• Alexandru, A., , de Elía R. , , and Laprise R. , 2007: Internal variability in regional climate downscaling at the seasonal scale. Mon. Wea. Rev., 135, 32213238, doi:10.1175/MWR3456.1.

    • Search Google Scholar
    • Export Citation
  • Anderson, C. J., and et al. , 2003: Hydrological processes in regional climate model simulations of the central United States flood of June–July 1993. J. Hydrometeor., 4, 584598, doi:10.1175/1525-7541(2003)004<0584:HPIRCM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bisson, J. L., , and Roberge F. , 1983: Prévisions des apports naturels: Expérience d’Hydro-Québec. Proc. Workshop on Flow Prediction, Toronto, ON, Canada, IEEE, 148161.

  • Bougeault, P., 1985: A simple parameterization of the large-scale effects of cumulus convection. Mon. Wea. Rev., 113, 21082121, doi:10.1175/1520-0493(1985)113<2108:ASPOTL>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Colin, J., , Déqué M. , , Radu R. , , and Somot S. , 2010: Sensitivity study of heavy precipitations in Limited Area Model climate simulation: Influence of the size of the domain and the use of the spectral nudging technique. Tellus, 62A, 591604, doi:10.1111/j.1600-0870.2010.00467.x.

    • Search Google Scholar
    • Export Citation
  • Côté, J., , Gravel S. , , Méthot A. , , Patoine A. , , Roch M. , , and Staniforth A. , 1998: The operational CMC-MRB global environmental multiscale (GEM) model. Part I: Design considerations and formulation. Mon. Wea. Rev., 126, 13731395, doi:10.1175/1520-0493(1998)126<1373:TOCMGE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and et al. , 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597, doi:10.1002/qj.828.

    • Search Google Scholar
    • Export Citation
  • Déqué, M., , and Somot S. , 2008: Analysis of heavy precipitation for France using ALADIN RCM simulations. Idojaras, 112, 179190.

  • DiNapoli, S. M., , and Misra V. , 2012: Reconstructing the 20th century high-resolution climate of the southeastern United States. J. Geophys. Res., 117, D19113, doi:10.1029/2012JD018303.

    • Search Google Scholar
    • Export Citation
  • Douville, H., , Royer J.-F. , , and Mahfouf J.-F. , 1995: A new snow parameterization for the Météo-France climate model. Part 1: Validation in stand-alone experiments. Climate Dyn., 12, 2135, doi:10.1007/BF00208760.

    • Search Google Scholar
    • Export Citation
  • Fischer, E. M., , Seneviratne S. I. , , Vidale P. L. , , Lüthi D. , , and Schär C. , 2007: Soil moisture–atmosphere interactions during the 2003 European summer heat wave. J. Climate, 20, 50815099, doi:10.1175/JCLI4288.1.

    • Search Google Scholar
    • Export Citation
  • Flesch, T. K., , and Reuter G. W. , 2012: WRF Model simulation of two Alberta flooding events and the impact of topography. J. Hydrometeor., 13, 695708, doi:10.1175/JHM-D-11-035.1.

    • Search Google Scholar
    • Export Citation
  • Harris, I., , Jones P. D. , , Osborn T. J. , , and Lister D. H. , 2014: Updated high-resolution grids of monthly climatic observations—The CRU TS3.10 Dataset. Int. J. Climatol., 34, 623642, doi:10.1002/joc.3711.

    • Search Google Scholar
    • Export Citation
  • Hernández-Díaz, L., , Laprise R. , , Sushama L. , , Martynov A. , , Winger K. , , and Dugas B. , 2013: Climate simulation over the CORDEX-Africa domain using the fifth-generation Canadian Regional Climate Model (CRCM5). Climate Dyn., 40, 14151433, doi:10.1007/s00382-012-1387-z.

    • Search Google Scholar
    • Export Citation
  • Herrmann, M. J., , and Somot S. , 2008: Relevance of ERA40 dynamical downscaling for modeling deep convection in the Mediterranean Sea. Geophys. Res. Lett., 35, L04607, doi:10.1029/2007GL032442.

    • Search Google Scholar
    • Export Citation
  • Herrmann, M. J., , Somot S. , , Calmanti S. , , Dubois C. , , and Sevault F. , 2011: Representation of spatial and temporal variability of daily wind speed and of intense wind events over the Mediterranean Sea using dynamical downscaling: Impact of the regional climate model configuration. Nat. Hazards Earth Syst. Sci., 11, 19832001, doi:10.5194/nhess-11-1983-2011.

    • Search Google Scholar
    • Export Citation
  • International Joint Commission, 2013: Plan of study for the identification of measures to mitigate flooding and the impacts of flooding of Lake Champlain and Richelieu River. IJC Rep., 128 pp. [Available online at www.ijc.org/files/publications/Final_PoS_LakeChamplain-RichelieuRiver.pdf.]

  • Kain, J. S., , and Fritsch J. M. , 1990: A one-dimensional entraining/detraining plume model and application in convective parameterization. J. Atmos. Sci., 47, 27842802, doi:10.1175/1520-0469(1990)047<2784:AODEPM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kanamitsu, M., , and Kanamaru H. , 2007: Fifty-seven-year California reanalysis downscaling at 10 km (CaRD10). Part I: System detail and validation with observations. J. Climate, 20, 55535571, doi:10.1175/2007JCLI1482.1.

    • Search Google Scholar
    • Export Citation
  • Kanamitsu, M., , Ebisuzaki W. , , Woollen J. , , Yang S.-K. , , Hnilo J. J. , , Fiorino M. , , and Potter G. L. , 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 16311643, doi:10.1175/BAMS-83-11-1631.

    • Search Google Scholar
    • Export Citation
  • Kotlarski, S., , Hagemann S. , , Krahe P. , , Podzun R. , , and Jacob D. , 2012: The Elbe River flooding 2002 as seen by an extended regional climate model. J. Hydrol., 472–473, 169183, doi:10.1016/j.jhydrol.2012.09.020.

    • Search Google Scholar
    • Export Citation
  • Laprise, R., , Hernández-Díaz L. , , Tete K. , , Sushama L. , , Šeparović L. , , Martynov A. , , Winger K. , , and Valin M. , 2013: Climate projections over CORDEX Africa domain using the fifth-generation Canadian Regional Climate Model (CRCM5). Climate Dyn., 41, 32193246, doi:10.1007/s00382-012-1651-2.

    • Search Google Scholar
    • Export Citation
  • Leung, L. R., 2012: Regional climate models. Climate Change Modeling Methodology: Selected Entries from the Encyclopedia of Sustainability Science and Technology, P. J. Rasch, Ed., Springer, 211–233, doi:10.1007/978-1-4614-5767-1_9.

  • Leung, L. R., , and Qian Y. , 2003: The sensitivity of precipitation and snowpack simulations to model resolution via nesting in regions of complex terrain. J. Hydrometeor., 4, 10251043, doi:10.1175/1525-7541(2003)004<1025:TSOPAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Leung, L. R., , Qian Y. , , and Bian X. , 2003a: Hydroclimate of the western United States based on observations and regional climate simulations of 1981–2000. Part I: Seasonal statistics. J. Climate, 16, 18921911, doi:10.1175/1520-0442(2003)016<1892:HOTWUS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Leung, L. R., , Qian Y. , , Han J. , , and Roads J. O. , 2003b: Intercomparison of global reanalyses and regional simulations of cold season water budgets in the western United States. J. Hydrometeor., 4, 10671087, doi:10.1175/1525-7541(2003)004<1067:IOGRAR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lim, Y.-K., , Stefanova L. B. , , Chan S. C. , , Schubert S. D. , , and O’Brien J. J. , 2011: High-resolution subtropical summer precipitation derived from dynamical downscaling of the NCEP/DOE reanalysis: How much small-scale information is added by a regional model? Climate Dyn., 37, 10611080, doi:10.1007/s00382-010-0891-2.

    • Search Google Scholar
    • Export Citation
  • Livneh, B., , Rosenberg E. A. , , Lin C. , , Nijssen B. , , Mishra V. , , Andreadis K. M. , , Maurer E. P. , , and Lettenmaier D. P. , 2013: A long-term hydrologically based dataset of land surface fluxes and states for the conterminous United States: Update and extensions. J. Climate, 26, 93849392, doi:10.1175/JCLI-D-12-00508.1.

    • Search Google Scholar
    • Export Citation
  • Lorenz, C., , and Kunstmann H. , 2012: The hydrological cycle in three state-of-the-art reanalyses: Intercomparison and performance analysis. J. Hydrometeor., 13, 13971420, doi:10.1175/JHM-D-11-088.1.

    • Search Google Scholar
    • Export Citation
  • Lucas-Picher, P., , Arora V. K. , , Caya D. , , and Laprise R. , 2003: Implementation of a large-scale variable velocity river flow routing algorithm in the Canadian regional climate model. Atmos.–Ocean, 41, 139153, doi:10.3137/ao.410203.

    • Search Google Scholar
    • Export Citation
  • Lucas-Picher, P., , Caya D. , , Biner S. , , and Laprise R. , 2008a: Quantification of the lateral boundary forcing of a regional climate model using an ageing tracer. Mon. Wea. Rev., 136, 49804996, doi:10.1175/2008MWR2448.1.

    • Search Google Scholar
    • Export Citation
  • Lucas-Picher, P., , Caya D. , , de Elía R. , , and Laprise R. , 2008b: Investigation of regional climate models’ internal variability with a ten-member ensemble of ten years over a large domain. Climate Dyn., 31, 927940, doi:10.1007/s00382-008-0384-8.

    • Search Google Scholar
    • Export Citation
  • Lucas-Picher, P., , Boberg F. , , Christensen J. H. , , and Berg P. , 2013a: Dynamical downscaling with reinitializations: A method to generate finescale climate datasets suitable for impact studies. J. Hydrometeor., 14, 11591174, doi:10.1175/JHM-D-12-063.1.

    • Search Google Scholar
    • Export Citation
  • Lucas-Picher, P., , Somot S. , , Déqué M. , , Decharme B. , , and Alias A. , 2013b: Evaluation of the regional climate model ALADIN to simulate the climate over North America in the CORDEX framework. Climate Dyn., 41, 11171137, doi:10.1007/s00382-012-1613-8.

