Estimating near-surface climatology of multi-reanalyses over the Greenland Ice Sheet

Abstract

This study uses meteorological records from Automatic Weather Stations (AWSs) to estimate the performance of global reanalysis products for monthly air temperature, relative humidity and wind speed over the Greenland Ice Sheet (GrIS). These products include the fifth generation European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis (ERA5), ECMWF Interim Reanalysis (ERAI), Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2), Climate Forecast System Reanalysis Version 1/Climate Forecast System Version 2 (CFSRv1/CFSRv2), and Japanese 55-year Reanalysis (JRA-55). The global reanalysis products generally perform better in summer than in winter, and their qualities vary by glaciological regime. No reanalysis is clearly identified as the optimal dataset for all meteorological parameters, seasons and regions. For all reanalyses, warm biases are observed in the accumulation zone, but cold biases are observed in the ablation area of the GrIS. ERAI, ERA5 and JRA-55 underestimate relative humidity during any month. While MERRA-2 overestimates wind speeds, underestimates are found for the other reanalyses excluding JRA-55 during all months. Despite the robust agreement between the AWS time series for all three variables and each reanalysis product averaged over the ice sheet, sudden jumps occur in annual mean wind speed in CFSR, and in annual relative humidity in JRA-55 from 2010 to 2011.