear forthe Richelieu which has a higher mean temperature than othertributaries. The period from 1970 to 2001 is very likely influencedby high winter NAO indices (positive anomalies).

For the projected series, the WS S/P ratio continues to decrease(Fig. 6). The overall mean annual decrease of the ratio is between0.15 to 0.2% per year. For the first time horizon (2020s), WS S/Pis approximately 40% for the St-François, St-Maurice, Yamachicheand Batiscan rivers. For the Richelieu, WS S/P is close to 25% duringthe course of the first horizon. When compared to the referenceperiod, the projected decrease in WS S/P is of 5% and of 10% forthe north shore and south shore tributaries respectively. For thesecond horizon, WS S/P decreases by about 4–5% compared tothe previous horizon. During the last horizon, the rate of decreasewill be slightly higher (5 to 6%). The WS S/P ratio is around 30% forthe north shore rivers (a value similar to the current value ob-

served on the Richelieu), 25% for the St-François and below 15%for the south Richelieu. The differences between the first horizonand the reference period and between each horizons are significantfor all rivers (p < 0.01). As in the case of WS Tmean, the variabilitybetween models is higher during the last horizon.

There is a clear latitudinal gradient in the natural variability ob-served in the WS S/P ratio during the reference period. The inter-annual variability in the WS S/P ratio is larger for the north shorerivers that have lower WS Tmean. This variability is expected todecrease in response to the projected increase in temperature dur-ing the winter/spring period. However, this change cannot be fullycaptured with the simulation technique used (perturbation meth-od) which preserves the inter-annual variability of temperaturesand precipitation amounts.

WS total precipitationTotal precipitation during the winter/spring period (WS Ptotal)

is characterized by substantial inter-annual variability (data notshown). For the reference period 1961–1990, the maximum pre-cipitation is registered in 1974–1975 and the minimum in 1964–1966 and after 1980. Although WS total precipitation is higherfor the Richelieu and the St-François, the inter-annual variabilityis similar for all rivers. Discrepancy between the WS total precipi-tation observed in south and north shore watersheds is largestduring the low precipitation period. Partly due to the high inter-annual variability, historical data for each river do not show any

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

1950 1980 2010 2040 2070 2100

WS

S/P

ratio

(%)

(c)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

1950 1980 2010 2040 2070 2100

WS

S/P

ratio

(%)

(a)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

1950 1980 2010 2040 2070 2100

(d)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

1950 1980 2010 2040 2070 2100

WS

S/P

ratio

(%)

(e)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

1950 1980 2010 2040 2070 2100

(b)

Fig. 6. Variations of the WS S/P ratio. (a) St-François; (b) Richelieu; (c) Batiscan; (d) Yamachiche; (e) St-Maurice. For the reference period and the three futures horizons, thedots represent the mean of all models. The robust lowess curve (with a 10 year smoothing) is represented with the red line. The variability between simulations is indicatedby the error type dash line (2r). The mean WS S/P ratio of each 30 years period is represented by the black line. (For interpretation of the references to colour in this figurelegend, the reader is referred to the web version of this article.)

Table 4Mean annual WS total precipitation over the reference period and the three simulated horizons.

All models Batiscan (cm) Yamachiche (cm) St-Maurice(LaGabelle) (cm)

St-François Richelieu (cm)

Reference period Mean 36.43 37.35 36.51 39.93 47.39Minimum 9.61 9.53 9.59 7.45 8.77Maximum 62.98 70.74 73.19 56.68 65.53

2010–2039 Mean 40.43 41.40 40.92 44.41 52.99Minimum 27.45 27.10 28.18 26.34 29.87Maximum 68.57 77.02 80.18 62.78 72.52




temporal trend (Kendall-Tau, p > 0.05). For the three simulatedhorizons, the annual WS total precipitation generally increases

from one horizon to the other (Table 4). The difference betweenthe mean of each horizon is however not significant, (p > 0.05).


The small changes in WS total precipitation suggest that varia-tions in the WS S/P ratio are mainly explained by modification ofthe WS Tmean. The average rate of decrease of the WS S/P ratiowith increasing temperature for the reference period and for thethree future horizons is �5.2% per 1 �C. WS S/P ratios are projectedto be lower than 25–30% for the Richelieu during all time horizonsand during the last horizon for the other rivers. This suggests thatthe potential for stocking water into the snow cover will be re-duced, leading to important consequences for the winter andspring discharges.

Hydrologic impacts of climate change hydrological scenarios

Mean annual dischargeFor all five rivers and for all horizons, changes in mean annual

discharge when compared to the reference period are generallylower than 15%. As noted in previous studies, there is a large var-iability among the results from the different climate models dueto the difference in the temperature and precipitation projections(Fig. 7). An increase in mean annual discharge is projected by themodel HadCM3 (+6.6–+17.7%) and by the CSIRO-Mk2 model(�1% to + 10 %). Conversely, the ECHAM4 model is projecting a de-crease in the mean annual discharge (�4% to �12%). The decreasein annual water volume estimated by this model is a result of theprojected low changes in total precipitation associated with a mod-erate increase in mean temperature (Fig. 3). These conditions con-tributed to an increase of water loss by evapotranspiration. All theprojected values for changes in the mean annual discharge are

-15.00

-10.00

-5.00

0.00

5.00

10.00

15.00

20.00

A2b B2b

Wei

ghte

d di

ffere

nce

(%)

-15.00

-10.00

-5.00

0.00

5.00

10.00

15.00

20.00

Wei

ghte

d di

ffere

nce

(%)

-15.00

-10.00

-5.00

0.00

5.00

10.00

15.00

20.00

Wei

ghte

d di

ffere

nce

(%)

2070 - 2099 BatiscanYamachicheRichelieuSt-FrançoisSt-Maurice

2040 - 2069

2010 - 2039

HadCM3 HadCM3 CSIRO-Mk2 A2

ECHAM4 A2

mean all simulations

CSIRO-Mk2 B2

ECHAM4 B2

CSIRO-Mk2 A2

ECHAM4 A2

mean all simulationsCSIRO

-Mk2 B2

ECHAM4 B2

HadCM3B2b

ECHAM4 A2

mean all simulationsCSIRO

-Mk2 B2

ECHAM4 B2

CSIRO-Mk2 A2

HadCM3A2b

HadCM3B2b

HadCM3A2b

Fig. 7. Relative mean annual discharges of each horizon in relation to the referenceperiod (weighted difference (%) = (Qmean annual future � Qmean annual reference period)/Qmean annual reference period � 100).

within the uncertainty zone (±19%) associated with HSAMI calibra-tion process.

Mean seasonal and monthly dischargeUnlike the mean annual discharge, changes in the temporal dis-

tribution of discharge (monthly and seasonal) can be important. Asexpected from the results of other studies in Québec (Minvilleet al., 2008; Quilbé et al., 2008), the analysis of the simulated sea-sonal discharges indicates that mean winter and mean spring dis-charges are the most altered by changes in climatic variables.Results show that all models projected an increase in winter dis-charge compared to the reference period (Fig. 8). Except for theYamachiche, the mean winter discharge will increase by an aver-age of 52 % (ranging from 42% to 70%, which represent the upperand lower quartile) for the 2020s and by 133% (ranging from 92%to 163%) for 2080s. These differences are statistically significant(p < 0.01). The projected increase for the Yamachiche is larger thanfor the other rivers. However, in this case, the value of the simu-lated discharge (0.76 m3/s) for the reference period was fairly low-er than the observed value (3.26 m3/s). The underestimation forthis river impinges on the analysis of the change projected forthe future horizons. It is postulated that the projected increase inmean winter discharges are slightly underestimated for all tribu-taries except for the Richelieu. This is due to a negative bias inmean winter discharge during calibration process (Table 3).

