Faster decline and higher variability in the sea ice thickness of the marginal Arctic seas when accounting for dynamic snow cover

Robbie D C Mallett,Julienne C Stroeve,Michel Tsamados,Vishnu Nandan,Jack C Landy,Rosemary Willatt,Glen E Liston

doi:10.5194/tc-15-2429-2021

Abstract

Abstract. Mean sea ice thickness is a sensitive indicator of Arctic climate change and is in long-term decline despite significant interannual variability. Current thickness estimations from satellite radar altimeters employ a snow climatology for converting range measurements to sea ice thickness, but this introduces unrealistically low interannual variability and trends. When the sea ice thickness in the period 2002–2018 is calculated using new snow data with more realistic variability and trends, we find mean sea ice thickness in four of the seven marginal seas to be declining between 60 %–100 % faster than when calculated with the conventional climatology. When analysed as an aggregate area, the mean sea ice thickness in the marginal seas is in statistically significant decline for 6 of 7 winter months. This is observed despite a 76 % increase in interannual variability between the methods in the same time period. On a seasonal timescale we find that snow data exert an increasingly strong control on thickness variability over the growth season, contributing 46 % in October but 70 % by April. Higher variability and faster decline in the sea ice thickness of the marginal seas has wide implications for our understanding of the polar climate system and our predictions for its change.

Highlights

Sea ice cover moderates the exchange of moisture, heat and momentum between the atmosphere and the polar oceans, influencing regional ecosystems, hemispheric weather patterns and global climate
How does the variability in mW99 and SnowModel-LG at a given point compare to the values recorded at Soviet drifting stations published by Warren et al (1999)? These values for interannual variability are not currently used in sea ice thickness retrievals
SnowModel-LG snow-depth variability at a given point is significantly higher, ranging from ∼ 75 % of the drifting station values in October to ∼ 115 % by the end of winter. We present this analysis of the point-like snow variability to illustrate that mW99 does not introduce enough variability at a given point to match that observed at drifting stations from year to year

Summary

Introduction

Sea ice cover moderates the exchange of moisture, heat and momentum between the atmosphere and the polar oceans, influencing regional ecosystems, hemispheric weather patterns and global climate. Sea ice thickness (SIT) is a key characteristic of the sea ice cover, as thicker sea ice weakens the coupling between the ocean and atmosphere systems. Thicker sea ice is more thermally insulating and limits heat transfer from the ocean to the atmosphere in winter and consequent thermodynamic growth (Petty et al, 2018a). The thickness of sea ice during snow accumulation dictates whether the sea ice surface drops below the waterline, potentially increasing thermodynamic sea ice growth through the formation of snow ice (Rösel et al, 2018). The impact of the endof-winter SIT distribution persists into the melt season with thick sea ice decreasing the transmission of solar radiation to the surface ocean and reducing the potential for in- and under-ice primary productivity (Mundy et al, 2005; Katlein et al, 2015). Correct assimilation of ice thickness into models offers opportunities for prediction of the sea ice state on seasonal timescales (Chevallier and Salas-Mélia, 2012; Blockley and Peterson, 2018; Schröder et al, 2019)

Methods

Results

Discussion

Conclusion