    • Search Google Scholar
    • Export Citation
  • Martynov, A., , Laprise R. , , Sushama L. , , Winger K. , , Šeparović L. , , and Dugas B. , 2013: Reanalysis-driven climate simulation over CORDEX North America domain using the Canadian Regional Climate Model, version 5: Model performance evaluation. Climate Dyn., 41, 29733005, doi:10.1007/s00382-013-1778-9.

    • Search Google Scholar
    • Export Citation
  • Mesinger, F., and et al. , 2006: North American Regional Reanalysis. Bull. Amer. Meteor. Soc., 87, 343360, doi:10.1175/BAMS-87-3-343.

    • Search Google Scholar
    • Export Citation
  • Minville, M., , Brissette F. , , and Leconte R. , 2008: Uncertainty of the impact of climate change on the hydrology of a Nordic watershed. J. Hydrol., 358, 7083, doi:10.1016/j.jhydrol.2008.05.033.

    • Search Google Scholar
    • Export Citation
  • Noilhan, J., , and Planton S. , 1989: A simple parameterization of land surface processes for meteorological models. Mon. Wea. Rev., 117, 536549, doi:10.1175/1520-0493(1989)117<0536:ASPOLS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Radu, R., , Déqué M. , , and Somot S. , 2008: Spectral nudging in a spectral regional climate model. Tellus, 60A, 898910, doi:10.1111/j.1600-0870.2008.00341.x.

    • Search Google Scholar
    • Export Citation
  • Rasmussen, R., and et al. , 2011: High-resolution coupled climate runoff simulations of seasonal snowfall over Colorado: A process study of current and warmer climate. J. Climate, 24, 30153048, doi:10.1175/2010JCLI3985.1.

    • Search Google Scholar
    • Export Citation
  • Riboust, P., , and Brissette F. , 2014a: Climate change impacts and uncertainties on spring flooding of Lake Champlain and the Richelieu River. J. Amer. Water Resour. Assoc., in press.

    • Search Google Scholar
    • Export Citation
  • Riboust, P., , and Brissette F. , 2014b: Analysis of Lake-Champlain/Richelieu River's historical 2011 flood. Can. Water Resour. J., in press.

    • Search Google Scholar
    • Export Citation
  • Ricard, J. L., , and Royer J. F. , 1993: A statistical cloud scheme for use in an AGCM. Ann. Geophys., 11, 10951115.

  • Rienecker, M. M., and et al. , 2011: MERRA: NASA’s Modern-Era Retrospective Analysis for Research and Applications. J. Climate, 24, 36243648, doi:10.1175/JCLI-D-11-00015.1.

    • Search Google Scholar
    • Export Citation
  • Rost, A., 2010: Snow board: A look at the importance of snow information. Meteorological Technology International, June 2010, UKiP, Dorking, United Kingdom, 26–29.

  • Rummukainen, M., 2010: State-of-the-art with regional climate models. Wiley Interdiscip. Rev.: Climate Change,1, 82–96, doi:10.1002/wcc.8.

  • Salzmann, N., , and Mearns L. O. , 2012: Assessing the performance of multiple regional climate model simulations for seasonal mountain snow in the upper Colorado River basin. J. Hydrometeor., 13, 539556, doi:10.1175/2011JHM1371.1.

    • Search Google Scholar
    • Export Citation
  • Sanchez-Gomez, E., , Somot S. , , and Déqué M. , 2009a: Ability of an ensemble of regional climate models to reproduce the weather regimes during the period 1961–2000. Climate Dyn., 33, 723736, doi:10.1007/s00382-008-0502-7.

    • Search Google Scholar
    • Export Citation
  • Sanchez-Gomez, E., , Somot S. , , and Mariotti A. , 2009b: Future changes in the Mediterranean water budget projected by an ensemble of regional climate models. Geophys. Res. Lett., 36, L21401, doi:10.1029/2009GL040120.

    • Search Google Scholar
    • Export Citation
  • Šeparović, L., , Alexandru A. , , Laprise R. , , Martynov A. , , Sushama L. , , Winger K. , , Tete K. , , and Valin M. , 2013: Present climate and climate change over North America as simulated by the fifth-generation Canadian Regional Climate Model (CRCM5). Climate Dyn., 41, 31673201, doi:10.1007/s00382-013-1737-5.

    • Search Google Scholar
    • Export Citation
  • Sheffield, J., , Livneh B. , , and Wood E. F. , 2012: Representation of terrestrial hydrology and large-scale drought of the continental United States from the North American Regional Reanalysis. J. Hydrometeor., 13, 856876, doi:10.1175/JHM-D-11-065.1.

    • Search Google Scholar
    • Export Citation
  • Stefanova, L., , Misra V. , , Chan S. , , Griffin M. , , O’Brien J. J. , , and Smith T. J. III, 2012: A proxy for high-resolution regional reanalysis for the southeast United States: Assessment of precipitation variability in dynamically downscaled reanalyses. Climate Dyn., 38, 24492466, doi:10.1007/s00382-011-1230-y.

    • Search Google Scholar
    • Export Citation
  • Sundqvist, H., , Berge E. , , and Kristjansson J. E. , 1989: Condensation and cloud parameterization studies with a mesoscale numerical weather prediction model. Mon. Wea. Rev., 117, 16411657, doi:10.1175/1520-0493(1989)117<1641:CACPSW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Verseghy, D. L., 2009: CLASS—The Canadian Land Surface Scheme (version 3.4): Technical documentation (version 1.1). Internal Rep., Climate Research Division, Science and Technology Branch, Environment Canada, 180 pp. [Available online at www.usask.ca/ip3/download/CLASS_v3_4_Documentation_v1_1.pdf.]

  • Voldoire, A., and et al. , 2013: The CNRM-CM5.1 global climate model: Description and basic evaluation. Climate Dyn., 40, 20912121, doi:10.1007/s00382-011-1259-y.

    • Search Google Scholar
    • Export Citation
  • von Storch, H., , Zorita E. , , and Cubasch U. , 1993: Downscaling of global climate change estimates to regional scales: An application to Iberian rainfall in wintertime. J. Climate, 6, 11611171, doi:10.1175/1520-0442(1993)006<1161:DOGCCE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Topography (m) of the Richelieu River basin and its surroundings used by ALADIN. The Richelieu River basin is indicated with red and black contours. Only the part of the basin on the U.S. side (with the boldface black contour) will be considered for the analysis. Lake Champlain is located in the middle of the Richelieu River basin.

  • View in gallery

    Richelieu River discharge (m3 s−1) observed during 1938–2011 at Carignan station with a 3-day running mean filter. The mean, min, and max for the period 1938–2010 are in green, blue, and red, respectively. The year 2011 is in black.

  • View in gallery

    Richelieu River max annual discharge (m3 s−1) observed during 1938–2011 at Carignan station.

  • View in gallery

    Topography (m) of the free RCM domains. The ALADIN (CRCM5) domain is in black (red). The small black inner domain indicates where the analyses will be performed. The Richelieu basin is indicated with the dark blue contour.

  • View in gallery

    The 2-m air temperature (°C) time series over the U.S. part of the Richelieu River basin: (a) 1990–2011 mean annual cycle and (b) 2004–11 monthly 2-m air temperature anomalies according to the 1990–2011 monthly means. In (a), the top-right values indicate the annual mean. In (b), the first and second columns at the bottom right indicate the 1990–2011 intermonth std dev and the 1990–2011 monthly temporal correlation with respect to CONUS.

  • View in gallery

    Mean 2-m air temperature (°C) in (first column) January and (third column) July from the two RCM simulations (ALADIN and CRCM5), ERAI, and two gridded observations (CONUS and CRU) during 1990–2011. The bias with respect to CRU is also shown in (second column) January and (fourth column) July.

  • View in gallery

    As in Fig. 5, but for precipitation (mm day−1).

  • View in gallery

    Mean precipitation (mm day−1) in (left) January and (right) July from the two RCM simulations (ALADIN and CRCM5), ERAI, and two gridded observations (CONUS and CRU) during 1990–2011.

  • View in gallery

    Mean precipitation (mm day−1) in May (left) 1990–2011 and (right) 2011 from the two RCM simulations (ALADIN and CRCM5), ERAI, and two gridded observations (CONUS and CRU).

  • View in gallery

    SWE (mm) over the U.S. part of the Richelieu River basin: (a) 2004–11 mean annual cycle and (b) 2004–11 time series using monthly values. In (b), the first and second columns next to the dataset names indicate the 1990–2011 monthly temporal correlation relative to NSA and the 2004–11 annual mean, respectively.

  • View in gallery

    Mean SWE (mm) in March (left) 2004–11 and (right) 2011 from the two RCM simulations (ALADIN and CRCM5), ERAI, and the NSA.

  • View in gallery

    Mean annual cycle of the water cycle variables over the U.S. part of the Richelieu basin during 1990–2011 from ALADIN, CRCM5, ERAI, NARR, and MERRA. Precipitation (PCP) is in blue, surface runoff (RUNOFF) is in red, snowmelt (SNME) is in green, drainage (DRAI) is in light blue, evapotranspiration (EVAP) is in black, and SWE is in gray. All variables (mm day−1) refer to the left y axis, except SWE (mm), which refers to the right y axis.

  • View in gallery

    Box plots of the 1990–2011 monthly (top) 2-m air temperature (°C), (middle) precipitation (mm day−1), and (bottom) SWE (mm) over the U.S. part of the Richelieu basin from ALADIN, CRCM5, ERAI, and CONUS. The 2011 values are indicated with the green asterisks.

  • View in gallery

    Daily time series of the 2011 precipitation (mm day−1) over the U.S. part of the Richelieu basin from ALADIN, CRCM5, ERAI, and CONUS. A 3-day running mean was applied. The temporal correlation of the 3-day running mean field with CONUS is indicated next to the datasets at the top right corner.

  • View in gallery

    Daily time series of the 2011 precipitation, runoff, SWE, and 2-m air temperature over the U.S. part of the Richelieu basin from (a) ALADIN and (b) CRCM5. The simulated discharge of the Richelieu River by the hydrological model HSAMI is indicated in black. A 3-day running mean is applied. Precipitation (mm day−1), runoff (mm day−1), and 2-m air temperature (°C) refer to the left axis. SWE (mm) refers to the right axis, and the discharge (m3 s−1) refers to the right-most axis.