For the spring season, the overall mean projected change is a de-crease of 8% (ranging from�18% to + 8%) for 2020s and of 26% (rang-ing from �16% to �40 %) for 2080s. Unlike the other models, theHadCM3 simulations projected a significant increase (16%) of thespring discharge for the Batiscan, Yamachiche, LaGabelle and Riche-lieu during the first horizon. The ECHAM4 and CSIRO-Mk2 projectthat the mean discharges will decrease by 14% (ranging from �19%to�7%) and by 33% for 2020s and 2080s, respectively. The projecteddecrease in mean spring discharge may be slightly overestimated forall tributaries except the Richelieu due to the negative bias in thespring mean discharge during the calibration (Table 3). For the firsthorizon, the projected changes for the Batiscan, Yamachiche andLaGabelle watersheds are within the uncertainty zone.

The projected reduction of the mean discharge is lower in thesummer than that expected for the spring. The mean reduction isunder 20 % for the three time horizons. During the two first hori-zons, some simulations (HadCM3 A2b and B2b and CSIRO-Mk2B2) showed an increase in the summer discharge. The ECHAM4(A2 and B2) model projected a significant decrease (15–30%) forall rivers and horizons. For the fall, the trend is variable betweenrivers and horizons. Changes in mean discharge are expected tobe lower than ±20%.

The reduction of the differences in discharge magnitude be-tween winter and spring for south shore rivers (Richelieu and St-François) mainly occur during the first horizon and will be drivenby an increase in the winter months discharges (January, Februaryand March) and a decrease in the May discharge (Fig. 8b). For thesecond horizon, mean monthly discharges decrease during Apriland May and continue to increase in January and February. Marchmean monthly discharges change slightly compared to the horizon2020s. For the last horizon, mean January and February dischargescontinue to increase and discharges from March to May do notchange compared to the previous horizon.

Changes for the north shore rivers (LaGabelle, Batiscan andYamachiche) appear to occur gradually throughout the three hori-zons (Fig. 8c–e). The first modifications will be that the meanmonthly discharges will increase in late winter and will decreasein May. For the second horizon, this trend is accentuated. Dis-charges from January to March are shown to increase. For the lasthorizon, mean monthly discharges continue to increase in January

0.0

100.0

200.0

300.0

400.0

500.0600.0

- 50.0

0.0

50.0

100.0

150.0

200.0

250.0

1 2 3 4 5 6 7 8 9 10 11 12Wei

ghte

d di

ffere

nce

(%) 2010-2039

0.0

100.0

200.0

300.0

400.0

500.0600.0

- 50.0

0.0

50.0

100.0

150.0

200.0250.0

1 2 3 4 5 6 7 8 9 10 11 12Wei

ghte

d di

ffere

nce

(%) 2040-2069

0.0

100.0

200.0

300.0

400.0

500.0600.0

- 50.0

0.0

50.0

100.0

150.0

200.0

250.0

1 2 3 4 5 6 7 8 9 10 11 12 Mea

n di

scha

rge

(m3 /

s)

Wei

ghte

d di

ffere

nce

(%) 2070-2099

Mea

n di

scha

rge

(m3 /

s)M

ean

disc

harg

e (m

3 /s)

0.0

200.0

400.0

600.0

800.0

1000.0

- 50.0

0.0

50.0

100.0

1 2 3 4 5 6 7 8 9 10 11 12

Wei

ghte

d di

ffere

nce

(%) 2010-2039

0.0

200.0

400.0

600.0

800.0

1000.0

- 50.0

0.0

50.0

100.0

1 2 3 4 5 6 7 8 9 10 11 12

Wei

ghte

d di

ffere

nce

(%) 2040-2069

0.0

200.0

400.0

600.0

800.0

1000.0

- 50.0

0.0

50.0

100.0

1 2 3 4 5 6 7 8 9 10 11 12

Wei

ghte

d di

ffere

nce

(%) 2070-2099

Mea

n di

scha

rge

(m3 /

s)M

ean

disc

harg

e (m

3 /s)

Mea

n di

scha

rge

(m3 /

s)

(a) (b)

- 60.00.0

60.0120.0180.0240.0300.0

1 2 3 4 5 6 7 8 9 10 11 12

Mea

n di

scha

rge

(m3 /

s)W

eigh

ted

diffe

renc

e (%

)

- 60.00.060.0120.0180.0240.0300.0

1 2 3 4 5 6 7 8 9 10 11 12Wei

ghte

d di

ffere

nce

(%)

- 60.00.0

60.0120.0180.0240.0300.0

1 2 3 4 5 6 7 8 9 10 11 12

Wei

ghte

d di

ffere

nce

(%)

Mea

n di

scha

rge

(m3 /

s)M

ean

disc

harg

e (m

3 /s)

0.0

5.0

10.0

15.0

20.0

25.0

- 200.00.0

200.0400.0600.0800.0

1000.01200.0

1 2 3 4 5 6 7 8 9 10 11 12Wei

ghte

d di

ffere

nce

(%) 2070-2099

Mea

n di

scha

rge

(m3 /

s)0.0

5.0

10.0

15.0

20.0

25.0

- 200.00.0

200.0400.0600.0800.0

1000.01200.0

1 2 3 4 5 6 7 8 9 10 11 12Wei

ghte

d di

ffere

nce

(%) 2040-2069

Mea

n di

scha

rge

(m3 /

s)

0.0

5.0

10.0

15.0

20.0

25.0

- 200.00.0

200.0400.0600.0800.0

1000.01200.0

1 2 3 4 5 6 7 8 9 10 11 12

Wei

ghte

d di

ffere

nce

(%) 2010-2039

Mea

n di

scha

rge

(m3 /

s)

2070-2099

2040-2069

2010-2039

(c) (d)

weighted differenceCSIRO-Mk2 A2

ECHAM4 A2HadCM3 A2b

CSIRO-Mk2 B2ECHAM4 B2

HadCM3 B2b1961-1990

Fig. 8. Mean monthly discharges (mean of each horizon) and relative mean monthly discharges in relation to the reference period. (a) St-François; (b) Richelieu; (c) Batiscan,(d) Yamachiche and (e) LaGabelle (LaGabelle).


and February but changes in March and May are less marked thanwhat is anticipated for the first horizon.

Winter/spring center-volume dateProjected changes in the temporal distribution of discharges

indicate that the spring flood associated with snow melt occursat an earlier date in the future. We have used the winter/springcenter-volume date (WS CV date) to evaluate the magnitude ofthe shift in the timing of the spring flood. For all tributaries exceptthe Batiscan, the trend toward an earlier WS CV date was already

initiated in the historical period. Observed data from 1932 to2004 show a significant trend (Kendall-Tau, p < 0.1) toward an ear-lier date of the WS CV date (data not shown). A shift of 8–12 dayswas observed during this period. The trend is generally stronger forthe period following 1960. However, for the Batiscan and LaGabellerivers, the trends are not significant (p > 0.1) for the period follow-ing 1960. The WS CV date is earlier for the southern rivers (Riche-lieu and St-François). From 1932 to 2004, two periods when theWS CV date was later than the mean of the period can be distin-guished (shown by the lowess curve). The first period is in the early

Table 6Kendall-Tau correlation coefficients between the WS CV date and climatic variables (2010

Mean all models

2010–2099 Batiscan St-Maurice(LaGabelle)

WS Tmean �0.62 �0.66Tmean March �0.49 �0.56Tmean April �0.55 �0.69WS PTotal �0.03 �0.02WS snow 0.41 0.39Snow March 0.37 0.34WS S/P ratio 0.43 0.38S/P ratio February 0.36 0.23S/P ratio March 0.44 0.40S/P ratio April 0.23 0.20

