Atmosphere surface storm track response to resolved ocean mesoscale in two sets of global climate model experiments

• Published in: Climate dynamics 2019-02, Vol.52 (3-4), p.2067-2089
• Main Authors: Small, R. Justin, Msadek, Rym, Kwon, Young-Oh, Booth, James F., Zarzycki, Colin
• Format: Article
• Language: English
• Subjects: Air-sea flux; Analysis; Atmospheric models; Baroclinic mode; Bias; Boundary currents; Boundary layer stability; Boundary layers; Climate; Climate models; Climatology; Coastal environments; Computer simulation; Cyclones; Dynamic meteorology; Earth; Earth and Environmental Science; Earth Sciences; Eddies (Fluid dynamics); Experiments; Fronts; Geophysics/Geodesy; Global climate; Global climate models; Heat flux; Heat transfer; Latent heat; Latent heat flux; Low level; Low pressure systems (Meteorology); Meridional wind; Meteorology & Atmospheric Sciences; Ocean circulation; Ocean models; Oceanic fronts; Oceanography; Oceans; Reduction; Resolution; Sciences of the Universe; Sea surface; Sea surface temperature; Sensitivity; Stability; Storm tracks; Storms; Surface stability; Surface temperature; Test procedures; Tracking; Troposphere; Vortices; Western boundary currents
• ISSN: 0930-7575; 1432-0894
• DOI: 10.1007/s00382-018-4237-9

It has been hypothesized that the ocean mesoscale (particularly ocean fronts) can affect the strength and location of the overlying extratropical atmospheric storm track. In this paper, we examine whether resolving ocean fronts in global climate models indeed leads to significant improvement in the simulated storm track, defined using low level meridional wind. Two main sets of experiments are used: (i) global climate model Community Earth System Model version 1 with non-eddy-resolving standard resolution or with ocean eddy-resolving resolution, and (ii) the same but with the GFDL Climate Model version 2. In case (i), it is found that higher ocean resolution leads to a reduction of a very warm sea surface temperature (SST) bias at the east coasts of the U.S. and Japan seen in standard resolution models. This in turn leads to a reduction of storm track strength near the coastlines, by up to 20%, and a better location of the storm track maxima, over the western boundary currents as observed. In case (ii), the change in absolute SST bias in these regions is less notable, and there are modest (10% or less) increases in surface storm track, and smaller changes in the free troposphere. In contrast, in the southern Indian Ocean, case (ii) shows most sensitivity to ocean resolution, and this coincides with a larger change in mean SST as ocean resolution is changed. Where the ocean resolution does make a difference, it consistently brings the storm track closer in appearance to that seen in ERA-Interim Reanalysis data. Overall, for the range of ocean model resolutions used here (1° versus 0.1°) we find that the differences in SST gradient have a small effect on the storm track strength whilst changes in absolute SST between experiments can have a larger effect. The latter affects the land–sea contrast, air–sea stability, surface latent heat flux, and the boundary layer baroclinicity in such a way as to reduce storm track activity adjacent to the western boundary in the N. Hemisphere storm tracks, but strengthens the storm track over the southern Indian Ocean. A note of caution is that the results are sensitive to the choice of storm track metric. The results are contrasted with those from a high resolution coupled simulation where the SST is smoothed for the purposes of computing air–sea fluxes, an alternative method of testing sensitivity to SST gradients.