Comparing elevation and freeboard from IceBridge and four different CryoSat-2 retrackers for coincident sea ice observations

The airborne IceBridge and spaceborne Cryosat-2 missions observe polar sea ice at different altitudes with different footprint sizes and often at different time and locations. Many studies use different retrackers to derive Cryosat-2 surface elevation, which we find causes large differences in the elevation and freeboard comparisons of IceBridge and Cryosat-2. In this study, we compare sea ice surface elevation and freeboard using 8 coincident CryoSat-2, ATM, and LVIS observations with IceBridge airplanes under flying the Cryosat-2 ground tracks. We apply identical ellipsoid, geoid model, tide model, and atmospheric correction to CryoSat-2 and IceBridge data to reduce elevation bias due to their differences. IceBridge's ATM and LVIS elevation and freeboard and Snow Radar snow depth are averaged at each CryoSat-2 footprint for comparison. The four different Cryosat-2 retrackers (ESA, GSFC, AWI, and JPL) show distinct differences in mean elevation up to 0.35 meters over leads and over floes, which suggests that systematic elevation bias exists between the retrackers. The mean IceBridge elevation over leads is within the mean elevation distribution of the four Cryosat-2 retrackers. The mean IceBridge elevation over floes is above the mean elevation distribution of the four Cryosat-2 retrackers. After removing the snow depth from IceBridge elevation, over floe, the mean elevation of IceBridge is within the mean elevation distribution of the four Cryosat-2 retrackers. By identifying the strengths and weaknesses of the retrackers, this study provides a mechanism to improve freeboard retrievals from existing methods.