0.0

10.0

20.0

30.0

40.0

50.0

60.0

-50.00.0

100.0

200.0

300.0

400.0

500.0

1 2 3 4 5 6 7 8 9 10 11 12

Wei

ghte

d di

ffere

nce

(%)

2010-2039

0.010.020.030.040.050.060.070.0

0.0

100.0

200.0

300.0

400.0

500.0

1 92 3 4 5 6 7 8 10 11

Wei

ghte

d di

ffere

nce

(%)

0.010.020.030.040.050.060.070.0

- 50.00.0

100.0

200.0

300.0

400.0

500.0

1 2 3 4 5 6 7 8 9 10 11 12Wei

ghte

d di

ffere

nce

(%)

2040-2069

2070-2099

Mea

n di

scha

rge

(m3 /

s) -50.0 12

(e)

weighted differenceCSIRO-Mk2 A2ECHAM4 A2HadCM3 A2b

CSIRO-Mk2 B2ECHAM4 B2HadCM3 B2b1961-1990

Fig. 8 (continued)

Table 5Change of the WS CV date for the reference period and the three simulated horizons. Dat

All models Batiscan Yamach

Horizon 1 2010–2039 Mean 12–04 15–04Lower quartile 05–04 07–04Upper quartile 21–04 24–04



Reference period Mean 25–04 27–04Lower quartile 24–04 19–04Upper quartile 01–05 28–04

Difference between horizon 1 and reference period (days) �13Difference between horizon 2 and 1 (days) �11Difference between horizon 3 and 2 (days) �10Difference between horizon 3 and reference period (days) �34Slope of the temporal trend 2010–2100 �0.328


1940s and the second period is in the early 1970s. These periodsare known to have been characterized by low winter temperatureswith high total precipitation. For the studied tributaries, low WSTmean and high WS precipitation (snow and total) were observedfor these two periods. After 1950–1960, water regulation on riversfor hydro electricity production may also have contributed to thechange in the WS CV date.

Simulated data for the next century show that the WS CV datewill occur at an earlier date (Table 5). The differences between themean WS CV dates for each horizon are significant for all rivers(Mann–Whitney test, p < 0.05). Transition towards an earlier WSCV date will be more rapid for the rivers located at higher latitudes.For the north shore tributaries, the WS CV date will occur 9–12 days earlier at each horizon (Table 5). For the south shore rivers,the changes will be a shift of 4–8 days per horizon. At the last hori-zon, it is projected that the mean WS CV date will be 22 (Richelieu)to 34 days (Batiscan) earlier than what was observed during thereference period. The temporal shift in the WS CV date is largerfor the Yamachiche (36 days) than for the Batiscan. However,uncertainty of the hydrological simulations for January is high forthe Yamachiche thus leading to potentially larger error in the pro-jected value of WS CV date. These results are consistent with re-sults reported by Minville et al. (2008) who indicated that thetime of peak discharge projected for a basin located in central Que-bec will be about 23 days earlier in 2080 horizon with a large var-iability between models (6–46 days). They are also consistent withresults obtained for the North-American Eastern coastal region(Hayhoe et al., 2007). For this lower latitude region, the peakstreamflow in spring is projected to be in advances of 10–15 daysby the end of the century.

–2099). Significant correlations (p < 0.05) are in bold.

Yamachiche Richelieu St-Francois

�0.66 �0.42 �0.50�0.58 �0.38 �0.40�0.53 �0.38 �0.39�0.04 �0.02 �0.04

0.43 0.29 0.370.37 0.28 0.450.38 0.29 0.390.33 0.33 0.270.46 0.24 0.420.19 0.14 0.15

e is defined as DD-MM.

iche St-Maurice (LaGabelle) St-François Richelieu

20–04 23–03 27–0315–04 15–03 21–0326–04 31–03 31–03

10–04 15–03 21–0304–04 08–03 16–0317–04 23–03 27–03

31–03 08–03 17–0324–03 28–02 12–0310–04 16–03 23–03

01–05 08–04 08–0405–04 31–03 06–0416–04 18–04 19–04

�12 �11 �16 �12�11 �9 �8 �5�12 �11 �7 �4�36 �31 �31 �22�0.368 �0.324 �0.262 �0.161


Relation between WS CV date and climatic variablesAn earlier WS CV date might be caused by an earlier snow melt

and by changes in precipitation amounts or phase (Hodgkins et al.,2003). If an earlier WS CV date is due to early snow melt, this dateshould be related to late winter and early spring temperatures(Hodgkins et al., 2003). If an earlier WS CV date is linked to changesin precipitation, it should be associated with winter and earlyspring Ptotal and S/P ratio and with Tmean. For the northernNew-England region (northwestern Maine and Northern NewHampshire, USA), Hodgkins et al. (2003) have found that changesin the WS CV date in historical data (1900–2000) were mainlylinked with January to April temperature and precipitation. Forthe same region, Huntington et al. (2004) have shown that WSCV date was significantly and positively correlated with the annualor winter S/P ratios.

For the entire simulated period (from 2010 to 2099), the WS CVdate of all tributaries was strongly negatively correlated with theWS Tmean, as one might expect (Table 6). The correlation betweenWS CV date and WS Tmean is stronger for rivers located on thenorth shore of the St. Lawrence. For all tributaries, high correlationcoefficients are observed with March and April temperatures.There is no correlation between WS CV date and WS PTotal. How-ever, correlation coefficients between WS CV date and WS S/P ratioare significant and positive (+) for all tributaries. The correlationcoefficient between WS CV date and March S/P ratio is slightlystronger (+) than for other months for the LaGabelle, Batiscan,Yamachiche and St-François, For the Richelieu, the highest correla-tion value is with the February S/P ratio (data not shown). It isimportant to remember that the WS S/P ratio is negatively corre-lated to both WS Tmean and WS PTotal. The correlation with WSTmean is twice as much as that with WS PTotal for south shorerivers.

The strength of the correlation between climate variables andWS CV date is variable through time (Table 7). For all tributaries,the strength of the correlation with Tmean in April and Marchslightly decreases from the first to the last horizon. For the Bati-scan, St-François and Yamachiche, the relation with April Tmeanis no longer significant (p > 0.1) during the 2080s horizon.

For the north shore tributaries the strength of the correlationbetween the WS CV date and the monthly S/P ratio varies fromone horizon to the other (Table 7). For the first horizon, correla-tions are stronger for March and April than for the other months.For the other two horizons, the correlation coefficient is strongerfor February and March. The correlations for the south shore trib-

Table 7Significant correlations (p < 0.05 or p < 0.1 (�)) between WS CV date and monthly valuesimulated horizons. Non-significant correlation with any studied climate variables is indic

Horizon Batiscan St-Maurice (LaGabelle) Yamac

2010–2039 S/P : March (0.36) andApril (0.38)

S/P : March(0.30) andApril (0.26)

S/P : MApril (

P : NaN P : February� (0.23) andMarch� (0.22)

P : Feb

Tmean : March (�0.40)and April (�0.52)


Tmeanand A

2040–2069 S/P : February (0.33)and March (0.52)

S/P : March (0.43) S/P : Fand M

P : NaN P : February� (0.24) P : FebTmean : March (�0.34)and April (�0.34)


Tmeanand A

2070–2099 S/P : February (0.29)and March (0.45)

S/P : February (0.26)and March (0.39)

S/P : Fand M

P : February� (0.24) P : February� (0.24) P : Feb

Tmean : March (�0.28) Tmean :March (�0.43)and April (-0.40)

Tmean

utaries are always stronger in late winter months but are generallyweaker than for the north shore tributaries. For the Richelieu andSt-François, the WS CV date is positively correlated (p < 0.05) withthe March precipitation, which mainly consists of rain, for the hori-zon 2080s. For this horizon, a weak negative correlation (p < 0.1)with precipitation is also observed in January for the Richelieu.For the north shore tributaries, the positive correlation betweenWS CV date and total precipitation (snow) in February becomesstronger for the last horizon.

