• Bartlett, P. A., , MacKay M. D. , , and Verseghy D. L. , 2006: Modified snow algorithms in the Canadian land surface scheme: Model runs and sensitivity analysis at three boreal forest stands. Atmos.–Ocean, 44, 207222, doi:10.3137/ao.440301.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , Viterbo P. , , Beljaars A. , , Pan H. L. , , Hong S. Y. , , Goulden M. , , and Wofsy S. , 1998: Evaluation of land-surface interaction in ECMWF and NCEP/NCAR reanalysis models over grassland (FIFE) and boreal forest (BOREAS). J. Geophys. Res., 103 (D18), 23 07923 085, doi:10.1029/98JD02023.

    • Search Google Scholar
    • Export Citation
  • Brown, R. D., 2010: Analysis of snow cover variability and change in Quebec, 1948–2005. Hydrol. Processes,24, 1929–1954, doi:10.1002/hyp.7565.

  • Brown, R. D., , Brasnett B. , , and Robinson D. , 2003: Gridded North American monthly snow depth and snow water equivalent for GCM evaluation. Atmos.–Ocean, 41, 114, doi:10.3137/ao.410101.

    • Search Google Scholar
    • Export Citation
  • Brown, R. D., , Bartlett P. A. , , MacKay M. , , and Verseghy D. L. , 2006: Evaluation of snow cover in CLASS for SnowMIP. Atmos.–Ocean, 44, 223238, doi:10.3137/ao.440302.

    • Search Google Scholar
    • Export Citation
  • Brown, R. D., , Derksen C. , , and Wang L. , 2010: A multi-data set analysis of variability and change in Arctic spring snow cover extent, 1967–2008. J. Geophys. Res., 115, D16111, doi:10.1029/2010JD013975.

    • Search Google Scholar
    • Export Citation
  • Caya, D., , and Laprise R. , 1999: A semi-implicit, semi-Lagrangian regional climate model: The Canadian RCM. Mon. Wea. Rev., 127, 341362, doi:10.1175/1520-0493(1999)127<0341:ASISLR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Cesana, G., , and Chepfer H. , 2013: Evaluation of the cloud thermodynamic phase in a climate model using CALIPSO-GOCCP. J. Geophys. Res. Atmos., 118, 79227937, doi:10.1002/jgrd.50376.

    • Search Google Scholar
    • Export Citation
  • Chung, Y., , Belair S. , , and Mailhot J. , 2010: Simulation of snow on Arctic sea ice using a coupled snow–ice model. J. Hydrometeor., 11, 199–210, doi:10.1175/2009JHM1112.1.

    • Search Google Scholar
    • Export Citation
  • Dee, D., and Coauthors, 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
  • Déry, S. J., , and Brown R. D. , 2007: Recent Northern Hemisphere snow cover extent trends and implications for the snow-albedo feedback. Geophys. Res. Lett., 34, L22504, doi:10.1029/2007GL031474.

    • Search Google Scholar
    • Export Citation
  • Doelling, D., and Coauthors, 2013: Geostationary enhanced temporal interpolation for CERES flux products. J. Atmos. Oceanic Technol.,30, 1072–1090, doi:10.1175/JTECH-D-12-00136.1.

  • Dorsaz, F., , and Brown R. , 2008: Évaluation des simulations du couvert nival sur le Québec par les modèles MRCC 4.2.3 et GEMCLIM 3.3.0. Ouranos Rapport de stage Décembre 2008, 49 pp.

  • Dutra, E., , Balsamo G. , , Viterbo P. , , Miranda P. M. A. , , Beljaars A. , , Schär C. , , and Elder K. , 2010: An improved snow scheme for the ECMWF land surface model: Description and offline validation. J. Hydrometeor., 11, 899–916, doi:10.1175/2010JHM1249.1.

    • Search Google Scholar
    • Export Citation
  • Essery, R., , Morin S. , , Lejeune Y. , , and Ménard C. B. , 2013: A comparison of 1701 snow models using observations from an alpine site. Adv. Water Resour., 55, 131–148, doi:10.1016/j.advwatres.2012.07.013.

    • Search Google Scholar
    • Export Citation
  • Flanner, M. G., , Shell K. M. , , Barlage M. , , Perovich D. K. , , and Tschudi M. A. , 2011: Radiative forcing and albedo feedback from the Northern Hemisphere cryosphere between 1979 and 2008. Nat. Geosci., 4, 151155, doi:10.1038/ngeo1062.

    • Search Google Scholar
    • Export Citation
  • Gao, F., , Schaaf C. B. , , Strahler A. H. , , Roesch A. , , Lucht W. , , and Dickinson R. , 2005: MODIS bidirectional reflectance distribution function and albedo Climate Modeling Grid products and the variability of albedo for major global vegetation types. J. Geophys. Res., 110, D01104, doi:10.1029/2004JD005190.

    • Search Google Scholar
    • Export Citation
  • Gardner, A. S., , and Sharp M. J. , 2010: A review of snow and ice albedo and the development of a new physically based broadband albedo parameterization. J. Geophys. Res., 115, F01009, doi:10.1029/2009JF001444.

    • Search Google Scholar
    • Export Citation
  • Goulden, M. L., and Coauthors, 1998: Sensitivity of boreal forest carbon balance to soil thaw. Science, 279, 214217, doi:10.1126/science.279.5348.214.

    • Search Google Scholar
    • Export Citation
  • Groisman, P. Ya., , Karl T. R. , , and Knight R. W. , 1994: Observed impact of snow cover on the heat balance and the rise of continental spring temperatures. Science, 263, 198200, doi:10.1126/science.263.5144.198.

    • Search Google Scholar
    • Export Citation
  • Hall, A., , and Qu X. , 2006: Using the current seasonal cycle to constrain snow albedo feedback in future climate change. Geophys. Res. Lett., 33, L03502, doi:10.1029/2005GL025127.

    • Search Google Scholar
    • Export Citation
  • Hall, D. K., , Riggs G. A. , , Salomonson V. V. , , DiGirolamo N. E. , , and Bayr K. J. , 2002: MODIS snow-cover products. Remote Sens. Environ., 83, 181194, doi:10.1016/S0034-4257(02)00095-0.

    • Search Google Scholar
    • Export Citation
  • Hall, D. K., , Riggs G. A. , , and Salomonson V. V. , 2006: MODIS/Terra snow cover monthly L3 global 0.05Deg CMG, version 5. National Snow and Ice Data Center, Boulder, CO, digital media. [Available online at http://nsidc.org/data/docs/daac/modis_v5/mod10cm_modis_terra_snow_monthly_global_0.05deg_cmg.gd.html.]

  • Hansen, M., , DeFries R. , , Townshend J. , , Carroll M. , , Dimiceli C. , , and Sohlberg R. , 2003: Global percent tree cover at a spatial resolution of 500 meters: First results of the MODIS vegetation continuous fields algorithm. Earth Interact., 7 (10), 115, doi:10.1175/1087-3562(2003)007<0001:GPTCAA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Helfrich, S. R., , McNamara D. , , Ramsay B. H. , , Baldwin T. , , and Kasheta T. , 2007: Enhancements to, and forthcoming developments in the Interactive Multisensor Snow and Ice Mapping System (IMS). Hydrol. Processes, 21, 15761586, doi:10.1002/hyp.6720.

    • Search Google Scholar
    • Export Citation
  • Kato, S., , Loeb N. G. , , Rose F. G. , , Doelling D. R. , , Rutan D. A. , , Caldwell T. E. , , Yu L. , , and Weller R. A. , 2013: Surface irradiances consistent with CERES-derived top-of-atmosphere shortwave and longwave irradiances. J. Climate, 26, 27192740, doi:10.1175/JCLI-D-12-00436.1.

    • Search Google Scholar
    • Export Citation
  • Key, J. R., , Wang X. J. , , Stroeve J. C. , , and Fowler C. , 2001: Estimating the cloudy-sky albedo of sea ice and snow from space. J. Geophys. Res., 106 (D12), 12 48912 497, doi:10.1029/2001JD900069.

    • Search Google Scholar
    • Export Citation
  • Kimball, J. S., , McDonald K. C. , , Running S. W. , , and Frolking S. E. , 2004: Satellite radar remote sensing of seasonal growing seasons for boreal and subalpine evergreen forests. Remote Sens. Environ., 90, 243258, doi:10.1016/j.rse.2004.01.002.

    • Search Google Scholar
    • Export Citation
  • Liston, G. E., 2004: Representing subgrid snow cover heterogeneities in regional and global models. J. Climate, 17, 13811397, doi:10.1175/1520-0442(2004)017<1381:RSSCHI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lucht, W., , Schaaf C. B. , , and Strahler A. H. , 2000: An algorithm for the retrieval of albedo from space using semiempirical BRDF models. IEEE Trans. Geosci. Remote Sens., 38, 977998, doi:10.1109/36.841980.

    • Search Google Scholar
    • Export Citation
  • MacKay, M. D., , Seglenieks F. , , Verseghy D. , , Soulis E. D. , , Snelgrove K. R. , , Walker A. , , and Szeto K. , 2003: Modeling Mackenzie basin surface water balance during CAGES with the Canadian Regional Climate Model. J. Hydrometeor., 4, 748767, doi:10.1175/1525-7541(2003)004<0748:MMBSWB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • MacKay, M. D., , Bartlett P. , , Chan E. , , Dersen C. , , Guo S. , , and Leighton H. , 2006: On the simulation of regional scale sublimation over boreal and agricultural landscapes in a climate model. Atmos.–Ocean, 44, 289304, doi:10.3137/ao.440306.

    • Search Google Scholar
    • Export Citation
  • Male, D. H., , and Granger R. J. , 1981: Snow surface energy exchange. Water Resour. Res., 17, 609627, doi:10.1029/WR017i003p00609.

  • Markovic, M., , Jones C. G. , , Vaillancourt P. A. , , Paquin D. , , Winger K. , , and Paquin-Ricard D. , 2008: An evaluation of the surface radiation budget over North America for a suite of regional climate models against surface station observations. Climate Dyn., 31, 779–794, doi:10.1007/s00382-008-0378-6.

    • Search Google Scholar
    • Export Citation
  • Mellor, M., 1977: Engineering properties of snow. J. Glaciol., 19, 1566.

  • Milewska, E. J., , Hopkinson R. F. , , and Niitsoo A. , 2005: Evaluation of geo-referenced grids of 1961–1990 Canadian temperature and precipitation normals. Atmos.–Ocean, 43, 4975, doi:10.3137/ao.430104.

    • Search Google Scholar
    • Export Citation
  • Moody, E. G., , King M. D. , , Schaaf C. B. , , Hall D. K. , , and Platnick S. , 2007: Northern Hemisphere five-year average (2000–2004) spectral albedos of surfaces in the presence of snow: Statistics computed from Terra MODIS land products. Remote Sens. Environ., 111, 337345, doi:10.1016/j.rse.2007.03.026.

