Abstract:Much of the snowmelt season is characterized by a patchy surface; differential heating of the snow and snow-free surfaces results in a significant horizontal transport of energy that affects and contributes to the snowmelt. The calculation of the rate of energy advection requires some knowledge of the behaviour of the thermal boundary layer over the patches of snow and snow-free surfaces. We present the results from a series of field observations of the rate of growth of the thermal boundary layer over snow and snow-free patches. The results confirm that the boundary-layer growth can be described by a power function of the distance from the leading edge of the patch. For the case of the thermal boundary layer over a snow patch within a bare field, the boundary-layer growth is affected by the upwind surface roughness; the thermal boundary layer over a snow patch within a 'rough' field grows much more quickly than that in a 'smooth' field. KEY WORDS boundary layer; advection; snowmelt; sensible heat; snow patches INTRODUCTION In the calculation of the melting of a patchy snow cover, the energy advected from the adjacent bare soil to the snow surface is an important consideration. As the lower-layer air moves from the bare soil to the snow, it undergoes a considerable modification as energy is extracted. This additional flux of energy to the snow surface cannot be reliably calculated using traditional boundary-layer flux-profile relationships, since these are based on the assumption of a constant flux layer. To date, only a very few attempts have been made to provide estimates of this advective flux over a field. Shook et al. (1993) found that snow patch geometry and size frequency distributions could be described using fractal geometry. Synthetic snow-cover spatial distributions were created using the fractal sum of pulses method and in a gridded calculation; an approximation of Weisman's (1977) model of advection over snow was applied to estimate the areal melt (Shook, 1995; Shook and Gray, 1997). Liston (1995) applied a numerical atmospheric boundary-layer model, based on higher order turbulence closure assumptions, to simulate local advection during the melt of a hypothetical snow cover with simple geometrical properties; he used the model to demonstrate the increase in available melt energy as the snow cover is depleted. Essery (1999) also applied a modified boundary-layer model to assess the advection of energy with varying snow cover fraction.The boundary-layer analyses described by Weisman (1977) and Liston (1995) require the solution of a very complex set of equations describing the conservation of momentum, mass and energy, as well as the changes in atmospheric stability as a boundary layer develops. Weisman reduced his results to a simple parametric form; however, he was only able to provide a few sample values of the coefficients required to apply his