We have plotted WS CV date in relation with WS Tmean (resultsfrom all models) for the five tributaries and for the reference andsimulated periods (Fig. 9). This shows that WS CV date decreasesat a mean rate of 6 days with an increase of 1 �C of WS Tmeanbut when WS Tmean is above 4 �C, variations in the WS CV dateis no longer related to changes in WS Tmean. For a temperatureof 4 �C, the WS S/P ratio is around 20%. For these conditions, themean value of WS CV date is March 14 with a standard deviationof ±12 days. Results from the St-François frequently show lowervalues of WS CV date contributing a large part of the variability ob-served in the WS CV dates. The threshold of WS Tmean = 4 �C willbe reached during the last horizon only for the south shore rivers.For the north shore rivers, it is expected that WS CV date will con-tinue to be affected by the temperature increase through its effecton the S/P ratio during the winter/spring period after 2100.

Discussion

Effect of latitude on hydrological change

All simulations give a consistent picture of changes where thelatitude gradient of WS Tmean is likely the primary cause ofchanges in the WS S/P ratio, winter and spring snowmelt runoffand the WS CV dates in the studied area. The timing and durationof the maximum change period in the hydrological regime are alsoa function of the latitude of the watershed. Southern tributarieswill move quickly towards a new rain-based hydrological regimegiven that mean air temperatures are already relatively high forthese watersheds (�1.4 �C for the StFrancois and +0.2 �C for theRichelieu for 1990–1999). A strong link between WS CV datesand mean air temperatures was observed in historical data(1942–2000) for rivers in the North-Eastern of USA and in the Wes-tern part of North America (Hodgkins and Dudley, 2006; Stewartet al., 2004). The magnitude of the change towards an earlier date

s of climatic variables (temperature, total precipitation and S/P ratio) for the threeated by NaN.

hiche St-François Richelieu

arch (0.35) and0.26)

S/P : February� (0.24) andMarch (0.44)

S/P : March (0.24)

ruary� (0.24) P : NaN P : March (0.26)

: March (�0.42)pril (�0.60)



ebruary (0.28)arch (0.54)

S/P : March (0.38) S/P : NaN

ruary� (0.22) P : NaN P : March* (0.30): March (�0.39)

pril (�0.37)Tmean : March� (�0.23)and April� (�0.23)


ebruary (0.32)arch (0.53)

S/P : March� (0.23) S/P : February� (0.23)

ruary (0.27) P : March (0.25) P : January� (�0.24)and March (0.32)

: March (�0.33) Tmean : March� (�0.23) Tmean : April� (�0.25)

01-01

21-01

10-02

01-03

21-03

10-04

30-04

20-05

-6.0 -4.0 -2.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

WS

CV

date

WS Tmean (°C)

Reference period (1961-1990)RichelieuSt-Maurice

BatiscanSt-FrançoisYamachiche

March 14

Fig. 9. Relation between the WS CV date and WS Tmean for the reference period and 2010–2100 (results obtained with all the models and scenarios). The dash lines indicatedthe standard deviation (1r) around the mean for cases with WS Tmean > 4 �C.


with an increase of 1 �C obtained for many stations is similar thanthe one reported here.

When the WS S/P ratio is lower than 20–25% (corresponding toa value of WS Tmean of 4 �C), the contrast between mean winterand spring discharges is reduced. The WS CV date remains almostconstant when WS Tmean is increasing. During the last horizon, itis expected that both South shore rivers will have a low WS S/P ra-tio (lower than 25%), high mean WS temperature (higher than 4 �C)and a reduced number of days when Tmean below 0 �C. During thishorizon, it is projected that Tmean will be above 0 �C early inMarch (Julian day 60) for the Richelieu. These conditions lengthenthe period of moderate to high discharges. February, March andApril will contribute almost equally to the total WS discharge(20% for each month). For the reference period, 40 % of the totalWS discharge was observed during April for the St-François andduring April (29 %) and May (28%) for the Richelieu. For the Northshore rivers, it is expected that the WS S/P ratio will be over 25 %and WS Tmean lower than 4 �C during the last horizon (2080s). Itis presumed that changes in WS CV date will continue beyondthe period considered in this study (2100). For these rivers, it islikely that most of the streamflow will be observed in March andApril (>50 %) during the last horizon instead of April and May(>50%) as it has been observed during the reference period.

Potential impacts for sediment transport and river geomorphology

Discharge higher than the sediment transport thresholdHydrological changes projected for the St. Lawrence tributaries

can have significant impacts on the frequency and magnitude ofsediment transport processes. It is likely that a reduction of up to32% of the maximum spring mean discharge and of the frequencyof discharges higher than the current mean sediment transportthreshold discharge (discharge at which sediment transport oc-curred at the majority of the cross-sections along a river reach)during this period will reduce the size of particles and the sedi-

Table 8Frequency of days (mean per years) during the winter with discharge higher than the sedmaximum values are extracted from the six sets of simulation.

RichelieuPeriod Winter Freq. days with Q > 450

1961–1990 162010–2039 Mean all simulations (min, max) 41 (32, 47)2040–2069 Mean all simulations (min, max) 53 (46, 59)2070–2099 Mean all simulations (min, max) 59 (51, 66)

ment volumes that can be transported during that season (Boyeret al., 2009). Conversely, higher winter flows will increase the riskof having events that will produce sediment transport with result-ing geomorphological impacts especially when the rivers are cov-ered with ice (for early winter months, Tmean will be under 0 �Cfor all horizons). Although, sediment transport processes are notwell documented under an ice cover, it is generally assumed thatprevailing winter low flow conditions do not have a significant im-pact on river morphology. Such an assumption may need to be re-vised in the future. For all rivers in this study, the frequency of dayswith a discharge higher than the current mean sediment transportthresholds will increase during winter (Table 8). In comparisonwith the reference period, the frequency of days with dischargeshigher than the sediment transport threshold during the wintermay increase by 143 % for the St-François (horizon 2020s) and by1693% for the Batiscan (horizon 2050s). For all tributaries, this fre-quency would, however, remain under the value observed duringthe spring for the reference period. A net increase of this frequencyfor both winter and spring is expected only for the Richelieu (21%,28% and 31% for horizons 2020s–2080s).

Rare flood events during winterThe magnitude and frequency of rare events, defined as dis-

charges larger than three times the standard deviation, are impor-tant for channel stability. Despite the fact that the perturbationmethod used in this study does not allow an exhaustive analysisof the frequency/magnitude of the simulated hydrological data,our results suggest that the magnitude of rare large events duringwinter would increase compared to the reference period. The high-est increase would be observed on rivers located at higher latitudes(e.g. Batiscan). Although some of the simulations projected a max-imum winter discharge higher than the maximum spring dischargeobserved for the reference period, the magnitude of rare eventsduring winter are projected to be lower than for the reference per-iod spring value. Compared to the reference period, the frequency

iment transport threshold for three of the five studied tributaries. The minimum and

St-François Batiscanm3/s Freq. days with Q > 330 m3/s Freq. days with Q > 150 m3/s

7 116 (13, 19) 4 (3, 6)22 (19, 26) 9 (6, 13)29 (24, 35) 16 (11, 26)


of these events during winter would increase and decrease forhigher and lower latitude rivers, respectively. By the 2020s, the fre-quency of rare large events simulated during winter for the Bati-scan would be similar to the frequency of rare events observedfor the St-François during the reference period. For this horizon,the winter mean temperature for the Batiscan watershed is pro-jected to be similar to the winter mean temperature observed dur-ing the reference period in the St-François watershed (winterTmean ffi �8 �C). This result suggests that the projected increasein the frequency of rare large hydrological events for rivers in high-er latitudes is in line with what was observed in the St-Françoiswatershed. This tributary may be used as an analogue to studythe behavior that is likely to occur in winter for rivers like theBatiscan.

