Fig. 4: The contribution of remote forcing to high-pressure driven adiabatic processes over the GrIS during 1980–2018. | Nature Communications

Fig. 4: The contribution of remote forcing to high-pressure driven adiabatic processes over the GrIS during 1980–2018.

From: Discrepancies between observations and climate models of large-scale wind-driven Greenland melt influence sea-level rise projections

Fig. 4

Spatial patterns corresponding to the leading internal mode of co-variability between Northern Hemisphere annual mean a ERA5 500hPa geopotential height (Z500) and b sea surface temperatures (ERSSTv5, SST) (30°S–55°N). The corresponding expansion coefficients (EC, shown as units of standard deviations) for Z500 (blue) and SST (ERSSTv5, orange) as revealed by maximum covariance analysis (MCA, Methods) also shown in c. The shared fraction of covariance (SFC) is 54.6% and the correlation between the ECs is r = 0.77. The area-weighted global mean is removed before MCA from Z500 and the 60°S–60°N mean from SST. Panel c also shows the spatially averaged GrIS mass balance (MB) anomaly time series from The IMBIE Team6 (purple) and the correlation between the MB and the Z500/SST expansion coefficients over 1980–2018 in the legend (with brackets showing correlations over 1990–2012). In a, the regression of Z500 EC(1) onto the 200 hPa streamfunction (Ψ200) calculated from ERA5 is shown with contours overlaid (unit: 106m2s−1). Panel d shows the regression of Z500 EC(1) onto the MAR-simulated (Methods) surface air temperatures (SAT) (shading) and ERA5 200hPa horizontal winds (blue arrows) for the GrIS.

Back to article page