  • View in gallery

    Discharge (m3 s−1) of the Richelieu River at the Carignan station simulated by the hydrological model HSAMI using different forcing fields: (a) 1990–2011 mean annual cycle and (b) 2011 cycle. The numbers in parentheses correspond to the Nash–Sutcliffe efficiency coefficient of the simulated with the observed discharge computed with daily values. The Nash–Sutcliffe efficiency coefficient of (a) is computed using the daily values from 1990 to 2011.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 39 39 9
PDF Downloads 20 20 3

Reconstruction of the Spring 2011 Richelieu River Flood by Two Regional Climate Models and a Hydrological Model

View More View Less
  • 1 Centre pour l’étude et la simulation du climat à l’échelle régionale, Département des sciences de la Terre et de l’atmosphère, Université du Québec à Montréal, and Département de génie de la construction, École de Technologie Supérieure, Université du Québec, Montréal, Québec, Canada
  • | 2 Département de génie de la construction, École de Technologie Supérieure, Université du Québec, Montréal, Québec, Canada
  • | 3 Centre National de Recherches Météorologiques (CNRM-GAME), Météo-France/CNRS, Toulouse, France
  • | 4 Centre pour l’étude et la simulation du climat à l’échelle régionale, Département des sciences de la Terre et de l’atmosphère, Université du Québec à Montréal, Montréal, Québec, Canada
© Get Permissions
Full access

Abstract

Climate simulations made with two regional climate models (RCMs), the French Aire Limitée Adaptation Dynamique Développement International (ALADIN) and the Canadian Regional Climate Model, version 5 (CRCM5), operating on 10-km meshes for the period 1989–2011, and the Hydro-Québec hydrological model (HSAMI), are used to reconstruct the spring 2011 Richelieu River flood in the southern region of the province of Québec, Canada. The analysis shows that the simulated fields of 2-m air temperature, precipitation, and snow water equivalent by the RCMs closely match the observations with similar multiyear means and a high correlation of the monthly anomalies. The climatic conditions responsible for the 2011 flood are generally well simulated by the RCMs. The use of multidecadal RCM simulations facilitates the identification of anomalies that contributed to the flood. The flood was linked to a combination of factors: the 2010/11 winter was cold and snowy, the snowmelt in spring was fast, and there was a record amount of precipitation in April and May. Driven by outputs from the RCMs, HSAMI was able to reproduce the mean hydrograph of the Richelieu River, but it underestimated the peak of the 2011 flood. HSAMI adequately computes the water transport from the mountains to the river mouth and the storage effect of Lake Champlain, which dampens the flood over a long period. Overall, the results suggest that RCM simulations can be useful for reconstructing high-resolution climate information and providing new variables that can help better understand the causes of extreme climatic events.

Corresponding author address: Dr. Philippe Lucas-Picher, Centre ESCER, Dép. des sciences de la Terre et de l’atmosphère, UQÀM, P.O. Box 8888, Stn. Downtown, Montréal QC H3C 3P8, Canada. E-mail: plp@sca.uqam.ca

Abstract

Climate simulations made with two regional climate models (RCMs), the French Aire Limitée Adaptation Dynamique Développement International (ALADIN) and the Canadian Regional Climate Model, version 5 (CRCM5), operating on 10-km meshes for the period 1989–2011, and the Hydro-Québec hydrological model (HSAMI), are used to reconstruct the spring 2011 Richelieu River flood in the southern region of the province of Québec, Canada. The analysis shows that the simulated fields of 2-m air temperature, precipitation, and snow water equivalent by the RCMs closely match the observations with similar multiyear means and a high correlation of the monthly anomalies. The climatic conditions responsible for the 2011 flood are generally well simulated by the RCMs. The use of multidecadal RCM simulations facilitates the identification of anomalies that contributed to the flood. The flood was linked to a combination of factors: the 2010/11 winter was cold and snowy, the snowmelt in spring was fast, and there was a record amount of precipitation in April and May. Driven by outputs from the RCMs, HSAMI was able to reproduce the mean hydrograph of the Richelieu River, but it underestimated the peak of the 2011 flood. HSAMI adequately computes the water transport from the mountains to the river mouth and the storage effect of Lake Champlain, which dampens the flood over a long period. Overall, the results suggest that RCM simulations can be useful for reconstructing high-resolution climate information and providing new variables that can help better understand the causes of extreme climatic events.

Corresponding author address: Dr. Philippe Lucas-Picher, Centre ESCER, Dép. des sciences de la Terre et de l’atmosphère, UQÀM, P.O. Box 8888, Stn. Downtown, Montréal QC H3C 3P8, Canada. E-mail: plp@sca.uqam.ca

1. Introduction

The Richelieu River basin originates in small streams of the Appalachian Mountains, more precisely, on the western slopes of the Green Mountains and the eastern slopes of the Adirondack Mountains in the northeastern United States (Fig. 1). The water from these streams flows to Lake Champlain, which acts as a large reservoir. The water then leaves the lake toward the north in the Richelieu River, which in turn flows to the St. Lawrence River in the southern region of the province of Québec, Canada. In spring 2011, a major flood occurred on the shores of the Richelieu River and around Lake Champlain. A maximum river discharge of 1542.3 m3 s−1 was measured on 6 May 2011 at the station Carignan (Fig. 2), establishing a new record since the beginning of measurements in 1938, being 22% above the previous record set in 1993 (Figs. 2, 3). Exceeding flood stage for 67 days (International Joint Commission 2013), the flood was ranked third among Canada’s top 10 weather stories for 2011 and was one of Québec’s longest-lasting natural disasters ever, wearing out residents physically and mentally.

Fig. 1.
Fig. 1.

Topography (m) of the Richelieu River basin and its surroundings used by ALADIN. The Richelieu River basin is indicated with red and black contours. Only the part of the basin on the U.S. side (with the boldface black contour) will be considered for the analysis. Lake Champlain is located in the middle of the Richelieu River basin.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

Fig. 2.
Fig. 2.

Richelieu River discharge (m3 s−1) observed during 1938–2011 at Carignan station with a 3-day running mean filter. The mean, min, and max for the period 1938–2010 are in green, blue, and red, respectively. The year 2011 is in black.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

Fig. 3.
Fig. 3.

Richelieu River max annual discharge (m3 s−1) observed during 1938–2011 at Carignan station.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

Because of the sparse spatial distribution of weather stations and the lack of long meteorological records in this mountainous basin, it is difficult to precisely determine the causes of the 2011 flood. Moreover, many important variables of the water cycle, such as snow water equivalent (SWE), soil moisture, runoff, and evapotranspiration, are very difficult to measure, thus making the calculation of the water budget over the basin and the characterization of floods from observations even more challenging. To address the small density of weather stations and the lack of long meteorological records, meteorological reanalyses constitute good alternatives. However, with a horizontal mesh of 75 km or coarser, state-of-the-art global reanalyses such as the Interim European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-Interim, hereinafter ERAI; Dee et al. 2011) or the National Centers for Environmental Prediction (NCEP)–U.S. Department of Energy (DOE) Atmospheric Model Intercomparison Project (AMIP)-II reanalysis (R-2; Kanamitsu et al. 2002) are too coarse to adequately represent the local climate, especially in a mountainous region like the Appalachians. Finer reanalyses, like the North American Regional Reanalysis (NARR; Mesinger et al. 2006) with a grid mesh of 32 km, do exist, but they nevertheless contain many drawbacks related to reanalysis technique (Sheffield et al. 2012). Degradation, replacement, or changes of instruments may introduce errors and discontinuities in reanalyses. Also, the water cycles in most reanalyses do not close because they are generated from the collection of short simulations launched from initial conditions obtained using data assimilation. Also, many variables such as precipitation and evapotranspiration are not assimilated, but are rather obtained as a by-product of models that contain biases and rely on parameterizations. The limited performance of the reanalyses in reproducing the hydrological cycle puts their use for climate trend analyses and long-term water budget studies into question (Lorenz and Kunstmann 2012).

For more than 20 years, a recognized way to obtain long-term spatially and temporally coherent meteorological data at high resolution consists of dynamically downscaling a reanalysis using a regional climate model (RCM; Leung 2012). Based on physical laws such as the conservations of mass, momentum, and energy, RCMs are able to consistently simulate all variables of the water cycle and to close the hydrological and energy budgets. The dynamical downscaling concept behind RCMs is that small-scale climate details can be generated using a high-resolution climate model driven at its lateral boundaries by large-scale climate information (von Storch et al. 1993; Rummukainen 2010). Dynamical downscaling serves as a regionalization of coarse-resolution reanalyses without conducting expensive high-resolution data assimilation (Kanamitsu and Kanamaru 2007). In other words, the dynamical downscaling of a reanalysis with an RCM can be described as a regional reanalysis without data assimilation using observations. This concept has been recently explored by Kanamitsu and Kanamaru (2007), Herrmann and Somot (2008), Lim et al. (2011), DiNapoli and Misra (2012), and Stefanova et al. (2012). Moreover, when a reanalysis is downscaled, the evaluation of the RCM simulation against observations becomes feasible in a “time-series sense” if the RCM internal variability remains small (Fischer et al. 2007; Rummukainen 2010; Kotlarski et al. 2012).

Many studies compared the water budgets simulated by the RCMs and the reanalyses that are forcing the RCMs at their lateral boundaries. For example, in their analysis of the hydroclimate of the western United States, Leung et al. (2003a) showed that RCM simulations retain large-scale features of the reanalyses while adding useful information at the regional scales. Their study also revealed that the spatial distribution of temperature and precipitation are significantly improved in the RCM simulations compared to the reanalyses. In another study, Leung et al. (2003b) compared three global reanalyses and two simulations with different RCMs. Large differences in the moisture fluxes and precipitation of the different reanalyses were found, and they determined that the regional simulations were superior to the global reanalyses in terms of the spatial distribution of mean precipitation and precipitation anomalies.