Potential impacts for aquatic and riparian ecosystems

Aquatic and riparian ecosystems are spatially and temporallydynamic and are largely shaped by fluvial processes (Hughes,1997; Hupp, 1988; Malanson, 1993; McKenney et al., 1995; Nai-man et al., 1993; Naiman and Décamps, 1997). Flow velocity, waterdepth and hydrological fluctuations (seasonal variability and dura-tion, frequency and timing of floods) play a major role in the struc-ture and diversity of ecosystems. Changes in these variables caninduce complex responses in riparian vegetation and ecosystemdynamics that are not well understood at present (Crowder et al.,1996). Studies have shown that even if the mean annual dischargeis not changed significantly, variation at a finer temporal scale (e.g.monthly) can cause important modifications in riparian vegetation.

Effect of higher winter dischargesHydrological changes projected during winter and early spring

may affect riparian landscapes in the St. Lawrence and its tributar-ies. These changes are expected to be larger than what has beenobserved for the five studied St. Lawrence tributaries between1964 and 1997 (Charron et al., 2008). Although winter flows arerarely analyzed in ecological studies because vegetation is in dor-mancy, their impacts are important for the stability of the riverchannel and for preserving the substrate where seeds are stored.Higher winter flows (increase of 50–200% of the mean winter flow)may increase bed and bank erosion and consequently reduce thesurviving chance of plants and seeds stored into the bank sub-strate. Higher winter flows may also keep plants and seeds underwater for a longer period. This may damage some plants or preventa successful germination of seeds. Changes in river hydrology andsedimentology during the winter may also modify fish habitat.Higher winter flows may affect fish migratory behavior and en-hance the erosion and transport of fish eggs and larvae of winterspawning species (Bergeron et al., 1998; Fortin et al., 1992). Flowvelocity will be a limiting factor for fish that have low swimmingcapacity like the Atlantic Tomcod (Microgadus tomcod) (East andMagnan, 1989), a fish that is spawning into some of the St. Law-rence tributaries (Sainte-Anne and Batiscan rivers) between theend of December and February.

Effect of earlier spring flood and of lower spring dischargesEarlier spring floods (from 22 to 34 days) might be out of phase

with spring spawning species that is controlled by a combinationof photoperiod, temperature and flow. Lower spring flows (downby as much as 40%) will reduce the extent of areas available forspecies that are spawning into the St. Lawrence tributaries duringearly spring, like the walleye (Sander vitreus) and the Northern pike(Esox lucius) for example. Concerning the vegetation, lower waterlevels during the spring are likely to modify plants and seeds dis-persion patterns. Depending of the form of the river cross-section,it can also reduce the areas of the riparian zones affected by annual

flooding. The timing of germination of the species currently foundin these rivers might also be out of phase with high flows which areexpected to occur at an earlier date.

Transformations of the wetlands located at the mouths of theBatiscan, Yamachiche and St-François rivers are expected to bemore important than changes anticipated along riparian corridors.These zones will be influenced by seasonal changes in tributarieshydrology, by modifications of the St. Lawrence water levels, bothchanges may amplify the temporal shift between high river dis-charges and high St. Lawrence water levels (especially after hori-zon 2050s), and by changes in sedimentation and erosiondynamics. Deltas and tributary mouth bars along the St. Lawrenceare important for its biodiversity because of their heterogeneousand contrasting physical characteristics (Desgranges and Jobin,2003). Variable flooding conditions associated with seasonal flood-ing or daily tidal fluctuations create a complex and dynamic mo-saic of wetlands and aquatic habitats. Changes in hydrologicalcharacteristics and sediment dynamics of the tributaries are ex-pected to expand the areas of sediment deposition and to trans-form their structures. Two-dimensional modeling of the flow andsediment dynamics of these zones will need to be conducted in or-der to estimate the magnitude of this morphological development.It is, however, difficult to predict if diversity will increase or de-crease under the projected new conditions.

Potential impacts for water management (hydro electricity)

The projected shift in winter precipitation from snow to rainand its impact on the snowpack accumulation and on winter andspring discharges will influence water management at the wa-tershed scale. For tributaries with dams and reservoirs used for hy-dro electricity, like the St-Maurice and St-François, managers mayhave to adopt new operation strategies to account for higher win-ter discharges and lower spring discharges and for the higheruncertainty in runoff associated with rainfall compared to snow-melt (Gleick and Adams, 2000). Although dam and reservoir man-agers already adapt to the existing variable climatic and hydrologicconditions from one year to the other, global climate change pro-jections add a new level of uncertainty. A better knowledge ofthe projected changes will help to develop adaptation and designstrategies in order to optimize hydroelectric production while min-imizing risks.

Limitations due to hydrological and climate variability

Projected changes in discharge must be interpreted with caredue to the limitations link with the GCMs and with the perturba-tion method used to generate the climate variables. It is possiblethat the larger differences observed between GCMs projectionsduring the spring, and to a lesser extent during the winter, increasethe uncertainty in the estimate of the amplitude of the hydrologi-cal changes for these seasons, particularly for the last horizon. Thedirection of change, however, is clear as indicated by the conver-gence of GCMs projections. The use of the perturbation methodin this study does not allow us to fully appraise the potentialchanges that may occur in the inter-annual variability and in thefrequency of extreme meteorological events in response to climatechanges. Climate variable extremes resulting from this approachare those observed during the reference period enhanced or damp-ened according to the perturbation factors (Graham et al., 2007).Consequently, the method limits our ability to study the variabilityat small temporal scales and restricts the analysis to basic statisti-cal properties of the hydrological data. The limitation due to theassumption of the stationarity of the variance and other momentsis less problematic in this study because we are mainly focusing onhydrologic changes that are linked to the modification of snow


accumulation and snowmelt and consequently on the gradualshifting from a snow hydrological regime to a rain regime. Exten-sive study of changes in the frequency of high magnitude hydro-logical events associated with extreme meteorological events andin the inter-annual variability on hydrological characteristics ofthe watersheds was also limited.

Although it is generally presumed that the variability of climatevariables will be modified during the current century, it remainsuncertain how ongoing climate warming will influence this charac-teristic. An important part of the temporal variability in the wintermean surface air temperature over the northern hemisphere is dueto the NAO (Hurrell, 1996). However, the range of modeling resultsobtained at the moment using various approaches indicates thatsubstantial uncertainty still exists in the projection of the responseof the NAO to increasing GHG concentrations (Hurrell et al., 2006).A complete investigation of the climate variability is then currentlylimited even with the most sophisticated approach (Hurrell et al.,2006; Hurrell and Deser, 2009).

Conclusion

The results from this study clearly indicate that the climatechanges projected for the next century will induce important mod-ifications of the St. Lawrence tributaries hydrological regime with agradual shifting from snow to rain regime. It is projected that bythe end of the century the difference between mean monthly dis-charges will be small during winter and spring. Increase in temper-ature during the winter/spring period will drive most of the changethat are projected in river discharge. Warming of air temperaturein February to April will induce an important decrease in the pro-portion of precipitation falling as snow, with a simultaneous in-crease in winter runoff and a reduction in the volume of waterstored into the snowpack compared to the reference period. Thiswill lengthen the snow melting period and reduce spring runoff.With these changes, the WS center-volume date is expected tobe in advance by 22–34 days depending of the watershed. The lat-itude of the river governs the timing of occurrence of the maxi-mum change (sooner for tributaries located south) and theduration of the period affected by marked changes in the temporaldistribution of discharge (longer time scale for rivers located athigher latitudes).

