Snow Cover in General Circulation Models, Subproject No. 28:
David A. Robinson1
Ross Brown and Anne Walker3
CIRES, University of Colorado2
Atmospheric Environment Service (Canada)3
Methodology and Validation
Proper simulation of snow cover in climate models is important for numerous reasons such as snow-cover climate feedbacks (albedo-temperature, snow-cloud), and correct simulation of soil moisture, runoff and ground temperatures (Cohen and Rind, 1991; Marshall et al., 1994; Lynch-Stieglitz, 1994). Snow cover is also an excellent diagnostic for evaluation since both temperature and precipitation must be realistic in order to correctly model important snow cover properties such as duration, water-equivalent and density (Foster et al., 1996). Results of an AMIP I snow cover evaluation (Frei and Robinson, 1995; 1998) and recent GCM snow cover verification studies (Foster et al., 1996; Walland and Simmonds, 1996) demonstrated that to a first order, GCMs captured the seasonal cycle of NH snow cover extent. However, the results also revealed (1) major differences between model representations of snow cover, (2) seasonal biases in hemispheric snow covered area, and (3) consistently poor simulation of regional snow cover over Tibet and China linked to model deficiencies in simulating air temperature and precipitation. Among the problems that remain outstanding from AMIP-1, Gates et al. (1999) identify "the portrayal of the interactions between the land surface and hydrological processes." In this subproject, Rutgers University, the University of Colorado and the Canadian Atmospheric Environment Service are proposing to focus their combined expertise in snow cover remote sensing and data analysis to provide a systematic analysis of the snow cover performance of the models included in AMIP II.Objectives
The initial stage of this AMIP subproject will focus on verifying spatial and temporal aspects of snow cover simulations over the Northern Hemisphere relying mainly on NOAA weekly satellite snow cover observations (Robinson et al, 1993). A detailed regional evaluation of snow cover properties (snow water equivalent (SWE) and density) will be carried out over the Canadian prairies using snow course observations and SWE estimates from the Special Sensor Microwave/Imager (SSM/I) algorithm of Walker and Goodison (1993). Following this first step, an attempt will be made to diagnose the most significant model-to-model and model-observed differences by examining and verifying some of the main driving parameters (e.g. temperature and precipitation fields) and atmospheric circulation patterns (e.g. the PNA and NAO which were shown to be linked closely to snow cover variability over North America (NA) by Gutzler and Rosen (1992) and Brown and Goodison (1996)). Guidance on which driving parameters to investigate will be provided from sensitivity analysis of the CROCUS physical snowpack model (Brun et al., 1989 - updated in 1996). Physical processes will also need to be investigated. For instance, verification of simulated snow cover in BATS by Yang et al. (1997) suggested that special attention must be given to downward longwave radiation, the snow aging process (density and surface albedo) and the critical temperature for rain/snow separation.Methodology
5.1 Snow cover extent climatology comparisonData Requirements
This step will involve the computation of standard monthly statistics (mean, s.d., extremes) of continental (NA, Eurasia) and hemispheric snow covered area (SCA) over the 17-year period to determine how well the models simulate the main features of NH snow cover extent (seasonal cycle, extremes, interannual variability) as determined from the Rutgers-corrected version of the NOAA weekly snow cover dataset (Robinson et al., 1993). Monthly SCA difference (obs-modeled) plots will be used to identify areas with significant snow cover discrepancies. In these areas, an assessment will be made of the quality of satellite observations (where snow depth observations are available), and the simulated air temperature and precipitation fields. Inter-model differences will be assessed through examination of diagnostic variables such as ground temperature, surface albedo, surface radiation and energy fluxes, and runoff. The ECMWF and NCEP reanalysis data sets and available temperature and precipitation climatologies (e.g. Legates and Willmott, 1990; Huffman et al., 1997) will be used to examine model- observed and model-model differences.
