Ocean-atmosphere interaction in the Atlantic storm track response

to climate change

Tim Woollings

With Jonathan Gregory, Joaquim Pinto, Mark Reyers and David Brayshaw

Storm track changes as described in AR4

Yin 2005

Storm tracks shift polewards, consistent with change in the eddy-driven jets.

Upper vs lower level Teq – Tpole

Standard deviation of 2-6 day MSLP (ctrs 1/10 hPa)2080-99 (SRESA1B) – 1960-2000 (20C3M)Ulbrich et al (2008)

Regional storm track response more complicated

20c3m

IPCC AR4

How much uncertainty arises from differing MOC responses?

Use 20c3m (20th century control) 1960-99 and sresa1b 2060-99.

Method 1: Regression across model space

Δ MOC

Δ TAt each point in space:

Using EOF1 of SST as a proxy for Δ MOC to expand the sample size:

Storm track response regressed on MOC response

Estimating the ensemble mean storm track from the ensemble mean MOC

Method 2: Compare coupled and slab models (all scaled to 3K warming)

Coupled models

Slab models

Coupled - slab

Method 3: Compare with response to MOC hosing

Summary

• There is considerable uncertainty in the response of the North Atlantic jet stream / storm track to anthropogenic forcing.

• Neither climate sensitivity or control biases can explain the spread.

• Differing MOC responses account for about half of the spread between different climate models.

• Slab runs and atmosphere-only experiments show that there is some causality from ocean to atmosphere.

• The storm track response should be considered a coupled problem.

Woollings, T., J. M. Gregory, J. G. Pinto, M. Reyers & D. J. Brayshaw (2012). Response of the North Atlantic storm track to climate change shaped by ocean-atmosphere coupling, Nature Geoscience. doi:10.1038/ngeo1438