Thirty years ago, a large mass of cloud moved north-east towards the shores of NW France and SE England: the infamous, poorly-forecast, October Storm. Thirty years on, Hurricane Ophelia sits close to the Azores ready to move north-east possibly towards the NW Europe. Two very different storms, but both are connected by the low risk but considerable potential that exists in the overlapping of the tropical and extra-tropical cyclone seasons that occurs in September and October.
I wanted to use this post to talk about this "crossover" period to have a discussion around how the hurricane season could impact the European Windstorm season and have a think around how this might produce tail events.
A brief (simple) climatology of ex-hurricanes close to NW Europe
To start with, just to get a feel for when ex-tropical systems pass close to NW Europe, I've taken a look at the frequency of events (in half-monthly intervals) in the NHC database (which is likely to be incomplete, especially as storms move out of tropical waters, but hopefully indicative). I've solely tallied up those storms North of Southern Spain and E of 20 degrees West.
You can see how this crossover "season" where tropical seasons impinge upon NW Europe typically is in September and the first half of October.
So - how does the tropical season typically interact with the European windstorm season? The next two headings deal with two different ways this can occur.
European windstorms enhanced by tropical air
Those (including myself) who woke up during the "Great October Storm" of 1987 were treated - or slightly horrified in my case - by the strength of the wind. You don't really expect gusts above 90 mph in suburban London. Looking beyond the wind strength, one of the more interesting aspects of the October Storm was the warmth preceding the strong winds. Warm air, its alleged origins off the eastern seaboard of the US maybe associated with Hurricane Floyd earlier in the month, was mooted to have been wrapped into the system. During the late evening in the southern UK ahead of the strong winds, some places reported 17c - unseasonably warm for night-time temperatures.
The chart below shows how anomalously warm the air was about 36 hours before the October storm crossed NW France and SE England and how this warmth is part of a wider pool of warm air brought up from the tropical Atlantic.
The October Storm is likely an example of how extra-tropical developments can be invigorated by pre-existing warm air that may have had its origins in the still-warm tropical Atlantic. Without going into painful detail, the presence of warmer-than-usual air in an extra-tropical system can give it the potential for greater intensification through the release of "latent heat" as all the warm air ascends into the extra-tropical system and cools. Clearly the overlapping of the two seasons in September and October present the best opportunity for this type of "modification" of extra-tropical windstorms to occur.
Should we improve our understanding around whether "early-season" European windstorms have the potential to be more damaging in the tail given there is warmer air present than is available mid-winter?
European Windstorms formed from Decaying Hurricanes
It's not just the fact that warmer air can get wrapped into our extra-tropical storms: hurricanes can exist quite far into the central Atlantic. Clearly at the time of writing, we have a tropical system, Ophelia, close to the Azores due to move towards Europe. It's likely to undergo "extratropical transition" there: a process where a tropical system encounters the mid-latitude jetstream and gradually morphs into a extra-tropical cyclone. These storms can also reinvigorate as an extra-tropical cyclone even though tropical element of the storm could not be supported over colder waters.
This isn't particularly unusual. The chart here shows the locations of a number of storms from the NHC archive since 1990 that have approached NW Europe having had tropical origins - I've limited the winds to 50 kts or greater (58 mph - right at the bottom end of where damage typically starts to begin):
Two tracks are highlighted here: the more southerly one is Hurricane Floyd from 1993, that was extra-tropical for a lot of its life as it accelerated across the Atlantic but deepened as an extratropical cyclone close to France, and Hurricane Lili from 1998 that brought strong winds to Ireland as an extra-tropical storm. These type of events clearly do have some potential for damage.
A Grey Swan: A Hybrid Tropical/Extratropical European Landfall
So what does a "hybrid" system on the boundary of the hurricane / extra-tropical windstorm seasons look like? I would imagine that we'd likely see a "perfect (hybrid) storm" coming from:
a) a hurricane forming and remaining strong far out into the Atlantic over seas that are - for the time of year - anomalously warm and able to support a hurricane
b) a rapid acceleration under the influence of a strong mid-latitude jet-stream, carrying the system rapidly toward Europe, potentially re-intensifying as an extra-tropical system owing to the extra warmth in the system
Here's an example of what I'm talking about. The chart below [ignore the colours for now] shows a dot at 34W, 36N [ignore the dot to the east of that for now]. This is the location, in NHC history, of pretty much the closest Cat 3 hurricane to Europe (I've used 95kt as a cut-off: practically a Cat 3). This was approximately 1400 miles from Portugal. The fastest forward speed in the history of tropical systems in this part of the Atlantic basin (not shown here) is around 45 mph. This means it would take around 30 hours to reach the Portuguese coastline.
Could a category 3 storm completely lose its tropical characteristics and Cat 3 strength in this time? It strikes me that with the optimum atmospheric and oceanic set-up, we probably can't rule out tropical cyclones making landfall in southern Europe. It's probably a tail event, but this simple analysis suggests that it's not unfeasible.
Additionally, I've talked in a previous blog post about the seemingly inexorable increase in oceanic temperatures commensurate with our warming globe in the past 40 years. The orange area here is the maximum extent of the 26.5c isotherm (broadly, the sea temperature able to support minimal tropical activity) in the last 35 years of SSTs. The closest Cat 3 event to Europe coincidentally sits right on the edge of this line and we can use this simply in the next step of the discussion.
So what if the seas warmed by, say, 1c? The yellow area shows the maximum extent of the 26.5c isotherm if the ocean warmed uniformly by 1c: something not necessarily unlikely over the next century. You can immediately see if we move our Cat 3 storm to the edge of this boundary, how much closer to the southern European coastline (and other European parts of Europe) this is, more like 1400 km to its closest point.
Food for thought that potentially in a warming world we could see more areas at risk from tropical cyclones - something for that has been discussed in a fairly recent paper, for the interested reader.
A lot of the above is merely a fairly crude thought experiment, but it is worthwhile understanding, especially in light of Ophelia's future track, how the tropics and extra-tropics could potentially interact when their seasons overlap. Is it worth double-checking that European insurance contracts would pay out if a hurricane rather than a windstorm made landfall?