Saturday, October 5, 2019

Hurricanes, Ireland and Lorenzo

There has been a lot of discussion around hurricanes and Ireland this past week and a bit, associated with the remnants of Hurricane Lorenzo, coming as it has climatologically speaking hot on the heels of ex-hurricane Ophelia. Lorenzo ended up having relatively minor impacts upon the island of Ireland and there have been accusations it was over-hyped. There have also been, frankly wild, messages around hurricanes, climate change and the future impacts on Ireland bandied around. I thought it may be useful to put some thoughts down on the matter.

Was Lorenzo unusual?

While there are records for parts of the North Atlantic basin that go back for in excess of a Century, we only started observing the whole basin with the advent of satellite records in the 1970s. The eastern Atlantic basin before this time was fitfully observed, which means we can't be certain about earlier behaviour. Hurricanes are also infrequent events and it is thus hard to know whether single events are unusual in the context of a short record - did we just not observe such events by chance in the short record rather than them actually being truly unusual?

Caveats out of the way, Lorenzo was by far and away an outlier geographically in where Category 5 hurricanes (the strongest classification) have been known to have occurred. You'll note that for the <24 hours it was likely at category 5 it was very far away from Ireland.
Unusualness of the position of Lorenzo when category 5 in a historical record context. Source: Robert Rohde (@RArohde)

Other aspects of Hurricane Lorenzo were far more typical. The storm started out as a disturbance coming off Africa, moved westwards, strengthened, recurved to the north then north-east before becoming entrained in the mid-latitude circulation and transitioning to a post-tropical storm.
Storm histories for all recorded historical systems including their genesis, intensification and eventual decay either in-situ in the tropics or as extra-tropical systems. Colours denote intensity (see colour bar). Source: NOAA


Were the potential impacts on Ireland overhyped?

Let's start with the official forecasts from Met Eireann, which as the national meteorological service should be seen as the sole source for authoritative information (similarly the UK Met Office for Northern Ireland). From about a week out it was consistently suggested by the premier global forecast model from ECMWF that the post-tropical remnants of Lorenzo could impact Ireland.
 ECMWF deterministic model for 00Z on 4/10 initialised a week in advance. Note the tightly packed isobars off Cork which were the remnants of Lorenzo as forecast at that time. Source: www.wetterzentrale.de

By the start of the week, based upon continued run to run consistency of the forecast for the post-tropical remnants of Lorenzo to be over / near Ireland it was absolutely right of Met Éireann to start providing advice to the public warning of potential disruption. However, there was still considerable uncertainty in the track and intensity which persisted until much nearer the time. This is not unusual - predicting how a tropical storm system will interact with the mid-latitude system is not a simple problem. By 48 hours in advance and with Lorenzo undergoing extra-tropical transition just north of the Azores the predicted state had changed slightly:
Forecast of ex-hurricane Lorenzo initialised 48 hours in advance from ECMWF. Source: www.wetterzentrale.de

However, even this close in there were substantial disagreements between predictions arising from different global modelling groups. ECMWF and the Met Office bought the system in over Donegal / Mayo rapidly weakening as it moved over to Wales and SW England. Several other modelling groups took it up to the west of Scotland or even recorded it toward Greenland.

It was only this close in that Met Eireann issued amber and yellow specific warnings. These were appropriate to what the models were suggesting would occur. These warnings were then updated regularly and downgraded appropriately as it became clear that the system was weakening. Impacts and measured windspeed were clearly entirely consistent with the amber warnings issued. Thus Met Eireann did not overhype the storm or its likely impacts.

Members of the media and certain social media may be less innocent in the matter. Lorenzo was a category 5 hurricane several thousand kilometres and about 7 Days prior to eventual arrival in Ireland. The storm was never going to be that strong anywhere near Ireland. Yet, much reporting implied that it was still a category 5 hurricane. Arguably responsible reporting should have concentrated upon the actual state of the system upon arrival on Irish shores (a strong post-tropical system) rather than what it had been in, meteorologically speaking, a former lifetime.

What Lorenzo actually looked like at landfall in Donegal according to the Met Office analysis: i) confirms the typical post-tropical system structure and ii) the ECMWF model forecast from two days or so prior.
Met Office analysis chart for 00Z on Friday 4th Oct


Could Ireland be hit by an actual hurricane?

There have been some suggestions reported that Ireland could be hit by a hurricane in the near future. Not putting too fine a point on this: those suggestions are utterly false. Hurricanes are warm cored systems that require the sustained availability of ocean waters over 26 degrees to maintain their structure. Present day climatological sea surface temperatures of the surface Ocean (not coastal waters) in the proximity of Ireland are about 16 degrees:
Climatological SSTs for September from NOAA. Hurricanes can only be maintained over regions above 26 degrees (the oranges) which are very far from Ireland.

The 26 degree isotherm is several thousand kilometres to the SW of Ireland. Tropical systems can only transect for a very limited time over ocean surface temperatures below 26 degrees without losing their tropical characteristics:

Model output from the GFS model showing the transition from a warm cored hurricane to an extra-tropical system for Lorenzo. Colours denote temperatures, white lines denote isobars. Note how the system takes on very different structure and how pressure field broadens (weaker peak winds). Source: Simon Lee (@simonleewx)

That is before getting into finer points such as the effects of vertical shear that can rip apart a tropical system. The hurricane season in the North Atlantic is driven by the availability of a low shear environment which is a seasonally occurring phenomenon. Shear in the midlatitudes would tend to tear apart storms even if, somehow, the ocean surface temperature warmed sufficiently.

For a hurricane to hit Ireland would require late summer season sea surface temperatures to be sustained all the way from the tropics to Ireland at or above 26C, along with a sufficiently low shear environment along the path. That would require warming of the North Atlantic basin of the order 10 degrees. Because oceans warm slower than land this would equate to a global mean surface temperature warming of even more - likely somewhere north of 15 degrees. That is a warming not seen in even the most sensitive climate models under the most pessimistic emissions scenarios by 2100.

So, no, Ireland is not about to get hit by a hurricane. It is extremely unlikely to occur within the lifetime of anyone reading this blog irrespective of our collective choices as global society.

There are more than sufficient reasons for concern and subsequent undertaking of meaningful actions to address the significant challenges of climate change. This is not one of them.

But what about Ophelia?

Ophelia was also extra-tropical (an ex-hurricane) when it hit. Despite having hurricane force winds it was not a hurricane when it hit County Cork. It was, however, a very special case, transitioning much closer to Ireland than any known system. This was a very particular dynamical set-up whereby the hurricane interacted with a strong polar jet streak which maintained an outflow aloft and allowed it to retain warm-core tropical characteristics over colder than usual ocean surface temperatures. Even that  somewhat unique circumstance was insufficient to maintain tropical storm structure all the way to Ireland.