Investigating Relationships Between Land-Cover, Forest Structure, and In-Situ and Airborne Passive Microwave Snow Water Equivalent in a Boreal Forest Environment

The Meteorological Service of Canada (MSC) has developed a suite of land-cover sensitive algorithms to extract snow water equivalent (SWE) estimates from satellite passive microwave brightness temperatures. In the boreal forest, however, accurate passive microwave SWE retrievals using the MSC coniferous forest algorithm are hampered by consistent under-estimation. In February 2003, a collection of in situ and passive microwave airborne SWE datasets were acquired in a boreal forest environment in central Saskatchewan, Canada. In addition to the SWE datasets, information on land-cover from satellite imagery, and vegetative parameters from a forest inventory were also assembled. The analysis of the airborne passive microwave SWE estimates compared to in situ SWE measurements highlighted a distinct underestimation of airborne SWE estimates in all forested land-covers and an overestimation of SWE in open clear-cut areas. The magnitude of SWE underestimation in forested land-covers rises with an increase in the amount of forest vegetation-cover present. More specifically, the underestimation appears to be a function of increased canopy closure, rather than an increasing number of tree stems per hectare, as there is a strong negative correlation between airborne SWE estimates and increased canopy closure while a weak positive correlation exists between airborne SWE estimates and increasing number of tree stems per hectare. Better understanding of the relationships between land-cover, forest stand properties and snow cover variables can improve passive microwave SWE estimation in a boreal forest environment.