    • Search Google Scholar
    • Export Citation
  • Mortin, J., , Schrøder T. M. , , Walløe Hansen A. , , Holt B. , , and McDonald K. C. , 2012: Mapping of seasonal freeze-thaw transitions across the pan-Arctic land and sea ice domains with satellite radar. J. Geophys. Res., 117, C08004, doi:10.1029/2012JC008001.

    • Search Google Scholar
    • Export Citation
  • Music, B., , and Caya D. , 2007: Evaluation of the hydrological cycle over the Mississippi River basin as simulated by the Canadian Regional Climate Model (CRCM). J. Hydrometeor., 8, 969988, doi:10.1175/JHM627.1.

    • Search Google Scholar
    • Export Citation
  • Oleson, K. W., , Bonan G. B. , , Schaaf C. B. , , Gao F. , , Jin Y. , , and Strahler A. H. , 2003: Assessment of global climate model land surface albedo using MODIS data. Geophys. Res. Lett., 30, 1443, doi:10.1029/2002GL016749.

    • Search Google Scholar
    • Export Citation
  • Pedersen, C. A., , and Winther J. G. , 2005: Intercomparison and validation of snow albedo parameterization schemes in climate models. Climate Dyn., 25, 351362, doi:10.1007/s00382-005-0037-0.

    • Search Google Scholar
    • Export Citation
  • Platnick, S., , King M. D. , , Ackerman S. A. , , Menzel W. P. , , Baum B. A. , , Riedi J. C. , , and Frey R. A. , 2003: The MODIS cloud products: Algorithms and examples from Terra. IEEE Trans. Geosci. Remote Sens., 41, 459473, doi:10.1109/TGRS.2002.808301.

    • Search Google Scholar
    • Export Citation
  • Pomeroy, J. W., , Gray D. M. , , Shook K. R. , , Toth B. , , Essery R. L. H. , , Pietroniro A. , , and Hedstrom N. , 1998: An evaluation of snow accumulation and ablation processes for land surface modeling. Hydrol. Processes, 12, 23392367, doi:10.1002/(SICI)1099-1085(199812)12:15<2339::AID-HYP800>3.0.CO;2-L.

    • Search Google Scholar
    • Export Citation
  • Qu, X., , and Hall A. , 2007: What controls the strength of snow-albedo feedback? J. Climate, 20, 39713981, doi:10.1175/JCLI4186.1.

  • Ramsay, B., 1998: The interactive multisensor snow and ice mapping system. Hydrol. Processes, 12, 15371546, doi:10.1002/(SICI)1099-1085(199808/09)12:10/11<1537::AID-HYP679>3.0.CO;2-A.

    • Search Google Scholar
    • Export Citation
  • Riette, S., , and Caya D. , 2002: Sensitivity of short simulations to the various parameters in the new CRCM spectral nudging. Research activities in atmospheric and oceanic modelling, H. Ritchie, Ed., WMO/TD 1105, Rep. 32, 7.39–7.40. [Available online at http://www.wcrp-climate.org/WGNE/BlueBook/2002/individual-articles/07_s_bastien_riette.pdf.]

  • Roe, G. H., , and Baker M. B. , 2007: Why is climate sensitivity so unpredictable? Science, 318, 629632, doi:10.1126/science.1144735.

  • Roesch, A., , Wild M. , , Gilgen H. , , and Ohmura A. , 2001: A new snow cover fraction parameterization for the ECHAM4 GCM. Climate Dyn., 17, 933946, doi:10.1007/s003820100153.

    • Search Google Scholar
    • Export Citation
  • Roman, M. O., and Coauthors, 2009: The MODIS (Collection V005) BRDF/albedo product: Assessment of spatial representativeness over forested landscapes. Remote Sens. Environ., 113, 24762498, doi:10.1016/j.rse.2009.07.009.

    • Search Google Scholar
    • Export Citation
  • Rutan, D., , Rose F. , , Roman M. , , Manalo-Smith N. , , Schaaf C. , , and Charlock T. , 2009: Development and assessment of broadband surface albedo from Clouds and the Earth’s Radiant Energy System Clouds and Radiation Swath data product. J. Geophys. Res., 114, D08125, doi:10.1029/2008JD010669.

    • Search Google Scholar
    • Export Citation
  • Salomon, J. G., , Schaaf C. B. , , Strahler A. H. , , Gao F. , , and Jin Y. F. , 2006: Validation of the MODIS bidirectional reflectance distribution function and albedo retrievals using combined observations from the Aqua and Terra platforms. IEEE Trans. Geosci. Remote Sens., 44, 15551565, doi:10.1109/TGRS.2006.871564.

    • Search Google Scholar
    • Export Citation
  • Salzen, K. V., and Coauthors, 2013: The Canadian Fourth Generation Atmospheric Global Climate Model (CanAM4). Part I: Representation of physical processes. Atmos.–Ocean, 51, 104125, doi:10.1080/07055900.2012.755610.

    • Search Google Scholar
    • Export Citation
  • Schaaf, C. B., and Coauthors, 2002: First operational BRDF, albedo nadir reflectance products from MODIS. Remote Sens. Environ., 83, 135148, doi:10.1016/S0034-4257(02)00091-3.

    • Search Google Scholar
    • Export Citation
  • Schaaf, C. B., , Wang Z. , , and Strahler A. H. , 2011: Commentary on Wang and Zender—MODIS snow albedo bias at high solar zenith angles relative to theory and to in situ observations in Greenland. Remote Sens. Environ., 115, 12961300, doi:10.1016/j.rse.2011.01.002.

    • Search Google Scholar
    • Export Citation
  • Slater, A. G., and Coauthors, 2001: The representation of snow in land surface schemes: Results from PILPS 2(d). J. Hydrometeor., 2, 725, doi:10.1175/1525-7541(2001)002<0007:TROSIL>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Stroeve, J., , Box J. E. , , Gao F. , , Liang S. L. , , Nolin A. , , and Schaaf C. , 2005: Accuracy assessment of the MODIS 16-day albedo product for snow: Comparisons with Greenland in situ measurements. Remote Sens. Environ., 94, 4660, doi:10.1016/j.rse.2004.09.001.

    • Search Google Scholar
    • Export Citation
  • Sturm, M., , Holmgren J. , , König M. , , and Morris K. , 1997: The thermal conductivity of seasonal snow. J. Glaciol., 43, 2641.

  • Verseghy, D., 1991: Class—A Canadian land surface scheme for GCMs. I. Soil model. Int. J. Climatol., 11, 111133, doi:10.1002/joc.3370110202.

    • Search Google Scholar
    • Export Citation
  • Verseghy, D., , McFarlane N. , , and Lazare M. , 1993: Class—A Canadian land surface scheme for GCMs. II: Vegetation model and coupled runs. Int. J. Climatol., 13, 347370, doi:10.1002/joc.3370130402.

    • Search Google Scholar
    • Export Citation
  • Viterbo, P., , and Betts A. K. , 1999: Impact on ECMWF forecasts of changes to the albedo of the boreal forests in the presence of snow. J. Geophys. Res., 104 (D22), 27 80327 810, doi:10.1029/1998JD200076.

    • Search Google Scholar
    • Export Citation
  • Wang, L., , Derksen C. , , and Brown R. , 2008: Detection of pan-Arctic terrestrial snowmelt from QuikSCAT, 2000–2005. Remote Sens. Environ., 112, 37943805, doi:10.1016/j.rse.2008.05.017.

    • Search Google Scholar
    • Export Citation
  • Wang, L., , Wolken G. J. , , Sharp M. J. , , Howell S. E. L. , , Derksen C. , , Brown R. D. , , Markus T. , , and Cole J. , 2011: Integrated pan-Arctic melt onset detection from satellite active and passive microwave measurements, 2000–2009. J. Geophys. Res., 116, D22103, doi:10.1029/2011JD016256.

    • Search Google Scholar
    • Export Citation
  • Wang, Z., , and Zeng X. B. , 2010: Evaluation of snow albedo in land models for weather and climate studies. J. Appl. Meteor. Climatol., 49, 363380, doi:10.1175/2009JAMC2134.1.

    • Search Google Scholar
    • Export Citation
  • Wang, Z., and Coauthors, 2012: Evaluation of Moderate-Resolution Imaging Spectroradiometer (MODIS) snow albedo product (MCD43A) over tundra. Remote Sens. Environ., 117, 264–280, doi:10.1016/j.rse.2011.10.002.

    • Search Google Scholar
    • Export Citation
  • Zhang, X., , Vincent L. A. , , Hogg W. D. , , and Niitsoo A. , 2000: Temperature and precipitation trends in Canada during the 20th century. Atmos.–Ocean, 38, 395429, doi:10.1080/07055900.2000.9649654.

    • Search Google Scholar
    • Export Citation
  • Zhou, L., and Coauthors, 2003: Comparison of seasonal and spatial variations of albedos from Moderate-Resolution Imaging Spectroradiometer (MODIS) and Common Land Model. J. Geophys. Res., 108, 4488, doi:10.1029/2002JD003326.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    CRCM4 model domain (black outline) and the study area (red) in Québec overlaid on tree fraction (%) from MODIS (Hansen et al. 2003). The black dots show the location of surface stations used to verify the QSCAT MOD method and for plots in Fig. 11. The open diamonds show locations of surface stations used for extracting cloud fraction. The blue triangle represents the location of the grid cell shown in Fig. 2.

  • View in gallery

    Daily surface albedo from CRCM4 simulations of AIL, AIO, and AIW for a grid cell in the northern Québec tundra region (see location in Fig. 1) during 2000–02.

  • View in gallery

    Québec regionally averaged bias in screen level monthly (a) minimum, (b) maximum, and (c) mean air temperature (°C) in CRCM4 simulations: AIL, AIO, and AIW. Observed temperatures were obtained from the CANGRD dataset.

  • View in gallery

    Québec regionally averaged (red outlined area in Fig. 1) (a) monthly mean snow cover fraction from MODIS, IMS, and CRCM4; (b) monthly mean surface albedo from MODIS and CRCM4 under clear sky; (c) melt- onset date (day of year) from QSCAT and CRCM4; and (d) seasonal cycle of cloud fraction from observations and CRCM4.

  • View in gallery

    Maps of snow cover fraction in (top) September and (bottom) October from (a) MODIS, (b) AIL, (c) AIO, and (d) AIW.

  • View in gallery

    Differences in monthly surface albedo between CRCM4 and MODIS (CRCM4 − MODIS): (a) AIL, (b) AIO, and (c) AIW. (d) Monthly surface albedo from MODIS. (left to right) March–June.

  • View in gallery

    (a) Mean observed melt-onset date (day of year) from QSCAT. (b)–(d) CRCM4 bias (CRCM4 − QSCAT, days) for simulations of AIL, AIO, and AIW, respectively.

  • View in gallery

    Québec regionally averaged monthly differences in simulated and observed (top) TOA reflected and (bottom) surface incoming SW fluxes (CRCM4 − OBS) under (a),(c) clear-sky and (b),(d) all-sky conditions for: AIL, AIO, and AIW.

  • View in gallery

    As in Fig. 8, but for surface-absorbed SW fluxes.

  • View in gallery

    As in Fig. 8, but for TOA outgoing and surface incoming LW fluxes.

