Modelling albedo and distributed snowmelt across a low hill in Svalbard Richard Essery 1 , Eleanor Blyth 2 , Richard Harding 2 and Colin Lloyd 2 1 Corresponding author. Institute of Geography and Earth Sciences, University of Wales, Aberystwyth SY23 3DB, UK. Tel: þ44 1970 622784; Fax: þ44 1970 622659; E-mail: rie@aber.ac.uk 2 Centre for Ecology and Hydrology, Wallingford, OX10 8BB, UK. Received 27 August 2004; accepted in revised form 24 January 2005 Abstract A land-surface model is used to simulate the albedo and mass of patchy snowcovers during radiation- driven melt for three years at a site in Svalbard. Performing single energy and mass balance calculations for the combined snow-covered and snow-free parts of the surface gives a faster decrease in albedo than observed because too much of the solar radiation absorbed by the composite surface is used to melt snow. Representing the snowcover separately allows the model to be calibrated to give a good match to the observed albedo for each of the years studied. A single set of model parameters cannot, however, give a good simulation for all of the years. The average snow mass and snowcover fraction measured on a grid of points can be simulated using either a distributed version of the model or a more efficient tiled version supplied with the observed relationship between snow mass and fractional coverage. Parameters obtained by optimising the snow mass simulations are more consistent from year to year than from the albedo simulations. Keywords Snow albedo; snowcover; snowmelt; snow modelling Introduction The presence of snow has a large effect on the energy balance of the land surface: a large fraction of the incoming solar radiation is reflected from a high albedo snow surface and the high latent heat of fusion for snow absorbs much of the available energy during melting. Through its strong influence on exchanges of energy and moisture between the surface and the atmosphere, snowcover has an important role in the global climate and has to be represented in climate models, but several studies have shown it to be a difficult quantity to model accurately and consistently (e.g. Essery et al. 1999b; Slater et al. 2001; Bowling et al. 2003). Snowmelt is driven by radiation and turbulent heat transfers from the atmosphere: solar radiation is the dominant source of energy for the melting of snow persisting into the late spring or summer at high latitudes and high elevations, although turbulent fluxes are also important in mid-latitude snowmelt. Fresh snow has a high albedo, but the albedo of a snow surface decreases over time with changes in the grain structure of the snow (Wiscombe and Warren 1980) and the deposition of contaminants (Warren and Wiscombe 1980). Snow generally becomes patchy and dirty while melting, giving a surface with highly heterogeneous characteristics. Inevitably, land-surface models have to use simplified representations of snow processes and, in the case of global modelling applications, global parameters often have to be applied. Aging of snow is typically represented by making the albedo a function of the surface temperature or a function of the age and temperature history of the snow surface. Heterogeneity on scales smaller than the model resolution has to be parametrised; this is usually done through the introduction of a function relating the fraction of snowcover to the average snow depth in a gridbox. Models differ in how they perform surface flux calculations for heterogeneous snowcover: the snowcover fraction may be used Nordic Hydrology Vol 36 No 3 pp 207–218 q IWA Publishing 2005 207 Downloaded from http://iwaponline.com/hr/article-pdf/36/3/207/364614/207.pdf by guest on 04 May 2021