|DAVIS, K - UNIVERSITY OF MINNESOTA
|LIKNESS, G - UNIVERSITY OF MINNESOTA
Submitted to: Journal of Geophysical Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/22/1999
Publication Date: N/A
Interpretive Summary: Spring snowmelt poses difficult problems for weather prediction and hydrologic models. Better prediction is necessary to reduce the damage from floods such as the catastrophic 1997 Red River flood, to reduce the environmental damage of snowmelt runoff due to sediment and chemical transport, and to provide more accurate weather forecasts. The main impediment to model improvement is the lack of comprehensive data sets tha include measurements of both the variables that affect snowmelt and the resulting transport of energy and water. We conducted a field experiment at the Rosemount, MN Agricultural Experiment Station where we measured all components of the surface energy balance, including soil heat flux, surface evaporation rates, and radiant energy exchange. In addition we collected radiosonde measurements of temperature and humidity in the atmospheric boundary layer to provide a complete picture of the snowmelt process during gtwo melt periods. The first occurred under cloudy skies, while the second took place during clear weather. One melt was caused primarily by longwave radiation and sensible heat exchange with a relatively warm air mass, while direct solar heating was the primary energy source for the second melt. In both cases we were able to estimate snowmelt accurately by balancing all energy flow at the surface. As snow disappeared, the surface reflectance increased dramatically, surface temperatures and evaporation rates also showed rapid increases, and this combination led to increased interaction between the surface and the atmosphere in the form of greater turbulence and boundary layer development. The data will be used by scientists within NOAA and at a number of universities to test and improve snowmelt prediction in weather and hydrologic models.
Technical Abstract: Improved prediction of snowmelt requires comprehensive data collection, including surface, subsurface, and atmospheric processes, during the snowmelt period. We report results of field research in which all components of the surface energy balance were measured during two different snowmelt periods, along with boundary layer soundings. The two periods were equite different, the first being overcast and the second occurring under clear skies. However, snowmelt was estimated relatively well from the cumulative residual of the energy balance in both cases. Downwards infrared radiation and sensible heat flux were important contributors to the melt during overcast conditions, with net radiation providing about 2/3 of the energy for melt, and sensible heat providing the remainder. The sunny melt was dominated by direct solar heating of the surface. In both cases, estimation of melt as a residual of the energy balance agreed well with visual and gravimetric observations. The boundary layer soundings revealed the importance of advection, which was generally consistent with synoptic patterns during the period of the study. The data also showed a transition from advection-dominated to turbulence-dominated boundary layer budgets as the snowpack disappeared. The potential for convective cloud formation was also examined. Surface heating and entrainment outweighed adiabatic cooling and evaporation, resulting in the boundary-layer top relative humidity decreasing as the snow melted and turbulent mixing increased.