  • View in gallery

    Simulated and observed Tmin for the day before snow onset in (a) AIL and (b) AIW. The solid black line represents the 1:1 line, and the horizontal dashed black line in (b) represents 6°C, below which snowfall is possibly being partitioned in AIW. Results for AIO (not shown) are similar to AIW.

  • View in gallery

    Québec regionally averaged bias in screen level mean air temperature in CRCM4 simulations for AJO, AJP, and AIW. Simulations AJO and AJP have similar model configurations as in AIW but different precipitation phase options: 0°C and 0°–2°C polynomial functions, respectively. Observed temperatures were obtained from the CANGRD dataset.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 18 18 4
PDF Downloads 8 8 0

Application of Satellite Data for Evaluating the Cold Climate Performance of the Canadian Regional Climate Model over Québec, Canada

View More View Less
  • 1 Climate Research Division, Environment Canada, Toronto, Ontario, Canada
  • | 2 Climate Research Division, Environment Canada, Ouranos, Montreal, Québec, Canada
  • | 3 Climate Research Division, Environment Canada, Toronto, Ontario, Canada
  • | 4 Canadian Centre for Climate Modelling and Analysis, Environment Canada, Ouranos, Montreal, Québec, Canada
  • | 5 Centre d’Applications et de Recherches en Télédétection, Université de Sherbrooke, Sherbrooke, Québec, Canada
© Get Permissions
Full access

Abstract

This study evaluates key aspects of the snow cover, cloud cover, and radiation budget simulated by the Canadian Regional Climate Model, version 4 (CRCM4), coupled with two versions of the Canadian Land Surface Scheme (CLASS). CRCM4 coupled with CLASS version 2.7 has been used operationally at Ouranos since 2006, while, more recently, CRCM4 has been coupled experimentally with CLASS 3.5, which includes a number of improvements to the representation of snow cover processes. The simulations showed evidence of a systematic cold temperature bias. Evaluation of cloud cover and radiation fluxes with satellite data suggests this bias is related to insufficient cloud radiative forcing from a combination of underestimated cloud cover, excessive cloud albedo, and too low cloud emissivity in the model. This cold bias is reinforced by a positive snow albedo feedback manifest through earlier snow cover onset in the fall and early winter period. Snow albedo was found to be very sensitive to the treatment of albedo refresh but insignificantly influenced by the partitioning of solid precipitation in CLASS. This study demonstrates that atmospheric forcing can exert a significant impact on the simulation of snow cover and surface albedo. The results highlight the need to evaluate parameterizations in land surface models designed for climate models in fully coupled mode.

Corresponding author address: Libo Wang, Research Scientist, Climate Research Division, Science and Technology Branch, Environment Canada, 4905 Dufferin St., Toronto, ON M3H 5T4, Canada. E-mail: libo.wang@ec.gc.ca

Abstract

This study evaluates key aspects of the snow cover, cloud cover, and radiation budget simulated by the Canadian Regional Climate Model, version 4 (CRCM4), coupled with two versions of the Canadian Land Surface Scheme (CLASS). CRCM4 coupled with CLASS version 2.7 has been used operationally at Ouranos since 2006, while, more recently, CRCM4 has been coupled experimentally with CLASS 3.5, which includes a number of improvements to the representation of snow cover processes. The simulations showed evidence of a systematic cold temperature bias. Evaluation of cloud cover and radiation fluxes with satellite data suggests this bias is related to insufficient cloud radiative forcing from a combination of underestimated cloud cover, excessive cloud albedo, and too low cloud emissivity in the model. This cold bias is reinforced by a positive snow albedo feedback manifest through earlier snow cover onset in the fall and early winter period. Snow albedo was found to be very sensitive to the treatment of albedo refresh but insignificantly influenced by the partitioning of solid precipitation in CLASS. This study demonstrates that atmospheric forcing can exert a significant impact on the simulation of snow cover and surface albedo. The results highlight the need to evaluate parameterizations in land surface models designed for climate models in fully coupled mode.

Corresponding author address: Libo Wang, Research Scientist, Climate Research Division, Science and Technology Branch, Environment Canada, 4905 Dufferin St., Toronto, ON M3H 5T4, Canada. E-mail: libo.wang@ec.gc.ca

1. Introduction

Surface albedo is the fraction of incident solar energy reflected from the surface and defines the lower boundary for atmospheric radiative transfer (Rutan et al. 2009). Snow has a much higher albedo than other land surface types, explaining why the spatial and temporal variations of snow cover account for most of the variations in surface albedo at the local and hemispheric scales (Moody et al. 2007; Zhou et al. 2003). Therefore, accurate simulations of snow cover and snow albedo are critical in weather and climate prediction systems (Dutra et al. 2010; Viterbo and Betts 1999). The snow albedo positive feedback mechanism is recognized as a key process contributing to the amplified warming over northern latitudes (Déry and Brown 2007; Groisman et al. 1994; Flanner et al. 2011). However, the strength of the simulated snow albedo feedback exhibits a factor-of-three spread among the suite of general circulation models used in the Fourth Assessment Report of the Intergovernmental Panel on Climate Change because of differences between the models in the simulated surface albedo of snow-covered regions (Hall and Qu 2006; Qu and Hall 2007). Further, Roe and Baker (2007) showed that any uncertainty in the magnitude of a climate feedback decreases our ability to reduce uncertainty in global climate sensitivity. Moreover, Essery et al. (2013) showed that realistic representation of albedo was a key requirement for consistent model simulation of snow cover over different winter conditions.

Snow albedo has typically been empirically parameterized in land surface models (LSMs), which provide the lower boundary conditions to the atmosphere in climate models. As pointed out in Gardner and Sharp (2010), such parameterizations are often highly simplistic and often based on statistical fits to limited observations representative of characteristics of the snow and atmosphere for specific time periods and locations. When modeled albedo is compared to observations, relatively large biases are found in some of the commonly used LSMs (Dutra et al. 2010; Oleson et al. 2003; Slater et al. 2001; Wang and Zeng 2010; Zhou et al. 2003). For example, Oleson et al. (2003) found that, relative to observation from the Moderate Resolution Imaging Spectroradiometer (MODIS), the Community Land Model, version 2 (CLM2), overestimated albedo by 20% in snow-covered regions.

Snow depth can vary considerably at subgrid scales of regional or global climate models because of heterogeneities in land cover, terrain, and meteorological conditions (Liston 2004). This heterogeneity has a major impact on the snow cover fraction (SCF) and surface albedo of a grid cell. Most climate models only treat some of this heterogeneity; for example, CLASS takes account of different land cover types within a grid cell when estimating the gridcell SCF. Realistic simulation of SCF has been shown to play a major role in the accuracy of simulated surface albedo (Roesch et al. 2001; Pedersen and Winther 2005). Accurate representation of the timing of snow onset and snowmelt is also critical for realistic surface albedo and energy balance simulations during the transition seasons and for accurate simulations of the growing season and terrestrial carbon budget (Betts et al. 1998; Goulden et al. 1998).

The Canadian Regional Climate Model [CRCM, version 4 (CRCM4); Music and Caya 2007; Caya and Laprise 1999] is the current operational model used to produce climate change simulations for the Québec and North America domains at the Ouranos Consortium (Montreal, Canada). The Canadian Land Surface Scheme (CLASS), version 2.7, is the current land surface component of the CRCM4 (Verseghy 1991; Verseghy et al. 1993). More recent versions of CLASS (version 3.0 and higher) have addressed a number of weaknesses in the treatment of snow-related processes such as snow aging, canopy interception, and unloading (Bartlett et al. 2006; Brown et al. 2006). Previous offline evaluations have shown that CLASS provides realistic snow cover simulations in a number of different climates. However, CRCM4 has received only limited evaluation over northern Québec, in part because of a lack of evaluation data. Dorsaz and Brown (2008) were unable to reach any conclusions about the performance of CRCM4 over northern Québec because of a lack of observational data as well as major disagreements between the available evaluation datasets. A number of high-quality products from instruments on board the National Aeronautics and Space Administration’s (NASA) Earth Observing System are now available for over a decade. New satellite-derived datasets have been used in climate model evaluations (Cesana and Chepfer 2013; Salzen et al. 2013; Wang et al. 2011).

CLASS 3.5 is used in the current research version of CRCM4 at Ouranos. The purpose of this paper is to examine the effect of different treatments of snow processes in CRCM4–CLASS 2.7 and CRCM4–CLASS 3.5 at a larger scale than previous offline studies using mainly new satellite-derived observations. While evaluations of standalone simulations of land surface schemes are very common, this study uses fully coupled CRCM4 simulations so that interactions/feedbacks between the atmosphere and the snow-covered surface can be addressed.

Cloud radiative forcing plays a critical role in determining the surface energy balance over snow (Male and Granger 1981). One important interaction between the CRCM4 and CLASS is the exchange of radiation at the surface–atmosphere interface. To understand better evaluation results for snow-related processes, we also compare modeled cloud cover and radiation fluxes with satellite observations. An assessment of snow-related performance of CRCM4 based primarily on in situ observations has been carried out by Langlois et al. (2013, manuscript submitted to J. Hydrometeor.).

A brief description of the CRCM4 simulations, snow treatments in CLASS, and observational datasets are presented in section 2. Evaluation of surface air temperature, snow cover, surface albedo, and snowmelt onset are given in section 3, and cloud cover and radiation fluxes are examined in section 4. Discussion of the results and main conclusions are presented in section 5.

2. Data and methods

a. Study area

Figure 1 shows the whole model domain [Québec (QC)] used for the CRCM4 simulations, which include all of the eastern Canadian provinces and parts of Manitoba and Nunavut. Our main focus is the region occupied by the province of Québec and adjacent Labrador from Newfoundland and Labrador (the red outline in Fig. 1). This region (which will be referred to hereafter as Québec) is snow covered for 4–8 months each year, with annual mean maximum snow water equivalent values ranging from less than 100 mm in northern Québec to more than 300 mm in southern Québec (Brown et al. 2003; Brown 2010). The evaluation period was limited to 2000–09 because of the availability of the satellite datasets. In this paper all maps will show results for the entire QC model domain, but the statistics will only be computed for the Québec and Labrador subregion outlined in red on Fig. 1.

Fig. 1.
Fig. 1.

CRCM4 model domain (black outline) and the study area (red) in Québec overlaid on tree fraction (%) from MODIS (Hansen et al. 2003). The black dots show the location of surface stations used to verify the QSCAT MOD method and for plots in Fig. 11. The open diamonds show locations of surface stations used for extracting cloud fraction. The blue triangle represents the location of the grid cell shown in Fig. 2.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

b. CRCM4 simulation setup and treatment of snow in CLASS

The CRCM4 was run over the QC domain (111 × 87 grid points) with a horizontal resolution of 45 km (true at 60°N) in a polar stereographic projection. The simulation was driven at its lateral boundaries by time-dependent atmospheric fields taken from the Interim European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-Interim; Dee et al. 2011). A spectral nudging technique was applied to large-scale winds (Riette and Caya 2002) within the interior of the regional domain to keep the CRCM4’s large-scale flow close to its driving data. A series of experiments with different model configurations was performed and three CRCM4 simulations were selected for our evaluation (Table 1). These include the current operational CRCM4 run with CLASS 2.7 (hereafter named AIL) and two experimental CRCM4 runs with CLASS 3.5 (hereafter named AIO and AIW; Table 1). In these simulations, the atmospheric component of the CRCM4 only provides the total precipitation and the partitioning of snow or rain is provided by CLASS. Here we briefly summarize the treatment of snow cover in CLASS and the main differences in CLASS 3.5 versus CLASS 2.7.

