Estimating snow evaporation with GPS derived precipitable water vapour

Precipitable water vapour estimates derived from Global Positioning System (GPS) zenith wet delay measurements were used in conjunction with a basic snow evaporation model to verify observations of snow evaporation in an open urban area. A field study conducted at the University of Calgary campus on urban snow processes monitored snow evaporation for three different varieties of snow in varying stages of ripening between February and April of 2001. Evaporation rates for natural fresh snow, natural ripened snow, and mechanically displaced snow were then compared to the rates derived using the SuomiNet GPS. Relationships between simulated and observed evaporation rates for natural ripened and mechanically displaced snow were in very good agreement, with coefficients of determination of 76 and 85%, respectively. The preciptable water vapour component accounted for 38 and 47%, respectively, of the linear variation between the observations and the model. The wind accounted for the remainder of the linear variation. The analysis for natural fresh snow produced a coefficient of determination of only 48%, which was only significant at the 10% level. However, the wind component of the model accounted for 100% of the linear variation explained by the model. In general, mechanically displaced snow experienced three times less evaporation than natural snow. Model coefficients reflected the aerodynamic properties of the snow as well as the path length of precipitable water vapour measurement. The coefficients ranged from approximately 0.0107±0.0028 km−1 for natural snow and 0.0034±0.0008 km−1 for mechanically displaced snow.