Mapping Arctic Sea Ice Concentration Using Satellite Altimetry

Abstract

Remote sensing techniques have been used to observe Arctic sea ice since the 1970s, revealing widespread declines in extent and a transition to a predominantly seasonal ice cover. These changes have wider impacts on local communities, ecosystems, economies, and climate. Ensuring a long-term, replicable climate record from satellite observations is essential to monitor current trends and project future changes. Currently, Arctic-wide sea ice concentration is only available from passive microwave radiometers, which can lead to uncertainty in trends. This thesis aims to develop new observations of Arctic sea ice concentration from satellite altimetry by evaluating and improving sea surface classification. The first results chapter develops a methodology for evaluating and improving CryoSat-2 (Ku-band radar) lead and floe density estimates in the Northwest Passage using near-coincident satellite imagery, ultimately reducing the root mean squared difference from 20% to 5% and 47% to 6% for leads and floes, respectively, by developing a bias adjustment. This adjustment is applied in the second results chapter to generate a novel sea ice concentration record in the Canadian Arctic from CryoSat-2. The results are in high agreement with other products, presenting a complementary alternative to existing methods. Finally, in the third results chapter, bias-adjusted radar altimetry observations from CryoSat-2 and Sentinel-3, and laser altimetry observations from ICESat-2, are aggregated to map leads and floes across the Arctic, paving the way for higher spatial and/or temporal resolution observations than is possible with a single sensor. Collectively, these results demonstrate that satellite altimetry can be used to map sea ice concentration and that multi-sensor observations can be successfully combined to enhance sampling, thus advancing the use of satellite altimetry for sea ice monitoring.

Date of Award	18 Dec 2025
Original language	English
Awarding Institution	Northumbria University
Supervisor	Andrew Shepherd (Supervisor), Isobel R. Lawrence (Supervisor) & Rachel L. Tilling (Supervisor)