Table 1.

Summary of the main differences between the three CRCM4 simulations analyzed in this study. The CRCM4 simulations were run for the 1989–2010 period.

Table 1.

CLASS is a physically based land surface scheme with complete thermal and hydrological budgets. It includes three soil layers, a single snow layer with variable depth, and a vegetation canopy. It has a simple treatment of subgrid-scale heterogeneity by dividing each grid cell into up to four subareas (vegetated and bare soil, each with and without snow cover), which are treated separately. Four broad vegetation classes are considered: needleleaf trees, broadleaf trees, crops, and grass. Canopy snow processes such as interception/unloading, sublimation, and melt are included. In CLASS 2.7, 0°C air temperature at the lowest atmospheric model layer is used to partition precipitation into rain or snow. CLASS 3.1 and later versions provide an option for using a polynomial function to partition the fraction of precipitation that is snow, allowing for a range of mixed precipitation between 0° and 2°C or between 0° and 6°C (Bartlett et al. 2006). The latter option is used in the two experimental runs (AIO and AIW; Table 1). Fractional snow cover is assumed to be 100% if the snow depth (Z) reaches 0.1 m. Otherwise, the snow depth is fixed at 0.1 m and the fractional snow cover is computed as Z/0.1 m.

CLASS assumes a broadband fresh snow albedo of 0.84 (0.95 in the visible and 0.73 in the near-infrared wavelengths), which decreases exponentially in time to either a dry (0.7) or melting snow (0.5) minimum albedo value due to snow metamorphism and melt (Verseghy 1991). When there is new snowfall and the depth is greater than a threshold (hereafter named the albedo refreshment threshold) Salb, CLASS refreshes the snow albedo to 0.84. The value of Salb is set to 0.0013 (AIL), 0.005 (AIO), and 0.0001 m (AIW) in the three simulations (Table 1). The latter two values of Salb were selected because Salb = 0.005 m was initially recommended for use in CLASS 3.5; however, subsequent offline simulations showed that Salb = 0.0001 m gave the best overall snow albedo over eastern Canada (D. Verseghy 2013, personal communication). The value used in AIL is the default provided with CLASS 2.7, which performed well in the Snow Model Intercomparison Project (Brown et al. 2006). The refresh depth plays a critical role in the CRCM4 simulated albedo, as is evident in Fig. 2, where in some winters Salb = 0.005 m almost never refreshed snow albedo while Salb = 0.0001 leads to frequent albedo refresh.

Fig. 2.
Fig. 2.

Daily surface albedo from CRCM4 simulations of AIL, AIO, and AIW for a grid cell in the northern Québec tundra region (see location in Fig. 1) during 2000–02.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

c. Observed data

The CANGRD gridded monthly surface air temperature dataset (Zhang et al. 2000; Milewska et al. 2005) was used to evaluate daily minimum and maximum (Tmin and Tmax) screen-level air temperature simulated by the CRCM4 following MacKay et al. (2003, 2006). The monthly SCF was obtained from the MODIS/Terra snow cover monthly L3 0.05° Climate Modeling Grid (CMG) dataset (MOD10CM, collection5) available from the National Snow and Ice Data Center (Hall et al. 2006). This dataset provides mean monthly SCF based on daily global snow cover fractional information derived from a Normalized Difference Snow Index at a spatial resolution of 500 m (Hall et al. 2002). According to Dorothy Hall (2013, personal communication), this version of MODIS snow products may underestimate SCF in areas with complex terrain during the spring melt period because of the use of temperature information in the algorithm. Thus, SCF from the National Oceanic and Atmospheric Administration (NOAA) Interactive Multisensor Snow and Ice Mapping System (IMS) was also used in the comparison. The IMS dataset consists of binary snow/no snow information on a 24-km resolution polar stereographic projection grid (Helfrich et al. 2007; Ramsay 1998). Daily IMS data were converted to monthly snow cover duration fraction (SCF = total number of days with snow cover in a month divided by the number of days in the month). A multidataset analysis study by Brown et al. (2010) indicated that the IMS dataset showed delayed snowmelt during the spring period in the Arctic. Therefore, estimates from IMS and MODIS should provide the upper and the lower limits, respectively, of SCF over the ablation period. The MODIS and IMS data were spatially aggregated to the CRCM4 grid at 45-km resolution (using only data flagged as good quality for MODIS). At least 4 yr of data were required at each grid cell to compute the average SCF over the 2000–09 period.

Surface albedo was obtained from the MODIS Bidirectional Reflectance Distribution Function (BRDF) albedo product produced from cloud-free, multiangle, and high-quality surface reflectance data over every 8 days with 16 days of data acquisition (Schaaf et al. 2002). If sufficient observations are available to sample the surface reflectance anisotropy during a 16-day period, a full model inversion is attempted. For periods with insufficient or poor sampling, a lower-quality backup magnitude inversion is performed (Schaaf et al. 2002). While the backup magnitude inversion often produces acceptable values, the albedo values obtained are more likely to be affected by residual cloud contamination and atmospheric turbidity and are thus less consistent. The MODIS BRDF/albedo products provide both black-sky (direct beam contribution) and white-sky (entire diffuse contribution) albedos in seven spectral bands and three broad bands. Actual albedos can be derived from a linear combination of white- and black-sky albedos, depending on the fraction of diffuse skylight (Lucht et al. 2000). Data from the Terra and Aqua satellites are combined in the generation of the product at 500-m resolution in a sinusoidal projection. In addition, a CMG product is also provided in a geographic latitude/longitude projection at 0.05° resolution (Gao et al. 2005). This reduced resolution product provides the average of all pixels of all qualities in the 0.05° grid box. The quality flag only presents the most frequently occurring flag value of the underlying 500-m pixels. The accuracy of the product has been evaluated using field measurements at several locations, and in general, the accuracy of the MODIS albedo is within 5% of in situ observations (Roman et al. 2009; Salomon et al. 2006; Stroeve et al. 2005; Wang et al. 2012).

The white-sky albedo represents the integration of the black-sky albedo over all solar zenith angles and should be close to the modeled albedo. Since the fraction of diffuse skylight is unknown, white-sky albedo in the visible-infrared broad band (0.3–5.0 μm) from the MODIS 0.05° CMG-grid product (MCD43C3) is used in this study as in previous studies (e.g., Dutra et al. 2010). The MODIS data were spatially aggregated to the CRCM4 grid. Ideally, we should use good-quality flagged data only, for example, grid cells with QC = 0 and QC = 1, indicating 75% or more of the underlying 500-m pixels with full inversions. However, this results in large data gaps in our study area. Therefore, we also included data flagged as mixed quality (QC = 2, 75% or less full inversions and 25% or less fill values). Doing so was not found to make a significant difference in the albedo results (not shown). Data were not used in the period November–January, when the solar zenith angle in the Québec region is often 70° or larger and the MODIS albedo products are not considered reliable (Schaaf et al. 2011; Stroeve et al. 2005).

For satellite comparison, surface albedo at each grid cell was computed from CRCM4 output fields of downward solar incident flux at the surface (FSS) and downward net solar flux absorbed at the surface (FSG): α = (FSS − FSG)/FSS. Since snow albedo is on average 4%–6% higher under cloudy sky than clear-sky conditions (Key et al. 2001) and the MODIS albedo products are based on clear-sky observations only (Schaaf et al. 2002), we extracted surface albedo from the CRCM4 under clear skies where a clear sky was defined as <5% simulated total cloud amount.

The spring snowmelt onset information was obtained from the SeaWinds scatterometer on board the QuikSCAT (hereafter QSCAT) satellite, which has data from June 1999 to November 2009. Because of its high sensitivity to the appearance of liquid water in snow and all-weather capability, time series of QSCAT backscatter data have been widely used for melt detection in the Arctic (e.g., Kimball et al. 2004; Mortin et al. 2012; Wang et al. 2008). An algorithm capable of identifying multiple melt events was developed in Wang et al. (2008) for the terrestrial Arctic. The main melt-onset date (MOD) estimated using this algorithm has proved to be robust in evaluating snowmelt onset date in a global climate model (Wang et al. 2011). In this study, MOD was produced using the same algorithm as in Wang et al. (2008), except that to account for more frequent melt events in southern Québec, melt onset was determined if the daily QSCAT backscatter was 1.7 dB lower than the previous 3-day average (instead of 5-day average) for three or more consecutive days. The determination of the detection threshold (1.7 dB), melt-end date, and the main MOD followed Wang et al. (2008). The method was evaluated over Québec by examining the frequency of above-freezing mean daily surface air temperature from weather stations (see locations in Fig. 1) over the 3-day period up to and including the melt onset day following Wang et al. (2008). The number of positive air temperature occurrences was 4% two days before the estimated MOD, 45% one day before, and 76% on the estimated MOD, indicating that the melt detection algorithm performed as well in Québec as in the pan-Arctic where it was developed.

For the CRCM4-satellite evaluation of snowmelt onset date, CRCM4 grid cell MOD values were estimated from daily values of simulated snowmelt runoff. Melt onset was assumed when snowmelt runoff was greater than zero for seven or more consecutive days. We compared MOD estimates using 3-, 5-, and 7-day criteria. The results showed that 7-day was the appropriate value to capture the main melt onset from the CRCM4 (by examining the simulated daily snow water equivalent and surface air temperature). The mean MOD was computed over the 2000–09 period to match the satellite record. CLASS has a single-layer snowpack, which means the entire snowpack must become isothermal before meltwater is released. This is expected to introduce a delay in the melt response as compared to melt onset derived from satellite microwave sensors, which is mainly associated with the occurrence of surface melting of the snowpack. However, the timing of melt onset from QSCAT data corresponds to the early stage of the main melt event, when the snowpack is wet but still fully covering the ground (Wang et al. 2008), and was found to be compatible with snowmelt runoff timing in CLASS (Wang et al. 2011).

The total cloud fraction was obtained from the level-3 MODIS/Terra atmosphere monthly global 1° × 1° gridded product by the MODIS Science Team (MOD08_M3; Platnick et al. 2003). There are separate day and night cloud fraction data in the MODIS product. Daytime cloud detection accuracy is generally higher because of additional information in the solar spectral bands and was used in this study. Because both clouds and snow cover can be highly reflective in the visible and near-infrared bands and have low brightness temperature in the thermal bands, it is sometimes difficult to discriminate clouds from snow cover using visible and near-infrared satellite measurements. For comparison, daytime (0900–1700 local time) monthly mean cloud fraction was also computed from hourly observations at surface stations across Québec (see locations in Fig. 1).