One very challenging variable of the water cycle to simulate is the snowpack. Leung and Qian (2003) showed that the use of RCMs’ higher spatial resolution improves the simulation of snowpack because of the higher elevation of the topography in RCMs compared to reanalyses. They also found that the accuracy of the snow simulation is limited by factors such as deficiencies in the land surface model or biases in other model variables. In turn, Rasmussen et al. (2011) obtained a very good comparison of annual snowfall accumulation over the Colorado headwaters regions between the RCM Weather Research and Forecasting (WRF) Model and Snow Telemetry (SNOTEL) observations. Finally, Salzmann and Mearns (2012) assessed the performance of the RCMs used in the North American Regional Climate Change Assessment Program (NARCCAP) for snow in the upper Colorado River basin. They found that most RCMs significantly improved the SWE compared to the driving NCEP–National Center for Atmospheric Research (NCAR) reanalysis. However, RCMs generally simulated too little SWE mainly because of inadequately resolved topography and biases due to the deficiencies of the land surface schemes.

Many studies used an RCM driven with a reanalysis to simulate specific floods that occurred in the past. Anderson et al. (2003) performed an intercomparison of 13 RCMs and assessed their ability to reproduce a 1993 flood that occurred in the central United States. Flesch and Reuter (2012) successfully simulated two Alberta flooding events in June 2005 using the WRF Model. Other studies have coupled an RCM with a hydrological model to reproduce the hydrograph of specific floods. For instance, Kotlarski et al. (2012) were able to reproduce the August 2002 Elbe River flooding discharge using the RCM Regional Model (REMO) and the river routing scheme Hydrological Discharge (HD). Quantifying the temporal variability of the terrestrial water budget from subdaily to annual scales is central to understanding the mechanisms and predictability of hydrological hazards (Sheffield et al. 2012).

All of the studies mentioned above focused on flash floods caused by short and strong precipitation events that occurred in summer, and the RCM simulations lasted only a few months at most. This prevented the computation of RCM simulation climate statistics and thus of the model-simulated anomaly during the flood event. In this paper, we will focus on a spring flood that occurred in the southern region of Québec in 2011. High-resolution, 10-km simulations exceeding two decades were performed with two different RCMs to reconstruct the atmospheric and surface conditions over Québec and the northeastern United States. The selected RCM formulations include state-of-the-art sophisticated land surface schemes, allowing a realistic simulation of the surface conditions. Moreover, the long-simulated climate sample allows us to compute climate statistics and to determine the amplitude of the anomalies that led to the flood.

The following section describes the two RCMs, the hydrological model, and characteristics of their simulations. The analysis is divided into three sections. The evaluation of the RCM simulations with observationally based gridded datasets is done in section 3. During the evaluation, the RCM simulations will also be compared to the driving reanalysis to show the RCMs’ added value. Then, a statistical analysis is performed in section 4 to determine the causes of this exceptional flood. Finally, the simulated river discharge with the hydrological model is compared with the observed discharge in section 5. Conclusions are presented in section 6.

2. Experimental setup

a. The RCM ALADIN

The Aire Limitée Adaptation Dynamique Développement International (ALADIN) model, ALADIN-Climate, is a limited-area RCM developed at the Centre National de Recherches Météorologiques (CNRM) at Météo-France. It is the climate version of the ALADIN limited-area model used for weather forecasting. ALADIN-Climate has been mainly utilized over France (Déqué and Somot 2008; Colin et al. 2010), Europe (Radu et al. 2008; Sanchez-Gomez et al. 2009a), and the Mediterranean region (Sanchez-Gomez et al. 2009b; Herrmann et al. 2011). ALADIN-Climate is currently participating in the international project Coordinated Regional Climate Downscaling Experiment (CORDEX) over the African, Mediterranean, European, and North American domains. In this paper, we used ALADIN-Climate, version 5.3 (ALADIN-Climate v5.3; CNRM-ALADIN53 in the CORDEX nomenclature), as described by Lucas-Picher et al. (2013b). The main new features of ALADIN-Climate v5.3 (simply called ALADIN in the following) with respect to previous versions (Colin et al. 2010; Herrmann et al. 2011) are the new radiation schemes and the simulation of the land surface processes with an external interface called Surface Externalisée (SURFEX). The basic characteristics of ALADIN are presented in Table 1.

Table 1.

Description of ALADIN and CRCM5.

Table 1.

b. The RCM CRCM5

The Canadian Regional Climate Model, version 5 (CRCM5), was developed at the Centre pour l’Étude et la Simulation du Climat à l’Échelle Régionale (ESCER) at the Université du Québec à Montréal (UQÀM). It is based on a limited-area version of the Global Environment Multiscale (GEM) model used for numerical weather prediction at Environment Canada (Côté et al. 1998). The Canadian Land Surface Scheme, version 3.5 (CLASS 3.5; Verseghy 2009), has been implemented in CRCM5 with some modifications. The basic characteristics of CRCM5 are presented in Table 1. A detailed description of CRCM5 and its performance over North America can be found in Martynov et al. (2013). CRCM5 has contributed to the CORDEX project over North America (Martynov et al. 2013; Šeparović et al. 2013) and Africa (Hernández-Díaz et al. 2013; Laprise et al. 2013).

c. RCM simulations centered over Québec

ALADIN and CRCM5 were used to generate climate simulations for a domain centered over Québec using grid meshes of approximately 10 km (Fig. 4). The domains are slightly different because the RCMs use different projections: Lambert conformal for ALADIN and rotated latitude–longitude for CRCM5. The RCMs were driven at their lateral boundaries by ERAI (Dee et al. 2011). The ERAI fields that forced the RCMs at the atmospheric lateral boundaries are the horizontal wind components, temperature, specific humidity, and surface pressure. These variables are passed to the RCMs every 6 h and are linearly interpolated in time to provide data to the RCMs at every time step. The details of the RCM simulations are summarized in Table 2. Both simulations cover the period 1989–2011. To perform the analysis on the RCM simulations once they are in equilibrium, the upcoming analysis started in 1990, discarding the first simulated year. Using a relatively small domain size, the internal spectral nudging technique was not employed. The sea surface temperature of both simulations was taken from the closest ERAI sea cell. The lake temperature is simulated by the 1D freshwater lake (FLake) model in CRCM5, while in ALADIN, the lake temperature comes from the closest ERAI cell, which is considered to be water.

Fig. 4.
Fig. 4.

Topography (m) of the free RCM domains. The ALADIN (CRCM5) domain is in black (red). The small black inner domain indicates where the analyses will be performed. The Richelieu basin is indicated with the dark blue contour.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

Table 2.

Description of the ALADIN and CRCM5 simulations.

Table 2.

d. HSAMI hydrological model

HSAMI is a hydrological model that was developed by Hydro-Québec (Bisson and Roberge 1983; Minville et al. 2008). It has been used for 20 years in an operational frame for short- and medium-term flow forecasting. HSAMI is a lumped conceptual model that is linear reservoir based. In its most common configuration, HSAMI inputs are the daily minimum and maximum temperatures and precipitation that come from observations or weather forecast models. HSAMI determines the type of precipitation according to the mean temperature (minimum plus maximum divided by 2). If the mean is below −3°C, precipitation is considered snow; if the mean is above +3°C, precipitation is considered rain; and if the mean is between −3°C and +3°C, a linear interpolation of solid and liquid precipitation is made. To estimate the river discharge, HSAMI simulates snow accumulation, snowmelt, evapotranspiration, interception of rain and snow, infiltration, and vertical flow processes with a total of 23 parameters.

Because the Richelieu River is the only outlet of Lake Champlain, the discharge of the river depends mainly on the water level of the lake, which acts as a reservoir. Thus, the high storage capacity of Lake Champlain requires the creation of a reservoir model with a mass balance equation. The inputs to Lake Champlain come from upstream rivers and precipitation into the lake. The outputs of the lake are computed from the lake evaporation and the water flow at the lake outlet toward the Richelieu River. The 16 subwatersheds of Lake Champlain are modeled independently with HSAMI using a semidistributed approach. Then the simulated flows from the 16 watersheds are added to the residual between the precipitation and evapotranspiration to compute the total input of the lake. Finally, the Richelieu River flow is computed using an empirical function linking the lake level to the lake outlet river discharge. For further information, we refer the reader to Riboust and Brissette (2014a,b).

e. The Richelieu River basin delimitation

The Richelieu River basin delimitation was determined from the watershed boundary dataset of the U.S. Department of Agriculture (USDA). The Richelieu basin is a medium-sized basin with a drainage area of 23 899 km2. Most of the analyses will only be performed on the side of the basin that lies in the United States (84% of the total surface of the Richelieu basin; see Fig. 1), where many high-resolution datasets are available. The part of the Richelieu River basin in the United States is represented by 216 grid cells in ALADIN and 212 in CRCM5.

3. Evaluation of the RCM simulations with observationally based gridded datasets

a. The 2-m air temperature

In Fig. 5a, the mean annual cycle of 2-m air temperature from the RCM simulations; the ERAI; and the gridded observation datasets [Climatic Research Unit (CRU), version 3.2.1 (Harris et al. 2014), and conterminous United States (CONUS), version 1.1 (Livneh et al. 2013)] is presented. In general, the RCMs’ simulated 2-m air temperature is similar to the observations. In detail, CRCM5 is slightly too warm in winter and summer (1°–2°C), while ALADIN is around 1°C too cold in spring. CRU, CONUS, and ERAI are generally very similar for the full annual cycle. In Fig. 5b, the monthly 2-m air temperature anomaly for the same simulations and observations is presented. The time sequence of the 2-m air temperature anomaly is well simulated by the two RCMs, which are only forced at their lateral boundaries by ERAI. The temporal correlations of ALADIN and CRCM5 with CONUS are 0.94 and 0.96, respectively. This is a bit below the temporal correlation of ERAI and CRU, but still close to unity. The good simulation of the 2-m air temperature annual cycle, interannual variability, and time sequence is not necessarily surprising since temperature is a large-scale variable and the Richelieu basin is in the vicinity of the southern lateral boundary of the RCM domains.

Fig. 5.
Fig. 5.

