Climate Change Reshapes Britain’s Most Infamous Storms in Divergent Ways

The Great Storm of 1987, Storm Martin, Storm Arwen and Storm Ophelia each left deep scars on the collective memory of Britain, Ireland and France. Now a new study suggests that a warming climate will not treat these four storms equally. Using a large ensemble of climate model simulations, researchers have found that the frequency and intensity of cyclones following paths similar to each historical storm respond in markedly different ways to future warming, a result that challenges the usefulness of aggregate storm projections for planning purposes.

The study, published in the journal Weather and Climate Dynamics, was carried out by Farrell Morgan, Ben Harvey, Kevin Hodges and Oscar Martínez-Alvarado of the National Centre for Atmospheric Science and the University of Reading. The team applied a technique called the cyclone track analogue method, which involves searching climate model output for storms whose tracks closely resemble a specific historical cyclone during its development stage. By comparing how many such analogues appear in simulations of the present climate versus a future high-emission scenario, the researchers could quantify how the likelihood of each type of storm changes under warming, along with the hazards those storms would carry.

The climate simulations came from the CANARI Large Ensemble, a 40-member set of runs produced with the HadGEM3-GC3.1-MM global coupled model, the same model used in CMIP6. The ensemble covers the years 1950 to 2100, using historical forcing until 2014 and then the SSP3-7.0 scenario, which combines moderate-to-high greenhouse gas emissions with high aerosol emissions. The researchers compared a present-day period from 1980 to 2010 with a future period from 2070 to 2100, giving them 1200 simulated years in each window to work with. The model’s atmosphere component resolves features at roughly 60 kilometre grid spacing in the mid-latitudes, fine enough to capture individual cyclones.

To find analogues, the team first identified and tracked cyclones using 6-hourly 850 hectopascal relative vorticity, filtered to retain synoptic-scale features. Candidate storms were those reaching peak intensity within 300 kilometres of the historical cyclone’s position of maximum intensity, and analogues were those whose average track distance over the development stage fell below 500 kilometres. Storms whose peak vorticity exceeded 8.2 times ten to the minus five per second, the 90th percentile of all Northern Hemisphere tracks in the ERA5 reanalysis between 1980 and 2010, were classed as extreme analogues. For Ophelia, an additional filter required analogues to originate in the subtropical North Atlantic, between 0 and 60 degrees west and 25 to 35 degrees north, to ensure they represented post-tropical cyclones rather than ordinary mid-latitude lows.

The headline result concerns the Great Storm of 1987, the vigorous depression that crossed southern England on the night of 15 to 16 October 1987, killing at least 18 people, flattening millions of trees and producing gusts above 60 metres per second off the coast of Brittany. In the future simulations, the number of candidate cyclones reaching peak intensity in the same region actually falls, from 19.2 to 17.9 per decade. Yet the number of analogues rises sharply, from 2.5 to 2.9 per decade, an increase of 16.5 percent, because a larger fraction of the remaining storms follow the storm’s north-eastward path rather than approaching from the north-west. Extreme analogues of the Great Storm increase even more dramatically, by 52.3 percent, with their share of all analogues climbing from 17.3 to 30.4 percent.

Storm Arwen, which battered the UK with unusually strong northerly winds and heavy snow on 26 to 27 November 2021, cutting power to more than one million homes, points in the opposite direction. The frequency of Arwen analogues falls significantly in the future simulations, from 4.3 to 2.9 per decade, a reduction of 46.7 percent, driven both by fewer candidate cyclones and by a smaller fraction following Arwen’s southward track. Extreme analogues decline by 37.9 percent. The contrast with the Great Storm is striking because both storms reach peak intensity in the same general region near the North Sea, yet their analogue populations respond in opposite ways to the same background warming.