Monthly radiation fluxes at the top of the atmosphere (TOA) and at the surface (Sfc) were obtained from the Clouds and the Earth’s Radiant Energy System (CERES) Synoptic 1° × 1° latitude/longitude gridded (SYN1deg) products (Doelling et al. 2013). The CERES instrument measures broadband shortwave (SW), total, and window radiances, and longwave (LW) radiances are determined from the difference between the total and the SW radiances. In the SYN1deg product, cloud and radiation changes between the satellite overpass times are inferred from 3-hourly imager data from five geostationary (GEO) satellites. To maintain calibration traceability, GEO radiances are calibrated against MODIS and the derived GEO fluxes are normalized to the CERES measurements (Doelling et al. 2013). To determine the distribution of surface radiation, the CERES team relies on radiative transfer model calculations initialized using satellite-based cloud and aerosol retrievals and meteorological and aerosol assimilation data from reanalysis to characterize the atmospheric state (Kato et al. 2013).

3. Results

a. Surface air temperature

Compared to observed air temperature from the CANGRD dataset, the simulated Tmin values in the CRCM4 were generally colder throughout the year, except for AIL in summer (June–September) and AIO in June (Fig. 3a). The largest cold temperature bias in Tmin was in fall (October–November). In contrast, simulated Tmax was warmer than observed from March to September (except for AIW in June and July), but too cold for the rest of the year (Fig. 3b). The Québec region annual mean biases for the three simulations were −1.4° (AIL), −1.9° (AIO), and −2.8°C (AIW) for Tmin and 0.24° (AIL), −0.22° (AIO), and −0.89°C (AIW) for Tmax.

Fig. 3.
Fig. 3.

Québec regionally averaged bias in screen level monthly (a) minimum, (b) maximum, and (c) mean air temperature (°C) in CRCM4 simulations: AIL, AIO, and AIW. Observed temperatures were obtained from the CANGRD dataset.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

b. Snow cover

On average, the Québec region is fully snow covered (>98%) from December to March each year (Fig. 4a). SCF from MODIS and IMS is nearly identical in the fall, but IMS gives a greater SCF estimate than MODIS in the spring. We believe the actual SCF lies somewhere between the MODIS and the IMS estimates based on previous discussion (see section 2c). We use SCF from MODIS in the following analyses, but keep in mind that it represents the lower limit in the spring period.

Fig. 4.
Fig. 4.

Québec regionally averaged (red outlined area in Fig. 1) (a) monthly mean snow cover fraction from MODIS, IMS, and CRCM4; (b) monthly mean surface albedo from MODIS and CRCM4 under clear sky; (c) melt- onset date (day of year) from QSCAT and CRCM4; and (d) seasonal cycle of cloud fraction from observations and CRCM4.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

Compared to observed SCF in fall, AIL shows relatively good agreement while AIO and AIW both significantly overestimate SCF from September to November. These results are consistent with the September temperature biases (Tmin and Tmean) shown in Fig. 3. Snowfall evaluation in Langlois et al. (2013, manuscript submitted to J. Hydrometeor.) also showed that, compared to AIL and observations, AIO consistently overestimates snowfall over the entire model domain. The same conclusion should apply to AIW since it uses the same mixed precipitation phase option as in AIO. Maps of SCF show that snow starts much earlier in the CRCM4, especially in AIO and AIW, showing excessive snow-covered areas south of the snow line observed by MODIS (Fig. 5). During the spring melt period (May–June), AIO SCF follows the MODIS estimate well, while AIL follows closely the IMS estimate. AIW overestimates SCF compared to both observations in spring (Fig. 4a). This appears to be closely associated with biases in surface albedo in these simulations, as discussed below and shown in Figs. 4b and 6, and with the aforementioned temperature bias (Fig. 3).

Fig. 5.
Fig. 5.

Maps of snow cover fraction in (top) September and (bottom) October from (a) MODIS, (b) AIL, (c) AIO, and (d) AIW.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

Fig. 6.
Fig. 6.

Differences in monthly surface albedo between CRCM4 and MODIS (CRCM4 − MODIS): (a) AIL, (b) AIO, and (c) AIW. (d) Monthly surface albedo from MODIS. (left to right) March–June.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

c. Surface albedo

Figure 4b shows the Québec regionally averaged seasonal variations of surface albedo from MODIS and CRCM4. All three simulations overestimate surface albedo in August by about 5%, which is attributed to a positive bias in the CLASS albedo [Bartlett et al. (2006) had to reduce the snow-free canopy albedo in CLASS for simulations in black spruce and jack pine forests], coupled with a documented negative bias in the MODIS albedo (Salomon et al. 2006; Stroeve et al. 2005). In the period of February–March, all three simulations underestimate surface albedo, with AIO having the largest negative bias (−10%; Table 2), AIW the least (−2%), and AIL in between (−5%). In the spring, both AIL and AIW overestimate surface albedo in May and June, which is linked to overestimation of SCF in these two runs (Fig. 4a). AIO underestimates surface albedo in April and May, which is consistent with the much earlier snowmelt onset in this run (Fig. 4c). The large difference in surface albedo from AIO and AIW throughout the snow cover season suggests that snow albedo in CLASS is very sensitive to Salb since this is the only difference in the configuration of AIO and AIW (Table 1). All three simulations clearly underestimate the annual range in surface albedo, with AIO indicating the smallest dynamic range.

Table 2.

Québec regionally averaged surface albedo from MODIS and difference with CRCM4 (CRCM4 − MODIS) during the 2000–09 period. The numbers in parentheses show the difference relative to the surface albedo from CERES.

Table 2.

Figure 6 shows the spatial distribution of surface albedo from MODIS and the bias of simulated surface albedo in CRCM4 relative to MODIS during March–June. Surface albedo for vegetated areas (see Fig. 1 for the distribution of tree fraction) with snow is generally greater than 0.2, while it reaches its maximum value (0.7) in the Arctic tundra during March–April (Fig. 6d). Because of snow metamorphism and melting, the high albedo in the tundra decreases to 0.6 or less in May. Melt advances very rapidly across Québec, and by June, snow only covers a small area in the northernmost tundra. All three simulations show large positive biases (>0.1) along the snow margin, with AIW having the largest magnitude and AIO the least.

During the peak snow accumulation period (March–April), the spatial distribution pattern of the bias is quite different in the three simulations (Figs. 6a–c). AIL shows negative bias in most of northern Québec and positive bias in southern Québec where there is dense forest. The positive bias in southern Québec is due to the fact that in CLASS 2.7, when calculating the whole surface albedo, a constant effective albedo value is used for the albedo of gaps in the forest over snow without explicitly accounting for the effects of snow metamorphism, melt, or shading (Bartlett et al. 2006). AIO shows a large negative bias in most of the model domain, except for the southeast corner, where the surface is not always snow covered in winter. AIW represents the surface albedo relatively well for the tundra and sparse forest areas (within ±5%), while it exhibits negative bias in some areas in southern Québec, which is related to underestimation in the canopy albedo for forests with intercepted snow. All three simulations show large negative biases in areas adjacent to the southern coast of Hudson Bay, which may be related to inaccuracies in either the vegetation mapping or the representation of certain vegetation types in CLASS and needs to be further studied.

According to observations from MODIS, snow is melted completely in southern Québec in May, and it only covers a small area in the northern Québec tundra by June (Fig. 6d). In May and June, AIL generally exhibits a positive albedo bias in areas where snow is melted, consistent with previous findings of delayed snow ablation in CLASS (Pomeroy et al. 1998; Slater et al. 2001; Brown et al. 2006). While in those similar areas, AIO shows relatively good agreement with observations because of early melt onset (Fig. 7c). AIW exhibits a large positive albedo bias in most of Québec, which is over 0.4 in northern Québec in June. This is consistent with delayed melt onset and significant overestimation in SCF in this simulation (Figs. 4a, 7d). In AIW, there is still snow cover in the Québec tundra in July and snow cover has not melted even by August on Baffin Island (not shown).

Fig. 7.
Fig. 7.

(a) Mean observed melt-onset date (day of year) from QSCAT. (b)–(d) CRCM4 bias (CRCM4 − QSCAT, days) for simulations of AIL, AIO, and AIW, respectively.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

d. Snowmelt onset

On average, melt starts from late March to mid-April in southern Québec, with the date of melt onset occurring later with increasing latitude (Fig. 7a). For the Arctic tundra, melt does not typically start until late May to early June. The spatial distribution in melt onset is closely correlated with tree density (Wang et al. 2011).

Compared to observations from QSCAT, MOD is relatively well represented in AIL, while it is too early in AIO and too late in AIW (Fig. 7, Table 3). This is consistent with observed biases in surface albedo in these simulations in spring (Fig. 6). During the February–April period, both AIL and AIO underestimate Québec regionally averaged surface albedo (Fig. 4b), and thus, the surface absorbs more solar radiation than observation would suggest, which compensates for the cold temperature bias in these two simulations (Fig. 3). As a result, the mean temperature bias becomes positive in March for AIL and in April for AIO (Fig. 3c). Surface albedo in AIW shows good agreement with observation during the period February–April and so the cold bias is not compensated, which leads to delayed melt onset (Figs. 3c, 7d). This is consistent with results from a sensitivity analysis using a coupled snow–ice model in Chung et al. (2010), indicating that the timing of spring snowmelt was highly sensitive to uncertainties in surface albedo.

Table 3.

Québec regionally averaged melt-onset date (day of year) from QSCAT and CRCM4 and the differences (days) between CRCM4 and QSCAT.

Table 3.

The mean MOD bias in Québec during 2000–09 is 1.8 days for AIL, −8.8 days for AIO, and 5.9 days for AIW (Table 3). The spatial distribution of the mean MOD bias is shown in Fig. 7. The MOD bias is within 1 week for 82% of the Québec region in AIL. It is mostly negative (early melt) for the whole model domain in AIO, with biases earlier than 1 week for 74% of the Québec region. In AIW, although the positive bias (late melt) in some tundra area is over 3 weeks, there are some small negative biases in southern Québec (30% of the Québec region), which is related to the negative bias in surface albedo in April when snowmelt starts there (figure not shown). These small negative biases in southern Québec offset some of the large positive biases in northern Québec (Fig. 7d), explaining the modest regional mean bias in this simulation (Table 3).

4. Radiation fluxes

On average, cloud cover fraction is over 60% year round in Québec (Fig. 4d). The cloud fraction from MODIS is, in general, consistent with estimates from weather stations, except for a slight overestimation in January–February (Fig. 4d), which is probably because of difficulties in differentiating clouds from snow using MODIS data under poor light conditions. Compared to surface observations, all three simulations slightly underestimate cloud fraction throughout most of the year, with relatively large underestimations during the spring period (April–June).

We compared clear-sky and all-sky radiation fluxes from the three simulations with values from CERES. All three simulations show a relatively large overestimation in clear-sky TOA reflected SW radiation (SW↑) in summer and early fall (June–October) and underestimation during February to April (Fig. 8a). In May, there is a positive bias in AIL and AIW, but a negative bias in AIO. This is consistent with observed biases in surface albedo (Fig. 4b). For example, AIW has the largest positive bias in surface albedo in June (~21%), which resulted in the largest positive bias in TOA SW↑, nearly 40 W m−2. Under all-sky conditions, there are small positive and negative biases in SW↑ during the winter and spring (January–May; Fig. 8b), suggesting that clouds are partly compensating the negative biases under clear-sky conditions.