The 2-m air temperature (°C) time series over the U.S. part of the Richelieu River basin: (a) 1990–2011 mean annual cycle and (b) 2004–11 monthly 2-m air temperature anomalies according to the 1990–2011 monthly means. In (a), the top-right values indicate the annual mean. In (b), the first and second columns at the bottom right indicate the 1990–2011 intermonth std dev and the 1990–2011 monthly temporal correlation with respect to CONUS.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

The spatial distribution of the mean 2-m air temperature in January and July by the RCMs, ERAI, and the gridded observations is presented in Fig. 6. As mentioned above, the RCMs are a bit too warm in January and July compared to CRU. In July, the warmer valley in the center of the basin and the colder mountains on the eastern and western sides of the basin are better represented by the RCMs than ERAI and are closer to the gridded observations. This could be explained by the better simulation of the orographic forcing by the RCMs, which use topography fields with deeper valleys and higher mountains that are closer to the reality than for ERAI (not shown). The CONUS gridded dataset has more details than CRU because of its finer horizontal grid mesh (0.0625° for CONUS versus 0.5° for CRU) and larger amount of weather stations.

Fig. 6.
Fig. 6.

Mean 2-m air temperature (°C) in (first column) January and (third column) July from the two RCM simulations (ALADIN and CRCM5), ERAI, and two gridded observations (CONUS and CRU) during 1990–2011. The bias with respect to CRU is also shown in (second column) January and (fourth column) July.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

b. Precipitation

Figure 7 shows the mean annual cycle of precipitation by the RCMs, ERAI, and the observations. Both RCMs and ERAI are too wet in spring compared to CRU and CONUS, which agree quite well. It is possible that the RCMs’ spring wet bias was inherited from ERAI. However, it is also possible that CRU and CONUS underestimate precipitation because of the greater number of weather stations in the valley than in the mountains and because of the snowfall undercatch by the weather stations in winter. As for temperature, the RCMs are able to capture the monthly precipitation anomalies (Fig. 7b) even though precipitation is only indirectly forced at the lateral boundaries through specific humidity. Precipitation is a small-scale variable that is affected by local factors, especially in summer. Thus, the monthly temporal correlation of precipitation computed between the RCMs and CONUS is smaller than for 2-m air temperature, but still fairly high, with values above 0.74.

Fig. 7.
Fig. 7.

As in Fig. 5, but for precipitation (mm day−1).

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

The impact of the topography on precipitation is clearly shown in Fig. 8, which presents the spatial distribution of the mean precipitation in January and July. The Richelieu valley is dry and the mountains are wet. The snow belt on the eastern side of Lake Ontario (to the west of the Richelieu basin) is also well captured by the RCMs in January. These features, considered as added value, are better simulated by the RCMs than by ERAI, mainly because of the more realistic topography in the RCMs due to their higher horizontal resolution. The RCMs simulated the spatial distribution of precipitation over the basin well, but they have a tendency to become too wet to the north of the Richelieu basin.

Fig. 8.
Fig. 8.

Mean precipitation (mm day−1) in (left) January and (right) July from the two RCM simulations (ALADIN and CRCM5), ERAI, and two gridded observations (CONUS and CRU) during 1990–2011.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

As seen in Fig. 7, it is more challenging for RCMs to capture the monthly anomaly of precipitation compared to 2-m air temperature. Precipitation is a small-scale variable that is influenced by both local forcings, such as the topography, and large-scale features, such as moisture flux convergence. Figure 9 shows the multiyear mean May and the May 2011 precipitation. The large amount of precipitation in May 2011 contributed greatly to the Richelieu flood. CONUS and CRU reveal a strong positive anomaly of around 3–4 mm day−1 in May 2011 over the Richelieu basin. Based on observations, CRU and CONUS agree generally, but CONUS has more spatial details because of its finer grid. CRCM5 and ALADIN also show a strong positive anomaly, but there are many differences in the spatial distribution. As explained by Alexandru et al. (2007), the internal variability of the RCMs for precipitation is important. A low-pressure system can develop in different ways and take different paths inside the domain of an RCM that is only driven at its lateral boundaries. This behavior influences the spatial distribution of the simulated anomaly, as presented in Fig. 9 by the two RCMs. Nevertheless, the large-scale features of the precipitation anomaly are well simulated by the two RCMs, and this shows the potential for reconstructing the 2011 flood, especially taking into account that the watershed spatially integrates information over a large territory.

Fig. 9.
Fig. 9.

Mean precipitation (mm day−1) in May (left) 1990–2011 and (right) 2011 from the two RCM simulations (ALADIN and CRCM5), ERAI, and two gridded observations (CONUS and CRU).

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

c. Snow

Snow accumulation is a meteorological variable that is difficult to simulate accurately because it is the result of a long chain of physical processes. Initially depending on snowfall, snow accumulation is sensitive to the land surface scheme that handles snowpack and snowmelt. In ALADIN, the snow scheme of SURFEX uses three prognostic variables for a single-layer snowpack: the SWE mass, the snow cover density, and the snow albedo (Douville et al. 1995). The heat content of the snow is implicitly represented by the composite soil–vegetation–snow surface and restores prognostic temperatures implemented in the Interactions between Soil, Biosphere, and Atmosphere (ISBA) force–restore scheme (Noilhan and Planton 1989). In CRCM5, the snow scheme of CLASS uses four prognostic variables for a single-layer snowpack: the SWE mass, the snow temperature, the snow density, and the snow albedo. The internal energy of the snowpack is diagnosed through a change in mass or conduction of energy (Verseghy 2009).

In Fig. 10, the mean annual cycle of the simulated SWE is evaluated with the National Snow Analysis (NSA), available since 2003 from the National Oceanic and Atmospheric Administration (NOAA; Rost 2010). The mean annual cycles from the RCMs agree very well with the NSA dataset, with a maximum amount of SWE of around 80 mm occurring in February or March. On the other hand, the ERAI annual cycle is quite different, with a maximum of 120 mm in February and a rapid melt starting in March. The earlier melt in ERAI is probably induced by the warmer temperature in spring due to the lower mountains in ERAI than in the RCMs. Also, the ERAI snow analysis was affected from 1 July 2003 to 23 February 2010 by a known geolocation error introduced during the processing of the Interactive Multisensor Snow and Ice Mapping System (Dee et al. 2011). In Fig. 10b, the monthly time series from the RCMs are close to NSA, with a temporal correlation above 0.9. Throughout the years, ERAI often has a large positive bias. The climatological 2004–11 March SWE spatial distribution in Fig. 11 shows a close match between the RCMs and the NSA, with high SWE in the cold and snowy mountains and low SWE in the warm and dry valley. Also in Fig. 11, the positive anomaly of SWE in March 2011 that contributed to the 2011 spring river flood is well simulated by the RCMs and is close to NSA.

Fig. 10.
Fig. 10.

SWE (mm) over the U.S. part of the Richelieu River basin: (a) 2004–11 mean annual cycle and (b) 2004–11 time series using monthly values. In (b), the first and second columns next to the dataset names indicate the 1990–2011 monthly temporal correlation relative to NSA and the 2004–11 annual mean, respectively.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

Fig. 11.
Fig. 11.

Mean SWE (mm) in March (left) 2004–11 and (right) 2011 from the two RCM simulations (ALADIN and CRCM5), ERAI, and the NSA.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

d. Water budget

To get an idea of the full water budget for the Richelieu River basin over the United States, Fig. 12 shows the mean annual cycle of the water cycle variables simulated by the two RCMs, ERAI, NARR (Mesinger et al. 2006), and the Modern-Era Retrospective Analysis for Research and Applications (MERRA; Rienecker et al. 2011). There are large differences among the reanalyses despite the fact that they all assimilated observations. The largest difference is found for SWE, which is low for MERRA, and even lower for NARR, compared to the values of ALADIN and CRCM5, which are close to NSA (see Fig. 10). Precipitation from NARR and MERRA is closer to CRU than ERAI and the RCM simulations. The evapotranspiration from NARR and MERRA is similar and higher in summer than that of ERAI. The total runoff is divided in surface runoff (red) and drainage (blue), except for ERAI. The surface runoff and drainage are very different between the reanalyses and the RCMs. This shows the sensitivity of the land surface schemes of the different models to the computation of the total runoff from snowmelt, precipitation, and evapotranspiration.

Fig. 12.
Fig. 12.

Mean annual cycle of the water cycle variables over the U.S. part of the Richelieu basin during 1990–2011 from ALADIN, CRCM5, ERAI, NARR, and MERRA. Precipitation (PCP) is in blue, surface runoff (RUNOFF) is in red, snowmelt (SNME) is in green, drainage (DRAI) is in light blue, evapotranspiration (EVAP) is in black, and SWE is in gray. All variables (mm day−1) refer to the left y axis, except SWE (mm), which refers to the right y axis.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

The water budgets of ALADIN, CRCM5, and MERRA are closed (i.e., precipitation minus evapotranspiration equals runoff). In contrast, the water budgets of NARR and ERAI are not closed. The large differences of SWE and runoff, and the open water budgets of NARR and ERAI, show the inadequate representation of the water cycle in the reanalyses for water budget studies, such as the analysis of a flood. With a reliable estimate of SWE and water budgets that are closed, RCM simulations seem to be suitable to study the water budgets of watersheds and to determine the cause of drought and floods. Another advantage of RCM simulations is their internal consistency in time and space compared to reanalyses, which rely on observations from stations that are sparsely spatially distributed and are available intermittently, often for short time periods. Based on physical laws, the variables from the RCMs are also consistent between them. Finally, reanalyses and gridded observations rely on individual measurements that may contain errors and that are not homogenized, thus considerably reducing the consistency among the variables that are measured.

4. Why was this flood so exceptional?

In section 3, 2-m air temperature, precipitation, and SWE from the RCM simulations gave similar results as the observationally based gridded datasets, increasing the confidence in the RCM simulations. To determine why the 2011 flood was so exceptional, the monthly time series from 1990 to 2011 of the RCMs simulations, ERAI, and CONUS were used to compute the box plots shown in Fig. 13. All the datasets show that 2-m air temperature in January, February, and March 2011 was below the median. This means that most precipitation during this period fell as snow, of which little melted, contributing to the increase of the snowpack. All the datasets show that the temperature in April 2011 was close to normal and that May 2011 was above normal. These conditions favored sudden and fast snowmelt in spring.

Fig. 13.
Fig. 13.