Higher winter discharges may have a significant effect on rivergeomorphological processes as sediment transport events may oc-cur more frequently under ice-cover conditions than under currentconditions. Conversely, lower spring discharges may promote sed-imentation into the tributary and at their confluence with the St.Lawrence River. The combined effects of modifications in riverhydrology and geomorphological processes will likely impactriparian ecosystems.

The use of multiple climatic models in this study has allowed usto examine plausible scenarios of changes into the hydrological re-gime of the St. Lawrence tributaries. Convergence between themodels and the consistency of the results with other studies in-creases the reliability of the outcome of this study. The strong linkbetween winter-spring center-volume date and winter-springmean temperature and ratio of snow/precipitation (winter-spring)for the St. Lawrence tributaries suggests that the quality of thehydrological projections for northern rivers is highly sensitive totemperature which drives most of the changes in the winter-springratio of snow/precipitation.

The perturbation method used here has given interesting re-sults for an early phase assessment of the impact of climatechanges on the hydrological regime of the St. Lawrence tributariesand in the identification of regions that are expected to be moresensitive. A further step needs to be taken in order to reduce and

quantify the uncertainty associated with the use of GCMs projec-tion at local scale, downscaling methods and hydrological model-ing and to examine more closely the hydrological impact ofpossible change in climate variability. Care should also be takenin future simulations to represent discharge characteristics witha higher level of confidence in order to exhaustively assess changesin the frequency of discharge with active sediment transport (dis-charge higher than sediment transport thresholds), in the magni-tude of large events and in the temporal sequence of discharges.

Acknowledgements

This paper was funded by the Ouranos Consortium and the Nat-ural Science and Engineering Research Council of Canada (NSERC).This research is part of the program of the Canada Research Chairin Fluvial Dynamics. The comments from three anonymous review-ers were very helpful to improve this paper.

References

Andréasson, J., Bergström, S., Carlsson, B., Graham, L.P., Lindström, G., 2004.Hydrological change–climate change impact simulations for Sweden. Ambio: AJournal of the Human Environment 33 (4), 228–234.

Bates, B.C., Kundzewicz, Z.W., Wu, S., Palutikof, J.P. (Eds.), 2008. Climate Change andWater. Technical Paper of the Intergovernmental Panel on Climate Change. IPCCSecretariat, Geneva, 210 pp.

Bergeron, N.E., Roy, A.G., Chaumont, D., Mailhot, Y., Guay, E., 1998. Wintergeomorphological processes in the Sainte-Anne River (Quebec) and theirimpact on the migratory behaviour of Atlantic tomcod (Microgadus tomcod).Regulated Rivers: Research and Management 14, 95–105.

Bisson, J.L., Roberge, F., 1983. Prévision des apports naturels: Expérience d’Hydro-Québec. Compte-rendu de l’Atelier sur la prévision du débit, Toronto, November1983.

Boé, J., Terray, L., Martin, E., Habets, F., 2009. Projected changes in components ofthe hydrological cycle in French river basins during the 21st century. WaterResources Research 45, W08426. doi:10.1029/2008WR007437.

Boyer, C, Verhaar, P.M., Roy, A.G., Biron, P.M., Morin, J., 2009. Impacts ofenvironmental changes on the hydrology and sedimentary processes at theconfluence of St. Lawrence tributaries: potential effects on fluvial ecosystems.Hydrobiologia (Ecosystem studies of the St. Lawrence River). doi: 10.1007/s10750-009-9927-1.

Burakowski, E.A., Wake, C.P., Braswell, B., Brown, D.P., 2008. Trends in wintertimeclimate in the northeastern United States: 1965–2005. Journal GeophysicalResearch 113, D20114. doi: 20110.21029/22008JD009870.

Cayan, D.R., 1996. Interannual climate variability and snow pack in the westernUnited States. Journal of Climate 9, 928–948.

Charron, I., Lalonde, O., Roy, A.G., Boyer, C., Turgeon, S., 2008. Changes in riparianhabitats along major tributaries of the St. Lawrence River, Québec, Canada. Riverregulations and application 24, 617–631.

Chartier, I. (2006). Influence du modèle hydrologique sur la modélisation desapports dans une perspective de changements climatiques : une application surle bassin versant Baskatong. In: Proceedings, Canadian Dam Association AnnualConference, Quebec City.

Chaumont, D., Chartier, I., 2005. Développement de scénarios hydrologiques à desfins de modélisation de la dynamique sédimentaire des tributaires du Saint-Laurent dans un contexte de changements climatiques. 46 pp.

Chiew, F.H.S., Teng, J., Vaze, J., Post, D.A., Perraud, J.M., Kirono, D.G.C., Viney, N.R.,2009. Estimating climate change impact on runoff across southeast Australia:method, results, and implications of the modeling method. Water ResourcesResearch 45, W10414. doi:10.1029/2008WR007338.

Cleveland, W.S., Devlin, S.J., 1988. Locally weighted regression: an approach toregression analysis by local fitting. Journal of the American StatisticalAssociation 83, 596–610.

Cooley, K.R., 1990. Effects of Co2-induced climatic changes on snowpack andstreamflow. Hydrological Sciences Journal 35, 511–522.

Court, A., 1962. Measures of streamflow timing. Journal of Geophysical Research 67,4335–4339.

Croley II, T.E., 2003. Great Lakes Climate Change Hydrologic Impact Assessment: IJCLake Ontario-Saint Lawrence River Regulation Study. NOAA TechnicalMemorandum GLERL-126. NOAA, Great Lakes Environmental ResearchLaboratory, Ann Arbor, MI. 77 pp.

Crossley, J.F., Polcher, J., Cox, P., Gedney, N., Planton, S., 2000. Uncertainties linked toland-surface processes in climate change simulations. Climate Dynamics 16,949–961.

Crowder, A.A., Smol, J.P., Dalrymple, R., Gilbert, R., Mathers, A., Price, J., 1996. Ratesof natural and anthropogenic change in shoreline habitats in the Kingston Basin,Lake Ontario. Canadian Journal Fisheries Aquatic Science 53 (Suppl. 1), 121–135.

http://dx.doi.org/10.1029/2008WR007437

http://dx.doi.org/10.1029/2008WR007338


Denis, B., Laprise, R., Caya, D., 2003. Sensitivity of a regional climate model to thespatial resolution and temporal updating frequency of the lateral boundaryconditions. Climate Dynamics 20, 107–126.

DesGranges, J.-L., Jobin, B., 2003. Knowing, mapping and understanding Saint-Lawrence biodiversity, with special emphasis on bird assemblages.Environmental Monitoring and Assessment 88, 177–192.

Diaz-Nieto, J., Wilby, R.L., 2005. A comparison of statistical downscaling and climatechange factor methods: Impacts on low flows in the River Thames, UnitedKingdom. Climatic Change 69, 245–268.

Duan, Q., Sorooshian, S., Gupta, V.K., 1992. Effective and efficient globaloptimization for conceptual rainfall-runoff models. Water Resources Research28, 1015–1031.

East, P., Magnan, P., 1989. Étude de la vitesse de nage du Poulamon Atlantique(Microgadus tomcod) en relation avec les travaux de la traversée du fleuveSaint-Laurent par la ligne Radisson-Nicolet-Des-Cantons. Université du Québecà Trois-Rivières, Hydro-Québec Environnement. 18 pp.

Feser, F., 2006. Enhanced detectability of added value in limited area model resultsseparated into different spatial scales. Monthly Weather Review 134, 2180–2190.

