High-resolution climate models show skill in hurricane variability

February 2, 2015
Hurricane frequency shown in models
North Atlantic tropical cyclone interannual variability for different model resolutions together with observed hurricanes. The solid line shows the ensemble mean, while the shading indicates the ensemble range.

Climate extremes such as hurricanes and typhoons can cause death and destruction for communities around the world. Scientists as part of the US CLIVAR Working Group on Hurricanes are interested in studying both the predictability (such as on multi-year timescales) and possible future changes in the frequency and intensity of such storms.

A hierarchy of models, ranging from typical climate resolutions using grid cells with 130km sides to resolutions more often found in weather forecasting models with 25km sides, have been used to study the impact on simulated tropical cyclone performance. In common with other studies, this paper finds that the higher resolution models are able to simulate observed interannual variability with considerable skill in basins such as the North Atlantic and Northwest Pacific. Part of the cause for this in the Atlantic seems to be due to interacting processes over Africa creating more structures from which hurricanes can grow.

In the future projection for 2100, there is a reduction in the frequency of these events by about 50% in the Southern Hemisphere, with changes in the Northern Hemisphere more varied. For example, the North Atlantic has less activity while the Pacific sees shifts in where the events happen, with increases occurring around Hawaii and the Central Pacific. One problem with the model used in the study is that the simulated storms have weak winds compared to the observations, which makes it difficult to say whether future storms are likely to become more intense or not in a warmer world.

Written by 
Malcolm Roberts, Met Office Hadley Centre

Malcolm J. Roberts1, Pier Luigi Vidale2, Matthew S. Mizielinski1, Marie-Estelle Demory2, Reinhard Schiemann2, Jane Strachan3, Kevin Hodges4, Ray Bell5, and Joanne Camp1

1Met Office Hadley Centre, Exeter, United Kingdom

2National Centre for Atmospheric Science, Department of Meteorology, University of Reading, Reading, United Kingdom

3Willis Research Fellow, Knowledge Transfer Partnership, University of Reading, Reading, United Kingdom

4Natural Environment Research Council Centre for Earth Observation, University of Reading, Reading, United Kingdom

5Department of Meteorology, University of Reading, Reading, United Kingdom