A Validation of ERA5 Reanalysis Data in the Southern Antarctic Peninsula—Ellsworth Land Region, and Its Implications for Ice Core Studies

Dieter Tetzner,Claire Allen,Elizabeth Thomas

doi:10.3390/geosciences9070289

Dieter Tetzner, Claire Allen + Show 1 more

Open Access

https://doi.org/10.3390/geosciences9070289

Copy DOI

Journal: Geosciences	Publication Date: Jun 29, 2019
Citations: 117	License type: CC BY 4.0

Affiliation: British Antarctic Survey, University of Cambridge

Abstract

Climate reanalyses provide key information to calibrate proxy records in regions with scarce direct observations. The climate reanalysis used to perform a proxy calibration should accurately reproduce the local climate variability. Here we present a regional scale evaluation of meteorological parameters using ERA-Interim and ERA5 reanalyses compared to in-situ observations from 13 automatic weather stations (AWS), located in the southern Antarctic Peninsula and Ellsworth Land, Antarctica. Both reanalyses seem to perform better in the escarpment area (>1000 m a.s.l) than on the coast. A significant improvement is observed in the performance of ERA5 over ERA-Interim. ERA5 is highly accurate, representing the magnitude and variability of near-surface air temperature and wind regimes. The higher spatial and temporal resolution provided by ERA5 reduces significantly the cold coastal biases identified in ERA-Interim and increases the accuracy representing the wind direction and wind speed in the escarpment. The slight underestimation in the wind speed obtained from the reanalyses could be attributed to an interplay of topographic factors and the effect of local wind regimes. Three sites in this region are highlighted for their potential for ice core studies. These sites are likely to provide accurate proxy calibrations for future palaeoclimatic reconstructions.

Highlights

In the last decades, the Antarctic Peninsula (AP) and West Antarctica (WA) have been experiencing dramatic climate changes [1]
Mean values of the performance indicators show that even though ERA5 datasets increase the normalized mean absolute error (NMAE) and normalized bias (NBIAS) by 4% and 6%, respectively, normalized root-mean-square error (NRMSE) decreases by 1%
Our results show that in both reanalyses mean surface air temperatures are cold biased in the coastal (ERA-Interim:−1.2 ◦ C, ERA5:−0.51◦ C) area, but slightly warm biased at the escarpment (ERA-Interim:+0.28◦ C, ERA5:+0.14 ◦ C)

Summary

Introduction

The Antarctic Peninsula (AP) and West Antarctica (WA) have been experiencing dramatic climate changes [1]. Instrumental records suggest that since the 1950s, AP surface air temperatures are among the most rapidly warming on Earth [1,2] This has been accompanied by ice shelf collapse and accelerated glacier flow, enhanced mass loss from melting glaciers [3] and increased snowfall [4,5,6], all having a direct and measurable impact on global sea levels [7,8]. In addition to the station network, since the early 1980s, several automatic weather stations (AWS) and weather transmitters/GPS stations (hereinafter referred to as AWS) have been deployed in the region [9] These stations have helped to improve the spatial coverage of meteorological observations, as well as provide data from remote locations. The region stretching between the southern AP and continental WA, the Ellsworth Land (EL) region, is still data sparse, hindering our ability to determine how far the climate extremes observed in the northern AP extend into the continent and to place these recent changes in a longer-term context [10]

Objectives

Methods

Discussion

Conclusion