LAUSR

Abstract As one of the most remote and inaccessible regions in the world, the sea level in the Arctic is still today in large parts uncertain. Sea level estimates from GRACE and satellite altimetry show disagreements in areas of the Arctic while only few in-situ measurements are present to validate satellite products. By using an independent in-situ based dataset of hydrographic data (DTUSteric), this study compares different available datasets from GRACE and altimetry. The best agreement (R = 0.76) is reached between the combination of DTUSteric and the JPL Mascon solution and the altimetry product from Centre of Polar Observation and Modelling (CPOM). A large residual signal is found in the East Siberian Sea, an area with no in-situ observations and in general uncertain satellite observations. The spatial correlation coefficients ranging from 0.32 to 0.76 reflects a large number of unknowns and uncertainties, with coarse GRACE resolution, extensive interpolation and lack of in-situ data in parts of the Arctic Ocean and uncertainty of the altimetric products being the largest sources of error. In general, we recommended that satellite observations are used carefully in the Arctic. Even this study shows that steric estimates from in-situ observations to some extent can explain the gap between GRACE and altimetry, the observed sea level trend in significant regions of the Arctic cannot be validated with in-situ hydrographic data.