Snow Thickness Estimation on First-Year Sea Ice from Microwave and Optical Remote Sensing and Melt Modelling

• Main Author: Zheng, Jiacheng
• Other Authors: Yackel, John, Else, Brent, Michael, Sideris
• Format: Master Thesis
• Language: English
• Published: Graduate Studies 2017
• Subjects: Physical Geography; Remote Sensing; Sea ice; Modelling; Snow; Microwave; Cryosphere; Canadian Arctic; Arctic; Canada; Dease Strait; Nunavut
• Online Access: http://hdl.handle.net/11023/3574; https://doi.org/10.11575/PRISM/27807

Late-winter snow thickness on first-year sea ice is estimated based on the duration of snowmelt. The study encompasses the late-winter to advanced-melt period. The beginning of snowmelt is detected using space-borne C-band microwave scatterometer measurements, and the end of snowmelt is detected using optical satellite measurements. The snowmelt duration is then used to invert a degree-day snowmelt model based on air temperature, and a melt coefficient is calibrated with in situ observations. The modelled snow thickness estimation is validated with distributed in situ measurements of snow thickness throughout Dease Strait, Nunavut, Canada. The mean snowmelt duration for the study sites is 24.6 ± 1.2 days, and the estimated mean snow thickness is 14.7 ± 3.0 cm. The overall performance of the model reveals a RMSE of 27.1% and a bias of 1.8%. The methodology shows promise, and it can easily be scaled up to estimate snow thickness on a regional basis.