Fig. 8.
Fig. 8.

Québec regionally averaged monthly differences in simulated and observed (top) TOA reflected and (bottom) surface incoming SW fluxes (CRCM4 − OBS) under (a),(c) clear-sky and (b),(d) all-sky conditions for: AIL, AIO, and AIW.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

At the surface, all three simulations show a large overestimation of clear-sky incoming SW radiation (SW) in the summer (Fig. 8c), which suggests an underestimation of SW radiation absorbed by the atmosphere in the model. All three simulations show some underestimation of SW↓ in the winter months (Fig. 8c), suggesting that the atmosphere is absorbing too much SW radiation. The corresponding all-sky plot in Fig. 8d shows that modeled clouds have changed the clear-sky summer positive bias in SW↓ to a negative bias (except for AIW in June) and have increased the negative winter bias. This suggests that clouds in the CRCM4 are too reflective, given that total cloud fraction is generally underestimated, especially, for example, during spring (Fig. 4d). The positive bias in June all-sky SW↓ in AIW is likely related to the large cold surface temperature bias (Fig. 3) and much-delayed snowmelt onset (Fig. 7), which contributes to a colder and dryer atmosphere in this simulation and to a large negative bias in June clear-sky LW↓ (Fig. 10c).

Under clear-sky conditions, the sign in the bias of surface-absorbed SW radiation is generally opposite to the bias in surface albedo in all three simulations each month (Fig. 9a, Table 2). An exception is the positive bias in AIL and AIO in September. Under all-sky conditions, all three simulations show negative biases in surface-absorbed SW radiation throughout the year, except for April, when there is positive bias in AIL and AIO (Fig. 9b).

Fig. 9.
Fig. 9.

As in Fig. 8, but for surface-absorbed SW fluxes.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

Under clear-sky conditions, all three simulations have negative biases in TOA outgoing LW radiation (LW↑) in fall and winter (October–March) and positive biases in April–September (Fig. 10a), with an exception of small negative biases in April and June in AIW. The negative biases in fall and winter are at least partly related to the large cold surface temperature bias shown in Fig. 3. All three simulations show negative biases in TOA LW↑ under all-sky conditions (Fig. 10b), and the most negative biases appear in the fall and winter (October–March). Given that fall and winter cloud fractions are simulated reasonably well, this suggests that cloud emissivity may be too small in the CRCM4. This is also consistent with the negative biases in downward LW radiation (LW↓) at the surface (Fig. 10d) and findings in MacKay et al. (2006) in an early version of the CRCM.

Fig. 10.
Fig. 10.

As in Fig. 8, but for TOA outgoing and surface incoming LW fluxes.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

At the surface, all three simulations have relatively large negative biases in clear-sky LW↓ during October–March and various biases for the rest of the year (Fig. 10c). This is consistent with results from Markovic et al. (2008), who suggested that the negative bias in surface LW↓ during the cold season is due to a combination of omission of trace gas contributions and a poor treatment of the water vapor continuum at low water vapor concentrations. Biases in all-sky LW↓ at the surface are negative year round (Fig. 10d).

In summary, insufficient cloud radiative forcing associated with low cloud fraction, high cloud reflectivity, and low cloud emissivity, in combination with inaccurate simulation of atmospheric properties, are likely contributing to the cold temperature bias in the model.

5. Discussion and conclusions

Simulated snow cover and surface albedo from a current operational CRCM4 simulation with CLASS 2.7 and two experimental CRCM4 simulations with CLASS 3.5 have been evaluated against satellite datasets over Québec and Labrador for a 10-yr period (2000–09). In the operational simulation (AIL), precipitation was deemed frozen or liquid based on a simple 0°C threshold temperature at the lowest atmospheric model layer, while in the experiments (AIO and AIW) mixed-phase precipitation occurs for temperatures between 0° and 6°C. All three simulations have a time-decaying snow albedo, with new snowfall refreshing snow albedo to its maximum value of 0.84. The three simulations used different 15-min model time step refresh thresholds (Salb) of 1.3, 5, and 0.1 mm respectively for AIL, AIO, and AIW. Taken in concert, these differences in precipitation phase and snow albedo refresh would appear to exert a strong influence on the simulation of surface albedo and snow cover season.

Both AIO and AIW show significant overestimation in SCF in the fall associated with early snow onset in the two simulations, while AIL agrees with observations relatively well. Early snow onset and thus overestimation in SCF greatly reduce absorbed solar radiation at the surface, resulting in a large cold temperature bias in CRCM4 in fall (Fig. 3). On the other hand, since CLASS uses air temperature to partition precipitation into rain or snow, a cold temperature bias would result in early snow onset in the CRCM4. To investigate the cause and effect, we compare simulated air temperature with observed on the day right before the simulated snow onset date (SOD) in the models in each fall during 2000–09. SOD at grid cells containing the weather stations were assumed if simulated daily snow water equivalent was greater than 0 mm for three or more consecutive days in the CRCM4. The results show that for the day right before the simulated SOD in CRCM4, simulated Tmin was colder than observed for 60%, 71%, and 75% of the days (~130 days in total) for AIL, AIO, and AIW, respectively (Fig. 11). In AIL, there were only 13 days with simulated Tmin less than 0°C and the cold bias occurred on 12 days. In AIO and AIW, the cold bias occurred on 84% and 89% of the days, respectively, when simulated Tmin was less than 6°C (99 days in total). The cold bias is more likely to cause early snow onset in AIO and AIW because of the additional snowfall from the mixed precipitation phase partitioning (between 0° and 6°C) in these two simulations. We also compared simulated Tmax with observed; however, in AIL none of the days had a simulated Tmax less than 0°C, and there were only a few days in AIO and AIW with simulated Tmax less than 6°C. Thus, early snow onset and overestimation in SCF in fall was mainly caused by a nighttime cold bias in the CRCM4. Once snow starts erroneously in the simulations, the cold bias is being reinforced through the snow albedo positive feedback mechanism, resulting in the largest cold bias in the fall months (Fig. 3). It also partially decouples the atmosphere from the land surface, allowing the near-surface atmosphere to cool down (especially in clear-sky/low-cloud-cover conditions).

Fig. 11.
Fig. 11.

Simulated and observed Tmin for the day before snow onset in (a) AIL and (b) AIW. The solid black line represents the 1:1 line, and the horizontal dashed black line in (b) represents 6°C, below which snowfall is possibly being partitioned in AIW. Results for AIO (not shown) are similar to AIW.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

During the peak snow accumulation period (February–April), surface albedo is underestimated in both AIL and AIO, and the extra SW radiation absorbed by the surface compensates for the cold temperature bias in these two simulations, as evidenced by the change to positive bias in Tmean in the spring (March–June) in both AIL and AIO (Fig. 3c). While melt onset in AIL agrees well with observations, it is too early in AIO, which is due to significant underestimation in surface albedo resulting from the unrealistically high Salb in this run. Simulated surface albedo in AIW shows good agreement with the observed values during February–April; thus, the cold temperature bias in this run is not compensated, and the Tmean bias in AIW is negative throughout the year (Fig. 3c), which causes delayed melt onset in AIW. In the spring melt period, the snowpack usually disappears in a few weeks (Wang et al. 2008). Delayed melt onset in AIW results in a significant overestimation in SCF and thus surface albedo, and the snow albedo positive feedback mechanism reinforces the cold temperature bias in the melt period (May and June; Fig. 3c). As a result, AIW still has snow cover in the Québec tundra in July, and snow cover is not melted even in August on Baffin Island.

Mean surface albedo varies widely among the simulations, with AIW yielding the largest values and AIO yielding the smallest values, consistent with the snow albedo refresh thresholds used. The annual range of surface albedo in all simulations is less than the observed estimates, with AIO exhibiting the least. The mean bias in melt-onset date (MOD) was found to be 1.8, 5.9, and −8.8 days for AIL, AIW, and AIO, respectively, with AIO also showing the least interannual variability.

To investigate the impact of the different precipitation phase expressions for partitioning snow and rain in CLASS, two new experiments with similar model configurations as in AIW but different precipitation phase options (0°C and 0°–2°C polynomial function) in CLASS 3.5 have been carried out. A comparison of the results from the two new simulations and AIW reveals that there is not much difference in surface albedo among the three simulations during both fall and spring (not shown). This suggests that the precipitation phase expression in CLASS 3.5 has little impact on snow albedo. Although CLASS 3.5 refreshes the snowpack albedo during mixed precipitation, since the air temperature is above 0°C and rain at the temperature of the air is also falling on the snowpack, most of the additional snow associated with allowing mixed-phase precipitation melts. Compared with observations, there are very limited improvements in the cold temperature bias from the new simulations (Fig. 12).

Fig. 12.
Fig. 12.

Québec regionally averaged bias in screen level mean air temperature in CRCM4 simulations for AJO, AJP, and AIW. Simulations AJO and AJP have similar model configurations as in AIW but different precipitation phase options: 0°C and 0°–2°C polynomial functions, respectively. Observed temperatures were obtained from the CANGRD dataset.

Citation: Journal of Hydrometeorology 15, 2; 10.1175/JHM-D-13-086.1

Cloud fraction is underestimated in all three simulations during spring, but only slightly underestimated throughout the remainder of the year. Under clear skies, all three simulations show large positive biases in surface SW↓ in summer and small negative biases in surface SW↓ in winter, suggesting inaccurate atmospheric absorption in the CRCM4 over Québec. Under all skies, the presence of cloud appears to offset these positive summer biases as well as increase small negative winter biases that were found under clear skies. Given that cloud fraction is at best biased slightly negative (significantly so in spring), this suggests that clouds are generally too reflective (i.e., cloud albedo is too high) in CRCM4 over this region.

Large negative biases in all-sky TOA outgoing LW↑ appear in all three simulations all year, but especially in fall and winter, when the cloud fraction appears to be reasonably well simulated with respect to surface observations. This, combined with the fact that surface incoming LW↓ is also too small, suggests that the simulated cloud emissivity may be too small. Under clear skies during fall and winter, all three simulations show large negative biases in TOA outgoing LW↑, which are partially related to the cold surface temperature bias. Large negative biases also appear in clear-sky surface incoming LW↓ in fall and winter, which suggests that atmospheric emissivity may be too small, consistent with the findings of Markovic et al. (2008).

Evaluation of fully coupled simulations of CRCM4 show that snow albedo in CLASS is very sensitive to the new snowfall depth threshold for refreshing the snowpack albedo (Salb). The results show that Salb = 5 mm is much too high and does not yield a realistic snow albedo. Surface albedo from the simulation with Salb = 0.1 mm (as in AIW) agrees well with observation, consistent with results from offline simulation experiments. The precipitation phase expression in CLASS has little impact on snow albedo simulation. A cold temperature bias in CRCM4 is responsible for the early snow start in fall and delayed snowmelt in spring. Insufficient cloud radiative forcing and biases associated with atmospheric properties in the CRCM4 appear to be responsible for the cold temperature bias in the model.