5.2 Snow depth climatology comparison
Most models compute snow mass and infer snow water equivalent using many different assumptions about snow density. Some of the more sophisticated snow models compute snow density but this information will need to be determined for each model. If modeled snow density information is available, this will be verified against available snow course observations (extensive data are available over Canada and the FSU). One approach to this problem is to use available snow course observations to define a NH monthly snow density climatology, which would be applied to all model snow mass fields. The disadvantage of this approach is that it ignores interannual variability in snow density, but it does allow all models to be compared in a consistent fashion. Modeled mean snow depths will be compared to available NH snow depth climatologies (RAND and USAF). These climatologies may be biased high compared to the AMIP II period, as the NOAA SCA data (1972 to date) show that NH snow cover anomalies were consistently negative during the late 1980s and early 1990s. An attempt will be made to quantify this (at least over NA) using available snow depth observations. Systematic differences in snow depths will be assessed by looking at how well models simulate important processes such as rain/snow separation, snow aging (density and albedo) and winter melt events. It is proposed to use the CROCUS snowpack model to help provide insight into possible sources of systematic error in GCM snow cover simulations. Snow cover albedo will be assessed through a pentad 1 degree by 1 degree albedo data set being developed at Rutgers.
Access to the high frequency model snow cover output is necessary for process studies and for verifying snow cover in shoulder seasons and in regions with ephemeral snow cover.
5.3 Snow water-equivalent climatology comparison
There are no known reliable hemispheric climatologies of SWE. However, two project members (DR and AF) in cooperation with colleagues at the Universities of Delaware and North Dakota, have just received funding from NASA to develop a detailed 40-year climatology of snow water equivalent (SWE) for the grassland regions of North America and Eurasia. The SWE climatology will be produced by blending SWE estimates from a physically-based snowpack model (SNTHERM), remotely sensed estimates (SMMR and SSM/I) and station observations. The consideration of the strengths and weaknesses of these three diverse data sets will lead to a more spatially and temporally detailed SWE climatology than has been produced in the past for any large region of the Earth. These results will be compared with those of Goodison and Walker (1994), who have had considerable success estimating SWE from SSM/I over the prairie region
of Canada. Wet snow events are identified following Walker and Goodison (1993) and are useful in computing monthly averages.
5.4 Assessment of correct spatial/temporal representation of snow cover
PC analysis will be applied to observed and modeled snow cover data sets over the 17-year period to identify coherent regions of snow cover variability following Frei et al. (1996). Comparison of rotated loading patterns and factor score time series for significant PCs will allow an objective assessment to be made of each model's ability to capture the dominant spatial and temporal modes of snow cover variability over the study period.
5.5 Assessment of snow cover climatic and synoptic associations
The relationship between interannual fluctuations in snow cover and atmospheric circulation will be examined following Gutzler and Rosen (1992) and Frei (1997) for both observed and modeled snow cover. This will involve correlation of monthly atmospheric circulation indices (PNA, NAO, SOI) with output from the PC analysis in 5.4, and investigation of snow cover climatic and synoptic associations through composite (high-snow/low-snow) analyses of surface (temperature and precipitation) and upper-air (500 mb height) data.
Upper-air monthly mean output:References
Single-level monthly mean output:
Surface air temperature
Mean sea-level pressure
Total precipitation rate
Snow depth (water equivalent)
Snow density (if available)
Surface eastward wind
Surface westward wind
Surface specific humidity
Surface sensible heat flux
Surface latent heat flux
Surface evaporation and sublimation rate
Surface incident SW radiation
Surface reflected SW radiation
Surface downwelling LW radiation
Surface upwelling LW radiation
Daily maximum surface temperature
Daily minimum surface temperature
Total cloud amount
High frequency: (6-hourly)
Snow depth (water equivalent); needed to compute daily snow cover statistics for comparison with observations e.g. start/end of snow cover, SWE maxima
Brown, R.D. and B.E. Goodison. 1996. Interannual variability in reconstructed Canadian snow cover, 1915-1992. J. Climate, 9:1299-1318.
