Sea Level Change in the Arctic Ocean: Quantifying Contributions to Present-Day Arctic Ocean Sea Level Change

Abstract

The Arctic Ocean is in the region with the fastest warming on Earth and is also the most inaccessible ocean in the world. Deglaciation of ice, ocean freshening, ocean warming changes the Arctic Ocean sea level and the relation to the global ocean circulation. Satellite measurements of Arctic sea level is challenged by varying sea ice and large ice mass changes on land. This PhD-thesis aims to quantify each of the contributions to sea level change in the Arctic Ocean and thereby improve the comprehension of the effects of climate change and validate sea level observations from remote sensing on the Arctic Ocean. Radar altimeters have observed sea level changes in the Arctic since 1991, but the sea ice cover significantly limits the observations. Challenges of separating younger sea ice and meltponds on top of ice from leads between sea ice floats and open ocean furthermore exacerbates altimetric observations, resulting in large discrepancies among products. Mass changes observed by the GRACE-satellites since 2002 are affected by large deglaciation on land that overshadows smaller changes in the ocean. Even though different procedures for separating changes on land and in ocean estimates of mass changes vary between 2 and 15 mm y−1 in the interior Arctic. To estimate sea level anomalies caused by changes in ocean salinity (halosteric) and temperature (thermosteric), studies often rely on the difference between GRACE and altimetry, which is an arguable approach considering the large ambiguity of the satellite products. In this study, an original approach is taken to create a satellite independent steric sea level product, called DTU Steric. From over 300k temperature and salinity profiles is a Arctic temperature and salinity grid compiled, which is used to compute steric changes. The product shows increased freshwater in the Arctic Ocean gives a significant sea level rise at the west coast of the Russian Arctic (10-20 mm y−1) and in the Beaufort Sea (10-15 mm y−1). Ocean warming causes a more uniform and smaller sea level rise (1-5 mm y−1) in most of the Arctic Ocean, while cooler water results in sea level fall in the east Russian Arctic (2 - 4 mm y−1). By combining DTU Steric and GRACE mascons from NASA JPL, good correlation (R=0.61- 0.76) is reached with altimetric observations. Large discrepancies are however in particular evident along the Russian Arctic. To validate the results with tide gauges along the coast requires accurate estimates of vertical land movement (VLM). In particular in the Arctic, where glacial isostatic adjustment (GIA) from prehistoric deglaciation and elastic VLM from changes in present-day ice loading (PDIL), creates significant VLM that affects tide gauges. By applying a high-resolution ice model from 1995-2015, a dynamic Arctic-wide elastic VLM-model is created. The measured VLM at 54 GNSS-sites north of 50°N is explained by the elastic VLM-model with GIA. The model shows, that the elastic VLM from PDIL is non-negligible in regions far away from glaciated areas. In Denmark creates PDIL an uplift of ∼ 0.5 mm y−1, which is significantly mitigating the associated barystatic sea level rise. Utilizing both DTU Steric and the VLM-product, are sea level trends from 1995-2015 from altimetry and at twelve VLM-corrected Arctic tide gauges assessed. Instead of GRACE, the modeled geoid changes from PDIL and GIA and dynamic mass changes from an oceanographic model are used as estimate for the mass component. For 11 of 12 tide gauges and 98% of Arctic Ocean is the sea level trend xplained within the standard error (1σ). Uncertainties associated with the dynamic mass contribution and the steric height sin areas with poor hydrographic data coverage attenuates the precision. By applying latest and upcoming satellite-missions can significantly improve the validation of the results presented.