This study suggests that new parameterizations in LSMs designed for climate models should be tested in fully coupled simulations before being used to produce/project climate change simulations. Results in this study show that atmospheric forcing can exert a significant impact on the simulation of snow cover. Given the strong interactions/feedbacks between the surface and the atmosphere, a realistic representation of snow cover and surface albedo may not be possible without accurate simulation of cloud and atmospheric properties in climate models. While evaluations of standalone simulations of LSMs are very common in intercomparison studies, they may not reflect model performance in the fully coupled mode since standalone simulations usually use unbiased observed forcing data, which is rarely the case with modeled forcing fields.

A number of high-quality products from instruments onboard NASA’s Earth Observing System (e.g., MODIS and CERES) have been available for over a decade. New satellites launched in recent years provide unprecedented access to important parameters related to climate change, such as cloud thermodynamic phase from the Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO; Cesana and Chepfer 2013). This study demonstrates that satellite-derived datasets are very useful in climate model evaluation.

Acknowledgments

The CRCM4 output used in this paper was generated and supplied by Ouranos. The authors express their thanks to Blaise Gauvin St-Denis from the Ouranos Scenario Group for providing output from the CRCM4 runs in netcdf format and also to Hélène Côté of the Ouranos Climate Simulation Team for providing model diagnostics results. This paper represents an Environment Canada contribution to the Ouranos collaborative project “Evaluation and improvement of the representation of snow cover in the Canadian Regional Climate models CRCM4 and CRCM5 over northern Québec” funded by the Fonds de Recherche en Sciences du Climat (FRSCO) and the Consortium Ouranos (A. Royer PI). The authors thank colleagues Diana Verseghy and Jason Cole for helpful discussions and comments to the early version of the manuscript and Ewa Milewska for providing cloud observation data at surface stations. Financial support for Libo Wang is from the Canadian Space Agency GRIP project “Linking Satellite-Derived Snow Cover Information with the Canadian Land Data Assimilation System.” The authors thank three anonymous reviewers for their helpful comments.

REFERENCES

  • Bartlett, P. A., , MacKay M. D. , , and Verseghy D. L. , 2006: Modified snow algorithms in the Canadian land surface scheme: Model runs and sensitivity analysis at three boreal forest stands. Atmos.–Ocean, 44, 207222, doi:10.3137/ao.440301.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , Viterbo P. , , Beljaars A. , , Pan H. L. , , Hong S. Y. , , Goulden M. , , and Wofsy S. , 1998: Evaluation of land-surface interaction in ECMWF and NCEP/NCAR reanalysis models over grassland (FIFE) and boreal forest (BOREAS). J. Geophys. Res., 103 (D18), 23 07923 085, doi:10.1029/98JD02023.

    • Search Google Scholar
    • Export Citation
  • Brown, R. D., 2010: Analysis of snow cover variability and change in Quebec, 1948–2005. Hydrol. Processes,24, 1929–1954, doi:10.1002/hyp.7565.

  • Brown, R. D., , Brasnett B. , , and Robinson D. , 2003: Gridded North American monthly snow depth and snow water equivalent for GCM evaluation. Atmos.–Ocean, 41, 114, doi:10.3137/ao.410101.

    • Search Google Scholar
    • Export Citation
  • Brown, R. D., , Bartlett P. A. , , MacKay M. , , and Verseghy D. L. , 2006: Evaluation of snow cover in CLASS for SnowMIP. Atmos.–Ocean, 44, 223238, doi:10.3137/ao.440302.

    • Search Google Scholar
    • Export Citation
  • Brown, R. D., , Derksen C. , , and Wang L. , 2010: A multi-data set analysis of variability and change in Arctic spring snow cover extent, 1967–2008. J. Geophys. Res., 115, D16111, doi:10.1029/2010JD013975.

    • Search Google Scholar
    • Export Citation
  • Caya, D., , and Laprise R. , 1999: A semi-implicit, semi-Lagrangian regional climate model: The Canadian RCM. Mon. Wea. Rev., 127, 341362, doi:10.1175/1520-0493(1999)127<0341:ASISLR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Cesana, G., , and Chepfer H. , 2013: Evaluation of the cloud thermodynamic phase in a climate model using CALIPSO-GOCCP. J. Geophys. Res. Atmos., 118, 79227937, doi:10.1002/jgrd.50376.

    • Search Google Scholar
    • Export Citation
  • Chung, Y., , Belair S. , , and Mailhot J. , 2010: Simulation of snow on Arctic sea ice using a coupled snow–ice model. J. Hydrometeor., 11, 199–210, doi:10.1175/2009JHM1112.1.

    • Search Google Scholar
    • Export Citation
  • Dee, D., and Coauthors, 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
  • Déry, S. J., , and Brown R. D. , 2007: Recent Northern Hemisphere snow cover extent trends and implications for the snow-albedo feedback. Geophys. Res. Lett., 34, L22504, doi:10.1029/2007GL031474.

    • Search Google Scholar
    • Export Citation
  • Doelling, D., and Coauthors, 2013: Geostationary enhanced temporal interpolation for CERES flux products. J. Atmos. Oceanic Technol.,30, 1072–1090, doi:10.1175/JTECH-D-12-00136.1.

  • Dorsaz, F., , and Brown R. , 2008: Évaluation des simulations du couvert nival sur le Québec par les modèles MRCC 4.2.3 et GEMCLIM 3.3.0. Ouranos Rapport de stage Décembre 2008, 49 pp.

  • Dutra, E., , Balsamo G. , , Viterbo P. , , Miranda P. M. A. , , Beljaars A. , , Schär C. , , and Elder K. , 2010: An improved snow scheme for the ECMWF land surface model: Description and offline validation. J. Hydrometeor., 11, 899–916, doi:10.1175/2010JHM1249.1.

    • Search Google Scholar
    • Export Citation
  • Essery, R., , Morin S. , , Lejeune Y. , , and Ménard C. B. , 2013: A comparison of 1701 snow models using observations from an alpine site. Adv. Water Resour., 55, 131–148, doi:10.1016/j.advwatres.2012.07.013.

    • Search Google Scholar
    • Export Citation
  • Flanner, M. G., , Shell K. M. , , Barlage M. , , Perovich D. K. , , and Tschudi M. A. , 2011: Radiative forcing and albedo feedback from the Northern Hemisphere cryosphere between 1979 and 2008. Nat. Geosci., 4, 151155, doi:10.1038/ngeo1062.

    • Search Google Scholar
    • Export Citation
  • Gao, F., , Schaaf C. B. , , Strahler A. H. , , Roesch A. , , Lucht W. , , and Dickinson R. , 2005: MODIS bidirectional reflectance distribution function and albedo Climate Modeling Grid products and the variability of albedo for major global vegetation types. J. Geophys. Res., 110, D01104, doi:10.1029/2004JD005190.

    • Search Google Scholar
    • Export Citation
  • Gardner, A. S., , and Sharp M. J. , 2010: A review of snow and ice albedo and the development of a new physically based broadband albedo parameterization. J. Geophys. Res., 115, F01009, doi:10.1029/2009JF001444.

    • Search Google Scholar
    • Export Citation
  • Goulden, M. L., and Coauthors, 1998: Sensitivity of boreal forest carbon balance to soil thaw. Science, 279, 214217, doi:10.1126/science.279.5348.214.

    • Search Google Scholar
    • Export Citation
  • Groisman, P. Ya., , Karl T. R. , , and Knight R. W. , 1994: Observed impact of snow cover on the heat balance and the rise of continental spring temperatures. Science, 263, 198200, doi:10.1126/science.263.5144.198.

    • Search Google Scholar
    • Export Citation
  • Hall, A., , and Qu X. , 2006: Using the current seasonal cycle to constrain snow albedo feedback in future climate change. Geophys. Res. Lett., 33, L03502, doi:10.1029/2005GL025127.

    • Search Google Scholar
    • Export Citation
  • Hall, D. K., , Riggs G. A. , , Salomonson V. V. , , DiGirolamo N. E. , , and Bayr K. J. , 2002: MODIS snow-cover products. Remote Sens. Environ., 83, 181194, doi:10.1016/S0034-4257(02)00095-0.

    • Search Google Scholar
    • Export Citation
  • Hall, D. K., , Riggs G. A. , , and Salomonson V. V. , 2006: MODIS/Terra snow cover monthly L3 global 0.05Deg CMG, version 5. National Snow and Ice Data Center, Boulder, CO, digital media. [Available online at http://nsidc.org/data/docs/daac/modis_v5/mod10cm_modis_terra_snow_monthly_global_0.05deg_cmg.gd.html.]

  • Hansen, M., , DeFries R. , , Townshend J. , , Carroll M. , , Dimiceli C. , , and Sohlberg R. , 2003: Global percent tree cover at a spatial resolution of 500 meters: First results of the MODIS vegetation continuous fields algorithm. Earth Interact., 7 (10), 115, doi:10.1175/1087-3562(2003)007<0001:GPTCAA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Helfrich, S. R., , McNamara D. , , Ramsay B. H. , , Baldwin T. , , and Kasheta T. , 2007: Enhancements to, and forthcoming developments in the Interactive Multisensor Snow and Ice Mapping System (IMS). Hydrol. Processes, 21, 15761586, doi:10.1002/hyp.6720.

    • Search Google Scholar
    • Export Citation
  • Kato, S., , Loeb N. G. , , Rose F. G. , , Doelling D. R. , , Rutan D. A. , , Caldwell T. E. , , Yu L. , , and Weller R. A. , 2013: Surface irradiances consistent with CERES-derived top-of-atmosphere shortwave and longwave irradiances. J. Climate, 26, 27192740, doi:10.1175/JCLI-D-12-00436.1.

    • Search Google Scholar
    • Export Citation
  • Key, J. R., , Wang X. J. , , Stroeve J. C. , , and Fowler C. , 2001: Estimating the cloudy-sky albedo of sea ice and snow from space. J. Geophys. Res., 106 (D12), 12 48912 497, doi:10.1029/2001JD900069.

    • Search Google Scholar
    • Export Citation
  • Kimball, J. S., , McDonald K. C. , , Running S. W. , , and Frolking S. E. , 2004: Satellite radar remote sensing of seasonal growing seasons for boreal and subalpine evergreen forests. Remote Sens. Environ., 90, 243258, doi:10.1016/j.rse.2004.01.002.

    • Search Google Scholar
    • Export Citation
  • Liston, G. E., 2004: Representing subgrid snow cover heterogeneities in regional and global models. J. Climate, 17, 13811397, doi:10.1175/1520-0442(2004)017<1381:RSSCHI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lucht, W., , Schaaf C. B. , , and Strahler A. H. , 2000: An algorithm for the retrieval of albedo from space using semiempirical BRDF models. IEEE Trans. Geosci. Remote Sens., 38, 977998, doi:10.1109/36.841980.