Brun, E., Martin, E. Simon, V., Gendre C., and Coléou C, 1989: An energy and mass model of snow cover from operational avalanche forecasting. J. Glaciology, 35:333-342.
Cohen, J., and D. Rind, 1991: The effect of snow cover on the climate. J. Climate, 4:689-706.
Foster, J., G. Liston, R. Koster, R. Essery, H. Behr, L. Dumenil, D. Verseghy, S. Thompson, D. Pollard and J. Cohen, 1996: Snow cover and snow mass intercomparisons of general circulation models and remotely sensed datasets. J. Climate, 9:409-426.
Frei, A., 1997: Towards a Snow Cover Fingerprint for Climate Change Detection. Dissertation, Rutgers University, 245pp.
Frei, A. and D.A. Robinson, 1995: Northern Hemisphere snow cover extent: comparison of AMIP results to observations. Proc. AMIP Scientific Conference, Monterey, California, 499-504.
Frei, A. and D.A. Robinson, 1998: Evaluation of snow extent and its variability in the Atmospheric Model Intercomparison Project. J. Geophys. Res. - Atmospheres, 103, 8859-8871.
Frei, A., D.A. Robinson and M.G. Hughes, 1996: A regional analysis of North America snow cover extent: climatic and synoptic associations from November through March. Proc. 53rd Eastern Snow Conference, Williamsburg, Virginia, 33-42.
Gates, W.L. and others, 1999: An overview of the results of the Atmospheric Model Intercomparison Project (AMIP I). Bull. Am. Met. Soc., 80, 29-55.
Goodison, B.E., and A.E. Walker, 1994: Canadian development and use of snow cover information from passive microwave satellite data. Proc. EAS/NASA International Workshop on Passive Microwave Remote Sensing Research Related to Land-Atmosphere Interactions, St. Lary, France, 245-262.
Gutzler, D.S., and R.D. Rosen, 1992: Interannual variability of wintertime snow cover across the Northern Hemisphere. J. Climate, 5:1441-1447.
Huffman, G.J. and 9 others, 1997: The global precipitation climatology project (GPCP) combined precipitation dataset. Bull. Amer. Met. Soc., 78:5-20.
Legates, D.R. and C.J. Willmott, 1990: Mean seasonal and spatial variability in gauge-corrected global precipitation. Int. J. Climatology, 10:111-127.
Lynch-Stieglitz, M., 1994: The development and validation of a simple snow model for the GISS GCM. J. Climate, 7:1842-1855.
Marshall, S., J.O. Roads, and G. Glatzmaier, 1994: Snow hydrology in a general circulation model. J. Climate, 7:1251-1269.
Robinson, D.A., 1996: Evaluating snow cover over northern hemisphere lands using satellite and in situ observations. Proc. 53rd Eastern Snow Conference, Williamsburg, VA, 13-19.
Robinson D.A., K.F. Dewey, and R.R. Heim, 1993: Global snow cover monitoring: an update. Bull. Amer. Met. Soc., 74, 1689-1696.
Walker, A.E., and B.E. Goodison, 1993: Discrimination of wet snow cover using passive microwave satellite data. Annals of Glaciology, 17:307- 311.
Walland, D.J. and I. Simmonds, 1996: Sub-grid-scale topography and the simulation of northern hemisphere snow cover. Intl. J. Climatology, 16:961-982.
Yang, Z., R.E. Dickinson, A. Robock and K. Ya. Vinnikov, 1997: Validation of the snow submodel of the Biosphere-Atmosphere Transfer scheme with Russian snow cover and meteorological observational data. J. Climate, 10:353-373.
For further information, contact David Robinson (email@example.com) or the AMIP Project Office (firstname.lastname@example.org).
Last update: 3 March 1999. This page is maintained by email@example.com