Box plots of the 1990–2011 monthly (top) 2-m air temperature (°C), (middle) precipitation (mm day−1), and (bottom) SWE (mm) over the U.S. part of the Richelieu basin from ALADIN, CRCM5, ERAI, and CONUS. The 2011 values are indicated with the green asterisks.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

After a dry January, precipitation in February and March 2011 was above normal in all of the datasets. This contributed to a large accumulation of snow in 2011, which is confirmed by the above-normal SWE of ALADIN and CRCM5 in February and March 2011. However, the 2011 SWE is not the highest one simulated by the RCMs for the period 1990–2011. The maximum SWE was simulated in 2008 by the RCMs. According to CONUS, April and May 2011 were the wettest months of April and May from 1990 to 2011. The RCMs simulated a wet April and May 2011, but they were not the wettest ones since 1990.

It seems that the 2011 Richelieu flood was caused by a combination of many factors. To summarize, winter 2010/11 was below normal for 2-m air temperature and above normal for precipitation. The precipitation was well above normal in April and May 2011. Taking into account the normal temperature in April and the warm May 2011, the snowpack, which was large but not the maximum during the period 1990–2011, melted rapidly. This rapid melt, in addition to the maximum precipitation in spring, explains the rise of the Lake Champlain water level and increase of Richelieu discharge, which remained very high from mid-April to mid-June. The RCMs and CONUS box plots are generally in agreement in winter and spring 2011. However, compared to CONUS, April and May 2011 of the RCM simulations were not the wettest ones. This behavior could be explained by the large internal variability of RCMs for precipitation and inherited from the ERAI lateral boundary forcing, which also underestimates precipitation in April and May 2011. The wet April and May, and the strong snowstorm in early March, can be seen in Fig. 14, which shows the 2011 daily precipitation from the RCMs, ERAI, and CONUS. In general, there is a good match between the daily precipitation of the RCMs and CONUS, with a temporal correlation above 0.8, slightly below the correlation of ERAI (0.93). The high correlation at a daily time scale gives confidence to simulate the right sequence of events that led to the flood with the RCMs.

Fig. 14.
Fig. 14.

Daily time series of the 2011 precipitation (mm day−1) over the U.S. part of the Richelieu basin from ALADIN, CRCM5, ERAI, and CONUS. A 3-day running mean was applied. The temporal correlation of the 3-day running mean field with CONUS is indicated next to the datasets at the top right corner.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

Showing the simulated daily precipitation, runoff, SWE, and 2-m air temperature from the RCMs, Fig. 15 gives an overview of what likely caused the 2011 flood. The simulated discharge by the hydrological model HSAMI described in the next section is also presented in Fig. 15. For the two RCMs, the 2-m air temperature was mainly below 0°C in January, February, and March. With the large amount of snowfall in February and especially the early March snowstorm, the mean SWE over the Richelieu basin reached 160 mm in ALADIN and 125 mm in CRCM5. The March snowstorm and the precipitation events from the beginning to the middle of March increased the river discharge from approximately 250 to 575 m3 s−1. The snow melted rapidly in April, generating runoff that contributed to a fast increase of the river discharge. High precipitation events occurred in April and May, maintaining high discharge despite the end of the melt in April. All of these factors contributed to the record river discharge from mid-April to mid-June 2011. At the beginning of June, the river discharge went down permanently when high evapotranspiration took place (not shown) in summer and the accumulated spring runoff in Lake Champlain was released by the Richelieu River.

Fig. 15.
Fig. 15.

Daily time series of the 2011 precipitation, runoff, SWE, and 2-m air temperature over the U.S. part of the Richelieu basin from (a) ALADIN and (b) CRCM5. The simulated discharge of the Richelieu River by the hydrological model HSAMI is indicated in black. A 3-day running mean is applied. Precipitation (mm day−1), runoff (mm day−1), and 2-m air temperature (°C) refer to the left axis. SWE (mm) refers to the right axis, and the discharge (m3 s−1) refers to the right-most axis.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

Variables shown in Fig. 15 are simulated by the RCMs and the hydrological model HSAMI. Thus, the sequence of events and the intensities are not expected to perfectly match reality. However, the RCMs give additional variables such as runoff and SWE that are not accurately measured over the full basin. The storage effect of Lake Champlain and the lag between the runoff generation and the river discharge increase at the lake outlet are also well simulated by the hydrological model. To reduce the complexity of the analysis, we did not focus on soil moisture and evapotranspiration, which are other variables simulated by the RCMs that probably only played a minor role in this flood.

In general, Figs. 15a and 15b are similar, but some differences can be noted. The precipitation events and intensities from the two RCM simulations are different in summer but rather similar in winter. This could be explained by the main summer convective precipitation in summer, which is linked to local factors. The large internal variability of RCMs in summer due to the slow atmospheric circulation at the lateral boundaries could also explain the divergence of the two RCMs (Lucas-Picher et al. 2008a,b). Conversely, the similar precipitation from the RCMs in winter is mainly due to the stratiform precipitation triggered by large-scale atmospheric circulation that is strongly constrained at the lateral boundaries in winter. The runoff in ALADIN is more variable and sensitive to precipitation and snowmelt than the runoff in CRCM5, depending on the land surface scheme formulation of each RCM. Overall, the 2-m air temperature is very well synchronized between the two RCMs but is higher for CRCM5 than ALADIN, which could explain the lower SWE in CRCM5.

5. Simulation of the river discharge with the hydrological model HSAMI

Another way to evaluate the RCMs’ simulations consists of comparing the river discharge simulated by a hydrological model using RCM outputs with the observed discharge. Daily minimum and maximum 2-m air temperatures and precipitation from the RCM simulations were given as input to HSAMI. Then the hydrological model was calibrated for each RCM dataset to compute the discharge at the Lake Champlain outlet, the origin of 90% of the total water of the Richelieu River basin (International Joint Commission 2013). More details on the HSAMI hydrological model and the calibration over the Richelieu basin are available in Riboust and Brissette (2014a). The same computation with outputs from ERAI and CONUS was performed to compare the simulated river discharge by HSAMI using different atmospheric forcings.

Figure 16a shows the 1990–2011 mean annual Richelieu River hydrograph computed with HSAMI using different forcings. In general, the river discharge computed using the ALADIN and CRCM5 outputs is close to the observed river discharge, as indicated by a Nash–Sutcliffe efficiency coefficient of 0.76 and 0.73, respectively, computed with the daily river discharge from 1990 to 2011. The mean maximum in spring is slightly underestimated and the mean minimum in summer is slightly overestimated using the outputs from ALADIN and CRCM5 compared to the observed values, but the results are satisfactory. The differences between the simulated and the observed river discharges could be associated with deficiencies in the atmospheric forcing and/or could be attributed to HSAMI, which is a conceptual model that introduces several approximations. The observed river discharge is also influenced by anthropogenic factors such as irrigation and dams that are not taken into account by the hydrological model. However, compared to most basins of comparable size in the United States, the Richelieu River basin has a limited amount of artificial flow regulation, such as lake outlet controls, reservoirs, and dams, facilitating the simulation of discharge with hydrological models.

Fig. 16.
Fig. 16.

Discharge (m3 s−1) of the Richelieu River at the Carignan station simulated by the hydrological model HSAMI using different forcing fields: (a) 1990–2011 mean annual cycle and (b) 2011 cycle. The numbers in parentheses correspond to the Nash–Sutcliffe efficiency coefficient of the simulated with the observed discharge computed with daily values. The Nash–Sutcliffe efficiency coefficient of (a) is computed using the daily values from 1990 to 2011.

Citation: Journal of Hydrometeorology 16, 1; 10.1175/JHM-D-14-0116.1

In Fig. 16b, the spring 2011 Richelieu River flood simulated by HSAMI with the RCM outputs is underestimated by a factor of around 30%. Conversely, the flood is well simulated using the gridded observation CONUS but is also underestimated using the ERAI outputs. The incorrect atmospheric forcing seems to be the main cause of the underestimated spring 2011 flood using the RCM outputs. From January to early April 2011, the simulated river discharges using CONUS, ALADIN, and CRCM5 are very similar. However, in April and May, the rise of the simulated river discharges using the CONUS output is much higher than using the RCMs’ outputs. Thus, the underestimation of the simulated precipitation in April and May 2011 (Fig. 13), combined with the calibration of HSAMI that considers an overestimation of the mean 1990–2011 spring precipitation (Fig. 7), probably explains the underestimation of the peak river discharge in 2011 using RCM outputs. The simulated river discharge using the ERAI output is close to the ones using the RCM outputs. Consequently, despite clear added value in the spatial distribution of the RCMs’ precipitation and 2-m air temperature, the simulation of the river discharge with RCM outputs is not significantly improved, partly because the RCM inherits the biases of its driving field. Also, it is likely that the biases of HSAMI and the combination of all the factors over the river basin reduce the improvements of the simulated river discharge with the RCM outputs that contain added value compared to ERAI.

6. Conclusions

Regional climate models have the potential to provide poor-man regional reanalyses when driven at their lateral boundaries by large-scale reanalyses (Kanamitsu and Kanamaru 2007; Stefanova et al. 2012; DiNapoli and Misra 2012). With their high resolution, RCMs simulate regional effects and provide a closed water cycle compared to reanalyses that use relatively coarse resolutions and need to be constantly reinitialized to assimilate observations. In this work, we reconstructed the sequence of events leading to the spring 2011 Richelieu River flood with two regional climate models (ALADIN and CRCM5) that were driven at their lateral boundaries with ERAI. The RCM simulations computed on a grid mesh of around 10 km exceed more than two decades, allowing the identification of anomalies that contributed to the flood. Also, both RCMs use state-of-the-art land surface schemes that improve the simulation of the water cycle. The outputs of the RCMs were also used to drive the hydrological model HSAMI to simulate the river discharge.

The simulated mean, variability, and sequence of events of the 2-m air temperature and precipitation by the RCMs are generally similar to those observed. The temporal correlation of precipitation over the Richelieu basin at a monthly and daily time scale between the RCMs and the observations is high. The spatial distribution of the 2-m air temperature and precipitation shows added value in the RCM simulations, with more precipitation and colder conditions in the mountains surrounding the Champlain valley compared to ERAI. The SWE simulated by the two RCMs is better than in the reanalysis, probably because of the more realistic topography in the RCMs, which favors snow accumulation in the mountains. Also, an analysis over the basin indicated that, unlike most reanalyses, the water budget of the RCMs is closed. The good performance of two RCMs, especially ALADIN, which is not commonly used in North America, increases the credibility and robustness of the approach.