    • Search Google Scholar
    • Export Citation
  • MacKay, M. D., , Seglenieks F. , , Verseghy D. , , Soulis E. D. , , Snelgrove K. R. , , Walker A. , , and Szeto K. , 2003: Modeling Mackenzie basin surface water balance during CAGES with the Canadian Regional Climate Model. J. Hydrometeor., 4, 748767, doi:10.1175/1525-7541(2003)004<0748:MMBSWB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • MacKay, M. D., , Bartlett P. , , Chan E. , , Dersen C. , , Guo S. , , and Leighton H. , 2006: On the simulation of regional scale sublimation over boreal and agricultural landscapes in a climate model. Atmos.–Ocean, 44, 289304, doi:10.3137/ao.440306.

    • Search Google Scholar
    • Export Citation
  • Male, D. H., , and Granger R. J. , 1981: Snow surface energy exchange. Water Resour. Res., 17, 609627, doi:10.1029/WR017i003p00609.

  • Markovic, M., , Jones C. G. , , Vaillancourt P. A. , , Paquin D. , , Winger K. , , and Paquin-Ricard D. , 2008: An evaluation of the surface radiation budget over North America for a suite of regional climate models against surface station observations. Climate Dyn., 31, 779–794, doi:10.1007/s00382-008-0378-6.

    • Search Google Scholar
    • Export Citation
  • Mellor, M., 1977: Engineering properties of snow. J. Glaciol., 19, 1566.

  • Milewska, E. J., , Hopkinson R. F. , , and Niitsoo A. , 2005: Evaluation of geo-referenced grids of 1961–1990 Canadian temperature and precipitation normals. Atmos.–Ocean, 43, 4975, doi:10.3137/ao.430104.

    • Search Google Scholar
    • Export Citation
  • Moody, E. G., , King M. D. , , Schaaf C. B. , , Hall D. K. , , and Platnick S. , 2007: Northern Hemisphere five-year average (2000–2004) spectral albedos of surfaces in the presence of snow: Statistics computed from Terra MODIS land products. Remote Sens. Environ., 111, 337345, doi:10.1016/j.rse.2007.03.026.

    • Search Google Scholar
    • Export Citation
  • Mortin, J., , Schrøder T. M. , , Walløe Hansen A. , , Holt B. , , and McDonald K. C. , 2012: Mapping of seasonal freeze-thaw transitions across the pan-Arctic land and sea ice domains with satellite radar. J. Geophys. Res., 117, C08004, doi:10.1029/2012JC008001.

    • Search Google Scholar
    • Export Citation
  • Music, B., , and Caya D. , 2007: Evaluation of the hydrological cycle over the Mississippi River basin as simulated by the Canadian Regional Climate Model (CRCM). J. Hydrometeor., 8, 969988, doi:10.1175/JHM627.1.

    • Search Google Scholar
    • Export Citation
  • Oleson, K. W., , Bonan G. B. , , Schaaf C. B. , , Gao F. , , Jin Y. , , and Strahler A. H. , 2003: Assessment of global climate model land surface albedo using MODIS data. Geophys. Res. Lett., 30, 1443, doi:10.1029/2002GL016749.

    • Search Google Scholar
    • Export Citation
  • Pedersen, C. A., , and Winther J. G. , 2005: Intercomparison and validation of snow albedo parameterization schemes in climate models. Climate Dyn., 25, 351362, doi:10.1007/s00382-005-0037-0.

    • Search Google Scholar
    • Export Citation
  • Platnick, S., , King M. D. , , Ackerman S. A. , , Menzel W. P. , , Baum B. A. , , Riedi J. C. , , and Frey R. A. , 2003: The MODIS cloud products: Algorithms and examples from Terra. IEEE Trans. Geosci. Remote Sens., 41, 459473, doi:10.1109/TGRS.2002.808301.

    • Search Google Scholar
    • Export Citation
  • Pomeroy, J. W., , Gray D. M. , , Shook K. R. , , Toth B. , , Essery R. L. H. , , Pietroniro A. , , and Hedstrom N. , 1998: An evaluation of snow accumulation and ablation processes for land surface modeling. Hydrol. Processes, 12, 23392367, doi:10.1002/(SICI)1099-1085(199812)12:15<2339::AID-HYP800>3.0.CO;2-L.

    • Search Google Scholar
    • Export Citation
  • Qu, X., , and Hall A. , 2007: What controls the strength of snow-albedo feedback? J. Climate, 20, 39713981, doi:10.1175/JCLI4186.1.

  • Ramsay, B., 1998: The interactive multisensor snow and ice mapping system. Hydrol. Processes, 12, 15371546, doi:10.1002/(SICI)1099-1085(199808/09)12:10/11<1537::AID-HYP679>3.0.CO;2-A.

    • Search Google Scholar
    • Export Citation
  • Riette, S., , and Caya D. , 2002: Sensitivity of short simulations to the various parameters in the new CRCM spectral nudging. Research activities in atmospheric and oceanic modelling, H. Ritchie, Ed., WMO/TD 1105, Rep. 32, 7.39–7.40. [Available online at http://www.wcrp-climate.org/WGNE/BlueBook/2002/individual-articles/07_s_bastien_riette.pdf.]

  • Roe, G. H., , and Baker M. B. , 2007: Why is climate sensitivity so unpredictable? Science, 318, 629632, doi:10.1126/science.1144735.

  • Roesch, A., , Wild M. , , Gilgen H. , , and Ohmura A. , 2001: A new snow cover fraction parameterization for the ECHAM4 GCM. Climate Dyn., 17, 933946, doi:10.1007/s003820100153.

    • Search Google Scholar
    • Export Citation
  • Roman, M. O., and Coauthors, 2009: The MODIS (Collection V005) BRDF/albedo product: Assessment of spatial representativeness over forested landscapes. Remote Sens. Environ., 113, 24762498, doi:10.1016/j.rse.2009.07.009.

    • Search Google Scholar
    • Export Citation
  • Rutan, D., , Rose F. , , Roman M. , , Manalo-Smith N. , , Schaaf C. , , and Charlock T. , 2009: Development and assessment of broadband surface albedo from Clouds and the Earth’s Radiant Energy System Clouds and Radiation Swath data product. J. Geophys. Res., 114, D08125, doi:10.1029/2008JD010669.

    • Search Google Scholar
    • Export Citation
  • Salomon, J. G., , Schaaf C. B. , , Strahler A. H. , , Gao F. , , and Jin Y. F. , 2006: Validation of the MODIS bidirectional reflectance distribution function and albedo retrievals using combined observations from the Aqua and Terra platforms. IEEE Trans. Geosci. Remote Sens., 44, 15551565, doi:10.1109/TGRS.2006.871564.

    • Search Google Scholar
    • Export Citation
  • Salzen, K. V., and Coauthors, 2013: The Canadian Fourth Generation Atmospheric Global Climate Model (CanAM4). Part I: Representation of physical processes. Atmos.–Ocean, 51, 104125, doi:10.1080/07055900.2012.755610.

    • Search Google Scholar
    • Export Citation
  • Schaaf, C. B., and Coauthors, 2002: First operational BRDF, albedo nadir reflectance products from MODIS. Remote Sens. Environ., 83, 135148, doi:10.1016/S0034-4257(02)00091-3.

    • Search Google Scholar
    • Export Citation
  • Schaaf, C. B., , Wang Z. , , and Strahler A. H. , 2011: Commentary on Wang and Zender—MODIS snow albedo bias at high solar zenith angles relative to theory and to in situ observations in Greenland. Remote Sens. Environ., 115, 12961300, doi:10.1016/j.rse.2011.01.002.

    • Search Google Scholar
    • Export Citation
  • Slater, A. G., and Coauthors, 2001: The representation of snow in land surface schemes: Results from PILPS 2(d). J. Hydrometeor., 2, 725, doi:10.1175/1525-7541(2001)002<0007:TROSIL>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Stroeve, J., , Box J. E. , , Gao F. , , Liang S. L. , , Nolin A. , , and Schaaf C. , 2005: Accuracy assessment of the MODIS 16-day albedo product for snow: Comparisons with Greenland in situ measurements. Remote Sens. Environ., 94, 4660, doi:10.1016/j.rse.2004.09.001.

    • Search Google Scholar
    • Export Citation
  • Sturm, M., , Holmgren J. , , König M. , , and Morris K. , 1997: The thermal conductivity of seasonal snow. J. Glaciol., 43, 2641.

  • Verseghy, D., 1991: Class—A Canadian land surface scheme for GCMs. I. Soil model. Int. J. Climatol., 11, 111133, doi:10.1002/joc.3370110202.

    • Search Google Scholar
    • Export Citation
  • Verseghy, D., , McFarlane N. , , and Lazare M. , 1993: Class—A Canadian land surface scheme for GCMs. II: Vegetation model and coupled runs. Int. J. Climatol., 13, 347370, doi:10.1002/joc.3370130402.

    • Search Google Scholar
    • Export Citation
  • Viterbo, P., , and Betts A. K. , 1999: Impact on ECMWF forecasts of changes to the albedo of the boreal forests in the presence of snow. J. Geophys. Res., 104 (D22), 27 80327 810, doi:10.1029/1998JD200076.

    • Search Google Scholar
    • Export Citation
  • Wang, L., , Derksen C. , , and Brown R. , 2008: Detection of pan-Arctic terrestrial snowmelt from QuikSCAT, 2000–2005. Remote Sens. Environ., 112, 37943805, doi:10.1016/j.rse.2008.05.017.

    • Search Google Scholar
    • Export Citation
  • Wang, L., , Wolken G. J. , , Sharp M. J. , , Howell S. E. L. , , Derksen C. , , Brown R. D. , , Markus T. , , and Cole J. , 2011: Integrated pan-Arctic melt onset detection from satellite active and passive microwave measurements, 2000–2009. J. Geophys. Res., 116, D22103, doi:10.1029/2011JD016256.

    • Search Google Scholar
    • Export Citation
  • Wang, Z., , and Zeng X. B. , 2010: Evaluation of snow albedo in land models for weather and climate studies. J. Appl. Meteor. Climatol., 49, 363380, doi:10.1175/2009JAMC2134.1.

    • Search Google Scholar
    • Export Citation
  • Wang, Z., and Coauthors, 2012: Evaluation of Moderate-Resolution Imaging Spectroradiometer (MODIS) snow albedo product (MCD43A) over tundra. Remote Sens. Environ., 117, 264–280, doi:10.1016/j.rse.2011.10.002.

    • Search Google Scholar
    • Export Citation
  • Zhang, X., , Vincent L. A. , , Hogg W. D. , , and Niitsoo A. , 2000: Temperature and precipitation trends in Canada during the 20th century. Atmos.–Ocean, 38, 395429, doi:10.1080/07055900.2000.9649654.

    • Search Google Scholar
    • Export Citation
  • Zhou, L., and Coauthors, 2003: Comparison of seasonal and spatial variations of albedos from Moderate-Resolution Imaging Spectroradiometer (MODIS) and Common Land Model. J. Geophys. Res., 108, 4488, doi:10.1029/2002JD003326.

    • Search Google Scholar
    • Export Citation
Save