The other two cases show intermediate behaviour. Martin, which struck France on 27 December 1999 with a recorded central pressure of 965 hectopascals, gusts above 40 metres per second and economic losses of six billion US dollars, sees its analogue rate decline modestly from 0.83 to 0.75 per decade, a change not statistically significant against the variability across ensemble members. However, the number of extreme Martin analogues doubles, from 8 to 16, raising their relative share from 8.1 to 17.8 percent. Ophelia, the post-tropical cyclone that made landfall in Ireland on 16 October 2017 with gusts exceeding 53 metres per second off County Cork, shows a weak analogue increase of 8.9 percent, but extreme analogues rise by 50 percent. The researchers caution that this signal is of similar size to the model’s bias for extreme cyclones, and that the low frequency of post-tropical cyclone occurrences limits statistical confidence.

Intensity and hazard metrics also diverge by case. Mean track-maximum relative vorticity changes range from minus 1.1 percent for Arwen and Ophelia analogues to a statistically significant 9.1 percent increase for Great Storm analogues. Yet every set of extreme analogues intensifies, by 1.0 to 4.7 percent in vorticity, and all analogue sets show robust increases in track-maximum precipitation, ranging from 12.0 to 22.9 percent for all analogues and 18.8 to 36.5 percent for extreme analogues. Wind speed at 850 hectopascals, used as an estimator of potential surface gusts, rises significantly by 6.7 percent for Great Storm analogues but falls significantly for Arwen analogues, while Martin and Ophelia show little change in track-maximum winds. Upper-tropospheric jet winds near each storm strengthen across all analogue sets, by 3.1 to 9.2 percent, and by 8.1 to 19.5 percent for extreme analogues.

The underlying mechanism, the researchers argue, is a dipole in baroclinicity, the atmospheric temperature contrast that fuels cyclone growth. Measured by the low-level Eady growth rate, baroclinicity declines at high latitudes south and east of Greenland, consistent with weakened lower-tropospheric temperature gradients under Arctic amplification, while increasing over the central North Atlantic, likely linked to changing sea surface temperature gradients. This pattern favours cyclones approaching from the south-west, like the Great Storm, and suppresses those descending from the north, like Arwen. The model evaluation supports these conclusions: the CANARI LE overestimates total North Atlantic cyclone numbers by only 2.9 percent, with biases of 10 to 15 percent confined mainly to the sub-polar region south of Greenland, and biases near the UK, Ireland and France are considerably smaller, though the model underestimates extreme cyclone numbers regionally by 14.4 percent and slightly weakens the most intense storms.

The broader implication is that cyclone-specific responses can diverge substantially from the aggregate signal. Across the whole region, the ensemble projects a 10.7 percent decline in total cyclone numbers and a shift toward fewer weak storms but more intense ones, a pattern consistent with previous multi-model studies. Yet the four case studies show that these averages conceal opposing regional stories: one iconic storm type becoming more frequent and more severe, another becoming substantially rarer. Because each cyclone type carries distinct hazards, from Arwen’s damaging north winds to the Great Storm’s south-westerly gusts and heavy rain, the authors argue that adaptation and resilience planning cannot rely on aggregate storm trends alone. They note that further work is needed, including analysis of a wider range of cyclones and of the roles of diabatic processes and large-scale modes of variability, before the full spectrum of cyclone-specific responses under warming is understood.

Subject of Research: Projected climate change effects on the frequency and intensity of cyclone track analogues of four historical European storms

Article Title: Climate change effects on analogues of contrasting extratropical cyclones over the UK, Ireland and France

Article References: Morgan, F., Harvey, B., Hodges, K., & Martínez-Alvarado, O. (2026). Climate change effects on analogues of contrasting extratropical cyclones over the UK, Ireland and France. Weather and Climate Dynamics, 7(4), 2087-2105. https://doi.org/10.5194/wcd-7-2087-2026

Image Credits: AI Generated

DOI: 10.5194/wcd-7-2087-2026

Keywords: extratropical cyclones, climate change, storm analogues, Great Storm 1987, Storm Arwen, CANARI Large Ensemble, baroclinicity, SSP3-7.0 scenario, UK storm risk, post-tropical cyclones, HadGEM3-GC3.1-MM, precipitation extremes