Taking advantage of the long RCM simulations, the analysis of the 2011 anomalies with respect to 1990–2011 indicates that the Richelieu flood was caused by a combination of several factors. At the end of January, the river discharge and the snowpack were normal and there was no indication that an important flood would occur a few months later. However, the 2011 winter was colder and snowier than normal. This contributed to the accumulation of a larger amount of snow at the end of winter 2011, which was nevertheless smaller than the amount in 2008. With above-normal temperatures in spring, the large melt of the snowpack increased the Richelieu River discharge. April and May 2011 were also the wettest of the last 22 years. The large and sudden snowmelt, combined with the high precipitation in spring, led to the flood of the Richelieu River from mid-April to mid-June. The two RCMs accurately simulated the anomalies responsible for the flood, except for the precipitation in April and May 2011, which was underestimated.

Using the RCM outputs, HSAMI was able to reproduce a high river discharge in 2011, which was, however, lower than the observed discharge. Considering the good simulation of the river discharge by HSAMI for 2011 using the gridded observations, we believe that the main factor that could explain the underestimation of the flood with the RCM outputs is the underestimation of precipitation in April and May 2011. To correct this underestimation and other misrepresentations of the RCMs, it would have been possible to bias correct the RCMs’ output before feeding HSAMI. However, we preferred to directly use the raw RCM outputs to determine their direct contribution to the computation of the river discharge. HSAMI accurately modeled the time lag between the runoff generation at each grid cell and the river flow until the river mouth. The hydrological model also reproduced the damping and storing effect of Lake Champlain, which absorbed and spread a large amount of water in spring over a long period. HSAMI is a rather simple conceptual model that only uses three daily driving variables: minimum and maximum 2-m air temperatures and precipitation. While HSAMI certainly introduces some approximations during the computation of the river discharge, it is easy to use and immediately available from hydrologists that have a great experience using it. It would have been possible to use an integrated river routing scheme like the one in Lucas-Picher et al. (2003). However, directly using the river runoff from the RCMs, the use of an integrated river routing would have introduced other approximations associated with the RCM land surface schemes.

The RCM simulations of ALADIN and CRCM5 were primarily performed to provide high-resolution climatic information over Québec. While the results were satisfactory, better results could have been obtained if the experimental setup had been optimized. Indeed, it might have been preferable to use a smaller domain with the Richelieu River basin located in the middle of the domain. Also, a double nesting approach to reach an even higher resolution could have better captured the local effects mainly due to orographic forcing. It would have been interesting to explore the internal variability using an ensemble of RCM simulations started from different initial conditions. The use of a sensitivity test using lateral boundary forcings from different reanalyses to see if the RCMs systematically inherit the biases from their driving fields would have also been interesting. Finally, different experimental configurations such as the use of spectral nudging (Radu et al. 2008; Herrmann et al. 2011) or frequent reinitializations (Lucas-Picher et al. 2013a) would ensure a proximity to the reanalysis temporal chronology, while generating reliable high-resolution climate information.

This study illustrates that an RCM can be used to reconstruct climate conditions at a regional scale. RCM simulations can provide high-resolution climate information where observations are sparsely available and new variables that cannot be measured. What has been found in this study over the Richelieu basin can also be valid over other regions. Such studies can also be useful for earlier historical situations, as far back in time as reanalyses can go, when few observations were available. Because RCMs are based on physical laws, the simulated variables are coherent in space, in time, and between them, compared to observationally based gridded information and reanalyses that have to deal with sparsely distributed weather stations that have short or incomplete time series. Finally, the datasets generated and the analysis will be useful for impact and adaptation studies in Québec, where few observing stations are available, especially in the north of the province, which is becoming of great interest for economic development.

Acknowledgments

This study was supported by a postdoctoral fellowship from Québec’s “Fonds de recherche du Québec—Nature et technologies” and a visiting scientist position at Météo-France to the first author and by a grant from the Canadian Network of Centres of Excellence (NCE) “Marine Environmental Observation, Prediction and Response” (MEOPAR).

REFERENCES

  • Alexandru, A., , de Elía R. , , and Laprise R. , 2007: Internal variability in regional climate downscaling at the seasonal scale. Mon. Wea. Rev., 135, 32213238, doi:10.1175/MWR3456.1.

    • Search Google Scholar
    • Export Citation
  • Anderson, C. J., and et al. , 2003: Hydrological processes in regional climate model simulations of the central United States flood of June–July 1993. J. Hydrometeor., 4, 584598, doi:10.1175/1525-7541(2003)004<0584:HPIRCM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bisson, J. L., , and Roberge F. , 1983: Prévisions des apports naturels: Expérience d’Hydro-Québec. Proc. Workshop on Flow Prediction, Toronto, ON, Canada, IEEE, 148161.

  • Bougeault, P., 1985: A simple parameterization of the large-scale effects of cumulus convection. Mon. Wea. Rev., 113, 21082121, doi:10.1175/1520-0493(1985)113<2108:ASPOTL>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Colin, J., , Déqué M. , , Radu R. , , and Somot S. , 2010: Sensitivity study of heavy precipitations in Limited Area Model climate simulation: Influence of the size of the domain and the use of the spectral nudging technique. Tellus, 62A, 591604, doi:10.1111/j.1600-0870.2010.00467.x.

    • Search Google Scholar
    • Export Citation
  • Côté, J., , Gravel S. , , Méthot A. , , Patoine A. , , Roch M. , , and Staniforth A. , 1998: The operational CMC-MRB global environmental multiscale (GEM) model. Part I: Design considerations and formulation. Mon. Wea. Rev., 126, 13731395, doi:10.1175/1520-0493(1998)126<1373:TOCMGE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and et al. , 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597, doi:10.1002/qj.828.

    • Search Google Scholar
    • Export Citation
  • Déqué, M., , and Somot S. , 2008: Analysis of heavy precipitation for France using ALADIN RCM simulations. Idojaras, 112, 179190.

  • DiNapoli, S. M., , and Misra V. , 2012: Reconstructing the 20th century high-resolution climate of the southeastern United States. J. Geophys. Res., 117, D19113, doi:10.1029/2012JD018303.

    • Search Google Scholar
    • Export Citation
  • Douville, H., , Royer J.-F. , , and Mahfouf J.-F. , 1995: A new snow parameterization for the Météo-France climate model. Part 1: Validation in stand-alone experiments. Climate Dyn., 12, 2135, doi:10.1007/BF00208760.

    • Search Google Scholar
    • Export Citation
  • Fischer, E. M., , Seneviratne S. I. , , Vidale P. L. , , Lüthi D. , , and Schär C. , 2007: Soil moisture–atmosphere interactions during the 2003 European summer heat wave. J. Climate, 20, 50815099, doi:10.1175/JCLI4288.1.

    • Search Google Scholar
    • Export Citation
  • Flesch, T. K., , and Reuter G. W. , 2012: WRF Model simulation of two Alberta flooding events and the impact of topography. J. Hydrometeor., 13, 695708, doi:10.1175/JHM-D-11-035.1.

    • Search Google Scholar
    • Export Citation
  • Harris, I., , Jones P. D. , , Osborn T. J. , , and Lister D. H. , 2014: Updated high-resolution grids of monthly climatic observations—The CRU TS3.10 Dataset. Int. J. Climatol., 34, 623642, doi:10.1002/joc.3711.

    • Search Google Scholar
    • Export Citation
  • Hernández-Díaz, L., , Laprise R. , , Sushama L. , , Martynov A. , , Winger K. , , and Dugas B. , 2013: Climate simulation over the CORDEX-Africa domain using the fifth-generation Canadian Regional Climate Model (CRCM5). Climate Dyn., 40, 14151433, doi:10.1007/s00382-012-1387-z.

    • Search Google Scholar
    • Export Citation
  • Herrmann, M. J., , and Somot S. , 2008: Relevance of ERA40 dynamical downscaling for modeling deep convection in the Mediterranean Sea. Geophys. Res. Lett., 35, L04607, doi:10.1029/2007GL032442.

    • Search Google Scholar
    • Export Citation
  • Herrmann, M. J., , Somot S. , , Calmanti S. , , Dubois C. , , and Sevault F. , 2011: Representation of spatial and temporal variability of daily wind speed and of intense wind events over the Mediterranean Sea using dynamical downscaling: Impact of the regional climate model configuration. Nat. Hazards Earth Syst. Sci., 11, 19832001, doi:10.5194/nhess-11-1983-2011.

    • Search Google Scholar
    • Export Citation
  • International Joint Commission, 2013: Plan of study for the identification of measures to mitigate flooding and the impacts of flooding of Lake Champlain and Richelieu River. IJC Rep., 128 pp. [Available online at www.ijc.org/files/publications/Final_PoS_LakeChamplain-RichelieuRiver.pdf.]

  • Kain, J. S., , and Fritsch J. M. , 1990: A one-dimensional entraining/detraining plume model and application in convective parameterization. J. Atmos. Sci., 47, 27842802, doi:10.1175/1520-0469(1990)047<2784:AODEPM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kanamitsu, M., , and Kanamaru H. , 2007: Fifty-seven-year California reanalysis downscaling at 10 km (CaRD10). Part I: System detail and validation with observations. J. Climate, 20, 55535571, doi:10.1175/2007JCLI1482.1.

    • Search Google Scholar
    • Export Citation
  • Kanamitsu, M., , Ebisuzaki W. , , Woollen J. , , Yang S.-K. , , Hnilo J. J. , , Fiorino M. , , and Potter G. L. , 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 16311643, doi:10.1175/BAMS-83-11-1631.

    • Search Google Scholar
    • Export Citation
  • Kotlarski, S., , Hagemann S. , , Krahe P. , , Podzun R. , , and Jacob D. , 2012: The Elbe River flooding 2002 as seen by an extended regional climate model. J. Hydrol., 472–473, 169183, doi:10.1016/j.jhydrol.2012.09.020.

    • Search Google Scholar
    • Export Citation
  • Laprise, R., , Hernández-Díaz L. , , Tete K. , , Sushama L. , , Šeparović L. , , Martynov A. , , Winger K. , , and Valin M. , 2013: Climate projections over CORDEX Africa domain using the fifth-generation Canadian Regional Climate Model (CRCM5). Climate Dyn., 41, 32193246, doi:10.1007/s00382-012-1651-2.

