Validation of reanalysis Southern Ocean atmosphere trends using sea ice data

William R Hobbs,Andrew R Klekociuk,Yuhang Pan

doi:10.5194/acp-20-14757-2020

Abstract

Abstract. Reanalysis products are an invaluable tool for representing variability and long-term trends in regions with limited in situ data, and especially the Antarctic. A comparison of eight different reanalysis products shows large differences in sea level pressure and surface air temperature trends over the high-latitude Southern Ocean, with implications for studies of the atmosphere's role in driving ocean–sea ice changes. In this study, we use the established close coupling between sea ice cover and surface temperature to evaluate these reanalysis trends using the independent, 30-year sea ice record from 1980 to 2010. We demonstrate that sea ice trends are a reliable validation tool for most months of the year, although the sea ice–surface temperature coupling is weakest in summer when the surface energy budget is dominated by atmosphere-to-ocean heat fluxes. Based on our analysis, we find that surface air temperature trends in JRA55 are most consistent with satellite-observed sea ice trends over the polar waters of the Southern Ocean.

Highlights

Atmospheric trends in the southern high latitudes have global importance
Sea ice trends are not themselves the focus of this work, except as an independent validation of reanalysis SAT, the observed sea ice concentration (SIC) trends are shown in Fig. 2 for illustrative purposes. (Note that while the trends are aggregated in Fig. 2 into seasons defined by sea ice melt/growth, for the SAT validation we used monthly trends, shown in the Supplement.) The trend patterns are well-established and have been described in many previous studies (e.g. Parkinson and Cavalieri, 2012; Hobbs et al, 2016; Comiso et al, 2017) and can be broadly summarized as a decrease in the Amundsen and Bellingshausen seas (60–120◦ W), with compensating increases in the western Ross Sea (150–180◦ E), Weddell Sea and King Haakon VII Sea (50◦ W–30◦ E)
We find several reanalysis products that reproduce reasonable surface air temperature trends, with JRA55 showing the consistently highest agreement with observed sea ice throughout the year

Summary

Introduction

Atmospheric trends in the southern high latitudes have global importance. Wind patterns are essential for driving the Southern Ocean overturning, which is responsible for most of the global ocean’s uptake of anthropogenic heat and approximately half its uptake of anthropogenic carbon (Frolicher et al, 2015). King and Harangozo (1998) demonstrated a close link between Antarctic Peninsula station temperature and local sea ice changes, Massonnet et al (2013) were able to reproduce Antarctic sea ice variability in a model driven by SAT, and both Kusahara et al (2017) and Schroeter et al (2018) showed the important role that thermodynamic forcing has in Antarctic sea ice trends We argue that the weak SAT–SIC relationship in summer is due to the direction of ocean–atmosphere heat flux in those months; since the net balance in the sea ice zone is from atmosphere to ocean, the surface energy budget is more a response to – rather than a driver of – the near-surface atmosphere

Data and method

Evaluation of SAT based on sea ice trends

Conclusions