Lake ice thickness (LIT), a thematic product of Lakes as an Essential Climate Variable (ECV), is sensitive to changes in air temperature and on-ice snow mass. Here, a novel and efficient analytic method (retracking approach) is presented for the estimation of LIT from Ku-band (13.6 GHz) radar altimetry data. The new retracker, referred to as LRM_LIT, is based on the physical modeling of the conventional radar echoes (also called Low Resolution Mode or LRM) over ice-covered lakes that show a characteristic step-like feature in their leading edge attributed to the reflection of radar waves at the snow-ice and ice-water interfaces. The method is applied to Jason-2 and Jason-3 data acquired over Great Slave Lake, Canada, over three ice seasons (2013-2016). As expected, the agreement between the Jason-2 and Jason-3 LIT estimates over their overlapping period (2016 ice season) is excellent with a mean bias error of 0.013 m and root mean square error (RMSE) of 0.024 m. LIT estimates from LRM_LIT are in good agreement with simulations from a thermodynamic lake ice model and in situ measurements with RMSE values of the order of few centimeters for the three winter seasons. The retracker also provides a robust way to assess the accuracy of LIT estimates which is in the order of 0.10 m when the ice cover is well established and prior to melt onset. In addition, LRM_LIT captures the seasonal transitions during the freeze-up and breakup periods and ice growth over different winter seasons, making it a promising method for monitoring inter-annual variability and trends in LIT from past and current conventional radar altimetry missions.
Read full abstract