Fortin, V., 2000. Le modèle météo-apport HSAMI: historique, théorie et application.Rapport de recherche, Institut de recherche d’Hydro-Québec, Varennes.

Fortin, R., Léveillé, M., Guénette, S., Laramée, P., 1992. Contrôle hydrodynamique del’avalaison des oeufs et des larves de poulamon atlantique (Microgadus tomcod)sous le couvert de glace de la rivière Sainte-Anne, Québec. Aquatic LivingResources 5, 127–136.

Fortin, J.P., Moussa, R., Bocquillon, C., Villeneuve, J.P., 1995. HYDROTEL, un modèlehydrologique distribué pouvant bénéficier des données fournies par latélédétection et les systèmes d’information géographique. Revue des sciencesde l’eau 8, 97–124.

Fortin, L.-G., Turcotte, R., Pugin, S., Cyr, J.-F., Picard, F., 2007. Impact deschangements climatiques sur les plans de gestion des lacs Saint-François etAylmer au sud du Québec. Canadian Journal of civil Engineering 34, 934–945.

Fowler, H.J., Blenkinsop, S., Tebaldi, C., 2007. Linking climate change modelling toimpact studies: recent advances in downscaling techniques for hydrologicalmodelling. International Journal of Climatology 27, 1547–1578.

Giorgi, F., Christensen, J., Hulme, M., von Storch, H., Whetton, P., Jones, R., Mearns, L.,Fu, C., Arritt, R., Bates, B., Benestad, R., Boer, G., Buishand, A., Castro, M., Chen, D.,Cramer, W., Crane, R., Crossly, J., Dehn, M., Dethloff, K., Dippner, J., Emori, S.,Francisco, R., Fyfe, J., Gerstengarbe, F., Gutowski, W., Gyalistras, D., Hanssen-Bauer, I., Hantel, M., Hassell, D., Heimann, D., Jack, C., Jacobeit, J., Kato, H., Katz,R., Kauker, F., Knutson, T., Lal, M., Landsea, C., Laprise, R., Leung, L., Lynch, A.,May, W., McGregor, J., Miller, N., Murphy, J., Ribalaygua, J., Rinke, A.,Rummukainen, M., Semazzi, F., Walsh, K., Werner, P., Widmann, M., Wilby, R.,Wild, M., Xue, Y., 2001. Regional climate information-evaluation andprojections. In: Houghton, J.T. et al. (Eds.), Climate change 2001: TheScientific Basis. Contribution of Working Group to the Third AssessmentReport of The Intergovernmental Panel on Climate Change. CambridgeUniversity Press, Cambridge, United Kingdom and New York, USA. 881 pp.

Gleick, P.H., Adams, D.B., 2000. Water: The Potential Consequences of ClimateVariability and Change for the Water Resources of the United States. Report ofthe Water Sector Assessment Team of the National Assessment of the PotentialConsequences of Climate Variability and Change, Pacific Institute for Studies inDevelopment, Environment and Security, Oakland, CA. ISBN #1-893790-04-05.

Gordon, C., Cooper, C., Senior, C.A., Banks, H.T., Gregory, J.M., Johns, T.C., Mitchell,J.F.B., Wood, R.A., 2000. The simulation of SST, sea ice extents and ocean heattransports in a version of the Hadley Centre coupled model without fluxadjustments. Climate Dynamics 16, 147–168.

Graham, L.P., Hagemann, S., Jaun, S., Beniston, M., 2007. On interpretinghydrological change from regional climate models. Climatic Change 81, 97–122.

Hartley, S., Keables, M.J., 1998. Synoptic associations of winter climate and snowfallvariability in New England, USA, 1950–1992. International Journal ofClimatology 18, 281–298.

Hay, L.E., Wilby, R.L., Leavesly, H.H., 2000. Comparison of delta change anddownscaled GCM scenarios for three mountainous basins in the United States.Journal of the American Water Resources Association 36 (2), 387–397.

Hayhoe, K., Wake, C.P., Huntington, T.G., Luo, L., Schwartz, M.D., Sheffield, J., Wood,E., Anderson, B., Bradbury, J., DeGaetano, A., Troy, T.J., Wolfe, D., 2007. Past andfuture changes in climate and hydrological indicators in the US Northeast.Climate Dynamics 28, 381–407.

Haylock, M.R., Cawley, G.C., Harpham, C., Wilby, R.L., Goodess, C.M., 2006.Downscaling heavy precipitation over the United Kingdom: a comparison ofdynamical and statistical methods and their future scenarios. InternationalJournal of Climatology 26, 1397–1415.

Hewitson, B., 2003. Developing perturbations for climate change impactassessments. Eos, Transactions, American Geophysical Union 84, 337–348.

Heywood, I., Cornelius, S., Carver, S., 2006. An Introduction to GeographicalInformation Systems, third ed. Prentice Hall, New Jersey. 464 pp.

Hirst, A.C., Gordon, H.B., O’Farrell, S.P., 1996. Global warming in a coupled climatemodel including oceanic eddy-induced advection. Geophysical Research Letters23, 3361–3364.

Hirst, A.C., O’Farrell, S.P., Gordon, H.B., 1999. Comparison of a coupled ocean-atmosphere model with and without oceanic eddy-induced advection. Part 1:Ocean spin-up and control integrations. Journal of Climate 13, 139–163.

Hodgkins, G.A., Dudley, R.W., 2006. Changes in the timing of winter–springstreamflows in eastern North America, 1913–2002. Geophysical ResearchLetters 33, L06402. doi:10.1029/2005GL025593.

Hodgkins, G.A., Dudley, R.W., Huntington, T.G., 2003. Changes in the timing of highriver flows in New England over the 20th Century. Journal of Hydrology 278,244–252.

Houghton, J. T., Ding, Y., Griggs, D.J., Noguer, M., Van der Linden, P.J., Dai, X., Maskell,K., Johnson, C.A., 2001. Climate Change 2001: The Scientific Basis. Contributiondu Groupe de travail au Troisième rapport d’évaluation du Groupe d’expertsintergouvernemental sur l’évolution du climat Cambridge University Press,Cambridge, Royaume-Uni and New York, NY, ÉU, 881 pp.

Hughes, F.M.R., 1997. Floodplain biogeomorphology. Progress in PhysicalGeography 21, 501–529.

Huntington, T.G., Hodgkins, G.A., Keim, B.D., Dudley, R.W., 2004. Changes in theproportion of precipitation occurring as snow in New England. Journal ofClimate 17, 2626–2636.

Hupp, C.R., 1988. Plant ecological aspects of flood geomorphology and paleofloodhistory. In: Baker, V.R., Kochel, R.C., Patton, P.C. (Eds.), Flood Geomorphology.Wiley, New York, pp. 335–356.

Hurrell, J.W., 1996. Influence of variations in extratropical wintertimeteleconnections on northern hemisphere temperature. Geophysical ResearchLetters 23, 665–668.

Hurrell, J.W., Deser, C., 2009. North Atlantic climate variability: the role of the NorthAtlantic Oscillation. Journal of Marine Systems 78, 28–41.

Hurrell, J.W., Visbeck, M., Busalacchi, A., Clarke, R.A., Delworth, T.L., Dickson, R.R.,Johns, W.E., Koltermann, K.P., Kushnir, Y., Marshall, D., Mauritzen, C., Mccartney,M.S., Piola, A., Reason, C., Reverdin, G., Schott, F., Sutton, R., Wainer, I., Wright,D., 2006. Atlantic climate variability and predictability: a CLIVAR perspective.Journal of Climate 19, 5100–5121.

Jain, S., Lall, U., 2000. Magnitude and timing of annual maximum floods: trends andlarge-scale climatic associations for the Blacksmith Fork River, Utah. WaterResources Research 36, 3641–3651.