    • Search Google Scholar
    • Export Citation
  • Leung, L. R., 2012: Regional climate models. Climate Change Modeling Methodology: Selected Entries from the Encyclopedia of Sustainability Science and Technology, P. J. Rasch, Ed., Springer, 211–233, doi:10.1007/978-1-4614-5767-1_9.

  • Leung, L. R., , and Qian Y. , 2003: The sensitivity of precipitation and snowpack simulations to model resolution via nesting in regions of complex terrain. J. Hydrometeor., 4, 10251043, doi:10.1175/1525-7541(2003)004<1025:TSOPAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Leung, L. R., , Qian Y. , , and Bian X. , 2003a: Hydroclimate of the western United States based on observations and regional climate simulations of 1981–2000. Part I: Seasonal statistics. J. Climate, 16, 18921911, doi:10.1175/1520-0442(2003)016<1892:HOTWUS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Leung, L. R., , Qian Y. , , Han J. , , and Roads J. O. , 2003b: Intercomparison of global reanalyses and regional simulations of cold season water budgets in the western United States. J. Hydrometeor., 4, 10671087, doi:10.1175/1525-7541(2003)004<1067:IOGRAR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lim, Y.-K., , Stefanova L. B. , , Chan S. C. , , Schubert S. D. , , and O’Brien J. J. , 2011: High-resolution subtropical summer precipitation derived from dynamical downscaling of the NCEP/DOE reanalysis: How much small-scale information is added by a regional model? Climate Dyn., 37, 10611080, doi:10.1007/s00382-010-0891-2.

    • Search Google Scholar
    • Export Citation
  • Livneh, B., , Rosenberg E. A. , , Lin C. , , Nijssen B. , , Mishra V. , , Andreadis K. M. , , Maurer E. P. , , and Lettenmaier D. P. , 2013: A long-term hydrologically based dataset of land surface fluxes and states for the conterminous United States: Update and extensions. J. Climate, 26, 93849392, doi:10.1175/JCLI-D-12-00508.1.

    • Search Google Scholar
    • Export Citation
  • Lorenz, C., , and Kunstmann H. , 2012: The hydrological cycle in three state-of-the-art reanalyses: Intercomparison and performance analysis. J. Hydrometeor., 13, 13971420, doi:10.1175/JHM-D-11-088.1.

    • Search Google Scholar
    • Export Citation
  • Lucas-Picher, P., , Arora V. K. , , Caya D. , , and Laprise R. , 2003: Implementation of a large-scale variable velocity river flow routing algorithm in the Canadian regional climate model. Atmos.–Ocean, 41, 139153, doi:10.3137/ao.410203.

    • Search Google Scholar
    • Export Citation
  • Lucas-Picher, P., , Caya D. , , Biner S. , , and Laprise R. , 2008a: Quantification of the lateral boundary forcing of a regional climate model using an ageing tracer. Mon. Wea. Rev., 136, 49804996, doi:10.1175/2008MWR2448.1.

    • Search Google Scholar
    • Export Citation
  • Lucas-Picher, P., , Caya D. , , de Elía R. , , and Laprise R. , 2008b: Investigation of regional climate models’ internal variability with a ten-member ensemble of ten years over a large domain. Climate Dyn., 31, 927940, doi:10.1007/s00382-008-0384-8.

    • Search Google Scholar
    • Export Citation
  • Lucas-Picher, P., , Boberg F. , , Christensen J. H. , , and Berg P. , 2013a: Dynamical downscaling with reinitializations: A method to generate finescale climate datasets suitable for impact studies. J. Hydrometeor., 14, 11591174, doi:10.1175/JHM-D-12-063.1.

    • Search Google Scholar
    • Export Citation
  • Lucas-Picher, P., , Somot S. , , Déqué M. , , Decharme B. , , and Alias A. , 2013b: Evaluation of the regional climate model ALADIN to simulate the climate over North America in the CORDEX framework. Climate Dyn., 41, 11171137, doi:10.1007/s00382-012-1613-8.

    • Search Google Scholar
    • Export Citation
  • Martynov, A., , Laprise R. , , Sushama L. , , Winger K. , , Šeparović L. , , and Dugas B. , 2013: Reanalysis-driven climate simulation over CORDEX North America domain using the Canadian Regional Climate Model, version 5: Model performance evaluation. Climate Dyn., 41, 29733005, doi:10.1007/s00382-013-1778-9.

    • Search Google Scholar
    • Export Citation
  • Mesinger, F., and et al. , 2006: North American Regional Reanalysis. Bull. Amer. Meteor. Soc., 87, 343360, doi:10.1175/BAMS-87-3-343.

    • Search Google Scholar
    • Export Citation
  • Minville, M., , Brissette F. , , and Leconte R. , 2008: Uncertainty of the impact of climate change on the hydrology of a Nordic watershed. J. Hydrol., 358, 7083, doi:10.1016/j.jhydrol.2008.05.033.

    • Search Google Scholar
    • Export Citation
  • Noilhan, J., , and Planton S. , 1989: A simple parameterization of land surface processes for meteorological models. Mon. Wea. Rev., 117, 536549, doi:10.1175/1520-0493(1989)117<0536:ASPOLS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Radu, R., , Déqué M. , , and Somot S. , 2008: Spectral nudging in a spectral regional climate model. Tellus, 60A, 898910, doi:10.1111/j.1600-0870.2008.00341.x.

    • Search Google Scholar
    • Export Citation
  • Rasmussen, R., and et al. , 2011: High-resolution coupled climate runoff simulations of seasonal snowfall over Colorado: A process study of current and warmer climate. J. Climate, 24, 30153048, doi:10.1175/2010JCLI3985.1.

    • Search Google Scholar
    • Export Citation
  • Riboust, P., , and Brissette F. , 2014a: Climate change impacts and uncertainties on spring flooding of Lake Champlain and the Richelieu River. J. Amer. Water Resour. Assoc., in press.

    • Search Google Scholar
    • Export Citation
  • Riboust, P., , and Brissette F. , 2014b: Analysis of Lake-Champlain/Richelieu River's historical 2011 flood. Can. Water Resour. J., in press.

    • Search Google Scholar
    • Export Citation
  • Ricard, J. L., , and Royer J. F. , 1993: A statistical cloud scheme for use in an AGCM. Ann. Geophys., 11, 10951115.

  • Rienecker, M. M., and et al. , 2011: MERRA: NASA’s Modern-Era Retrospective Analysis for Research and Applications. J. Climate, 24, 36243648, doi:10.1175/JCLI-D-11-00015.1.

    • Search Google Scholar
    • Export Citation
  • Rost, A., 2010: Snow board: A look at the importance of snow information. Meteorological Technology International, June 2010, UKiP, Dorking, United Kingdom, 26–29.

  • Rummukainen, M., 2010: State-of-the-art with regional climate models. Wiley Interdiscip. Rev.: Climate Change,1, 82–96, doi:10.1002/wcc.8.

  • Salzmann, N., , and Mearns L. O. , 2012: Assessing the performance of multiple regional climate model simulations for seasonal mountain snow in the upper Colorado River basin. J. Hydrometeor., 13, 539556, doi:10.1175/2011JHM1371.1.

    • Search Google Scholar
    • Export Citation
  • Sanchez-Gomez, E., , Somot S. , , and Déqué M. , 2009a: Ability of an ensemble of regional climate models to reproduce the weather regimes during the period 1961–2000. Climate Dyn., 33, 723736, doi:10.1007/s00382-008-0502-7.

    • Search Google Scholar
    • Export Citation
  • Sanchez-Gomez, E., , Somot S. , , and Mariotti A. , 2009b: Future changes in the Mediterranean water budget projected by an ensemble of regional climate models. Geophys. Res. Lett., 36, L21401, doi:10.1029/2009GL040120.

    • Search Google Scholar
    • Export Citation
  • Šeparović, L., , Alexandru A. , , Laprise R. , , Martynov A. , , Sushama L. , , Winger K. , , Tete K. , , and Valin M. , 2013: Present climate and climate change over North America as simulated by the fifth-generation Canadian Regional Climate Model (CRCM5). Climate Dyn., 41, 31673201, doi:10.1007/s00382-013-1737-5.

    • Search Google Scholar
    • Export Citation
  • Sheffield, J., , Livneh B. , , and Wood E. F. , 2012: Representation of terrestrial hydrology and large-scale drought of the continental United States from the North American Regional Reanalysis. J. Hydrometeor., 13, 856876, doi:10.1175/JHM-D-11-065.1.

    • Search Google Scholar
    • Export Citation
  • Stefanova, L., , Misra V. , , Chan S. , , Griffin M. , , O’Brien J. J. , , and Smith T. J. III, 2012: A proxy for high-resolution regional reanalysis for the southeast United States: Assessment of precipitation variability in dynamically downscaled reanalyses. Climate Dyn., 38, 24492466, doi:10.1007/s00382-011-1230-y.

    • Search Google Scholar
    • Export Citation
  • Sundqvist, H., , Berge E. , , and Kristjansson J. E. , 1989: Condensation and cloud parameterization studies with a mesoscale numerical weather prediction model. Mon. Wea. Rev., 117, 16411657, doi:10.1175/1520-0493(1989)117<1641:CACPSW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Verseghy, D. L., 2009: CLASS—The Canadian Land Surface Scheme (version 3.4): Technical documentation (version 1.1). Internal Rep., Climate Research Division, Science and Technology Branch, Environment Canada, 180 pp. [Available online at www.usask.ca/ip3/download/CLASS_v3_4_Documentation_v1_1.pdf.]

  • Voldoire, A., and et al. , 2013: The CNRM-CM5.1 global climate model: Description and basic evaluation. Climate Dyn., 40, 20912121, doi:10.1007/s00382-011-1259-y.

    • Search Google Scholar
    • Export Citation
  • von Storch, H., , Zorita E. , , and Cubasch U. , 1993: Downscaling of global climate change estimates to regional scales: An application to Iberian rainfall in wintertime. J. Climate, 6, 11611171, doi:10.1175/1520-0442(1993)006<1161:DOGCCE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
Save