Sensitivity of Ku- and X-Band Radar Observations to Seasonal Snow in Ontario, Canada

Abstract

Radar scatterometer observations at 17.2 GHz and 9.6 GHz were made of the snow cover in mid-latitude agricultural fields, using the University of Waterloo scatterometer, to determine the sensitivity of the frequency-dependent radar response to snow water equivalent. Observations were made in alfalfa fields near Maryhill, Ontario during the 2013–2014 and 2014–2015 winter seasons. Additional observations in the 2014–2015 season were made at Englehart, Ontario. The natural snowpack was scanned using polarimetric observations throughout the season as snow accumulated. Conditions typical of the mid-latitude snow environment, including non-frozen ground, snow melt, freezing rain, and agricultural vegetation were observed to confound the relationship between backscatter and SWE or snow depth. With these special cases removed from the analysis, the strongest relationship is observed at 17.2 GHz where VV-polarized backscatter increased linearly by 0.35 dB/cm SWE. Further, the backscatter increases linearly by 0.10 dB/cm of snow depth. Relationships at 9.6 GHz are weaker. These findings are unique. They highlight the suitability and challenges of radar remote sensing for estimating snow accumulation in a mid-latitude environment and demonstrate the use of polarization signatures for identifying the effects of short vegetation prevalent in this environment.