Jain, S., Lall, U., 2001. Floods in a changing climate: does the past represent thefuture? Water Resources Research 37, 3193–3205.

Johnson, F., Sharma, A., 2009. Measurement of GCM skill in predicting variablesrelevant for hydroclimatological assessments. Journal of Climate. doi:10.1175/2009JCLI2681.1.

Kay, A.L., Jones, R.G., Reynard, N.S., 2006. RCM rainfall for UK flood frequencyestimation II. Climate change results. Journal of Hydrology 318, 163–172.

Kendall, M.G., 1938. A new measure of rank correlation. Biometrika 30, 81–93.Kendall, M., 1975. Multivariate Analysis. Macmillan Co. III, New York, 210 p.Laprise, R., 2008. Regional climate modeling. Journal of Computational Physics 227,

3641–3666.Ludwig, R., May, I., Turcotte, R., Vescovi, L., Braun, M., Cyr, J.-F., Fortin, L.-G.,

Chaumont, D., Biner, S., Chartier, I., Caya, D., Mauser, W., 2009. The role ofhydrological model complexity and uncertainty in climate change impactassessment. Advances in Geosciences 21, 63–71.

Malanson, G.P., 1993. Riparian Landscapes. Cambridge University Press, Cambridge.296 pp.

Maurer, E.P., Hidalgo, H.G., 2008. Utility of daily vs. Monthly large-scale climatedata: an intercomparison of two statistical downscaling methods. Hydrologyand Earth System Sciences 12, 551–563.

McKenney, R., Jacobson, R.B., Wertheimer, R.C., 1995. Woody vegetation andchannel morphogenesis in low-gradient gravel-bed streams in the OzarkPlateaus, Missouri and Arkansas. Geomorphology 13, 175–198.

Meehl, G.A., Covey, C., Delworth, T., Latif, M., McAvaney, B., Mitchell, J.F.B., Stouffer,R.J., Taylor, K.E., 2007. The WCRP CMIP3 multimodel dataset – a new era inclimate change research. Bulletin of the American Meteorological Society 88,1383–1394.

Merritt, W.S., Alila, Y., Barton, M., Taylor, B., Cohen, S., Neilsen, D., 2006. Hydrologicresponse to scenarios change in sub watersheds of the Okanagan basin, BritishColumbia. Journal of Hydrology 326, 79–108.

Minville, M., Brissette, F., Leconte, R., 2008. Uncertainty of the impact of climatechange on the hydrology of a Nordic watershed. Journal of Hydrology 358, 70–83.

Mortsch, L.D., Quinn, F.H., 1996. Climate change scenarios for Great Lakes Basinecosystems studies. Limnology and oceanography 41, 903–911.

Naiman, R.J., Décamps, H., Pollock, M., 1993. The role of riparian corridors inmaintaining regional biodiversity. Ecological Applications 3, 209–212.

Naiman, R.J., Décamps, H., 1997. The ecology of interfaces – riparian zones. AnnualReview of Ecology and Systematics 28, 621–658.

Nash, J.E., Sutcliffe, J.V., 1970. River flow forecasting through conceptual models.Journal of Hydrology 10, 282–290.

Pope, V., Gallani, M.L., Rowntree, P.R., Stratton, R.A., 2000. The impact of newphysical parameterizations in the Hadley Centre climate model: HadAM3.Climate Dynamics 16, 123–146.

Prudhomme, C., Reynard, N., Crooks, S., 2002. Downscaling of global climate modelsfor flood frequency analysis: Where are we now? Hydrological Processes 16,1137–1150.

Prudhomme, C., Jakob, D., Svensson, C., 2003. Uncertainty and climate changeimpact on the flood regime of small UK catchments. Journal of Hydrology 277,1–23.

Quilbé, R., Rousseau, A.N., Moquet, J.-S., Trinh, N.B., Dibike, Y., Gachon, P., Chaumont,D., 2008. Assessing the Effect of climate change on river flow using generalcirculation models and hydrological modelling–Application to the ChaudièreRiver, Québec, Canada. Canadian Water Resources Journal 33, 73–94.

Roeckner, E., Arpe, K., Bengtsson, L., Brinkop, S., Dümenil, L., Esch, M., Kirk, E.,Lunkeit, F., Ponater, M., Rockel, B., Suasen, R., Schlese, U., Schubert, S.,Windelband, M., 1992. Simulation of the Present-day Climate with the

http://dx.doi.org/10.1029/2005GL025593

http://dx.doi.org/10.1175/2009JCLI2681.1

http://dx.doi.org/10.1175/2009JCLI2681.1


ECHAM4 Model: Impact of Model Physics and Resolution. Max-Planck Institutefor Meteorology, Report No. 93, Hamburg, Germany, 171 pp.

Seguí, P.Q., Ribes, A., Martin, E., Habets, F., Boé, J., 2009. Comparison of threedownscaling methods in simulating the impact of climate change on thehydrology of Mediterranean basins. Journal of Hydrology. doi:10.1016/j.jhydrol.2009.09.050.

Sen, P.K., 1968. Estimates of the regression coefficient based on Kendall’s Tau.Journal American Statistics Association 63, 1379–1389.

Stewart, I.T., Cayan, D.R., Dettinger, M.D., 2004. Changes in snowmelt runoff timingin Western North America under a ‘business as usual’ climate change scenario.Climatic Change 62, 217–232.

St-Hilaire, A., Lachance, M., Bobée, B., Ouarda, T.B.M.J., Gignac, C., Rioux, P.-J.,Gaudet, J., Thiémonge, N., Pion, C., 2003. Assessment of the impact ofmeteorological network density on the estimation of basin precipitation andrunoff: a case study. Hydrological Processes 17, 3561–3580.

Thompson, D., Wallace, J., 2001. Regional climate impacts of the NorthernHemisphere annular mode. Science 293, 85–89.

Toner, M., Keddy, P., 1997. River hydrology and riparian wetlands: A predictivemodel for ecological assembly. Ecological Applications 7, 236–246.

Verseghy, D.L., 1996. Local climates simulated by two generations of Canadian GCMland surface schemes. Atmosphere–Ocean 34, 435–456.

Visbeck, M., Hurrell, J.W., Polvani, L., Cullen, H.M., 2001. The North Atlanticoscillation: past present and future. Proceeding of the National AcademyScience 98, 12 876–12 877.

Whitfield, P.H., Cannon, A.J., 2000. Recent variations in climate and hydrology inCanada. Canadian Water Resources Journal 25, 19–65.

Whitfield, P.H., Wang, J.Y., Cannon, A.J., 2003. Modelling future streamflowextremes—floods and low flows in Georgia Basin, British Columbia. CanadianWater Resources Journal 28, 633–656.

Wilby, R.L., Wigley, T.M.L., Conway, D., Jones, P.D., Hewitson, B.C., Main, J., Wilks,D.S., 1998. Statistical downscaling of general circulation model output: Acomparison of methods. Water Resources Research 34, 2995–3008.

Wilby, R.L., Charles, S.P., Zorita, E., Timbal, B., Whetton, P., Means, L.O., 2004.Guidelines for Use of Climate Scenarios Developed from StatisticalDownscaling, IPCC Task Group on Scenarios for Climate Impact Assessment(TGCIA).

Zhang, X., Harvey, K.D., Hogg, W.D., Yuzyk, T.R., 2001. Trends in Canadianstreamflow. Water Resources Research 37, 987–998.



Journal of Hydrologyhuardda/articles/boyer10.pdf · to 6 of latitude. The rivers differ in their...

Documents

Transcript of Journal of Hydrologyhuardda/articles/boyer10.pdf · to 6 of latitude. The rivers differ in their...