Status:Ongoing analysis Material from: Linus, Tim, Fernando, David
Discussed in the following Daily reports:
Picture
1. Impact
On the 24 July a low pressure system crossed the Channel and produced significant amounts of rainfall along the South-East of England with many places reported amounts exceeding 30 mm in 24 hours. The system moved then along the coast of Netherlands, Germany and S Denmark. Here the rainfall amounts exceeded 30 mm as well. On the next day the cyclone gave considerable amounts of rain in southern Scandinavia.
2. Description of the event
The plot below show the 24-hour accumulated precipitation 24 July 06z-25 July 06z.
Note also that the system continued to deepen as it moved on towards N Europe, giving extreme gusts (for July) over the Netherlands in particular, but also notable gales in other countries such as Germany. This was a disruptive event with many trees down, and purportedly the worst July windstorm on record in the Netherlands. The plot below shows maximum gusts in kts reported between 00Z Fri 24th and 00Z Sun 26th (not clear why some Dutch obs are missing). Note that in summer 40kts is a lot.
3. Predictability
3.1 Data assimilation
3.2 HRES
The plot below show stamp map with all HRES forecasts valid for the same date and time with the position/intensity errors (reference Met Office analysis).
The plots below show the time-series of scores for Europe. It is evident that the HRES scores was affected by this case.
Evaluation of cy41r2 HRES
24-hour precipitation 24 Jul 06z - 25 Jul 06z
41r1
41r2
24-hour maximum wind gusts (25 Jul 00z - 26 Jul 00z)
41r1
41r2
3.3 ENS
The plots below shows stamp maps of MSLP and 24-hour precipitation, all valid 24 July.
MSLP and precipitation
The plot below shows the probability of exceeding the 2-year return period of 24-hour precipitation. The top-left panel show the stations where the return period was over 2 years (red symbols).
As regards the rain event over the UK the following animation from the cyclone database shows that there were hints in the ENS of the possibility of extreme rainfall (orange spots denote 30-50mm, see scale), associated with a cyclone, as much as 8 days in advance. Thereafter the signal decreased, but then re-asserted itself as the event approached.
Some of these frames, eg DTs 00UTC Tue to 00UTC Wed, also hint at two possible outcomes for the cyclone near the S of the UK, one near N France associated with less rain, and one near the UK associated with large rainfall.
This impression is borne out if one examines the cyclone-specific tracks and plumes, from around the first time that the cyclone was an identifiable feature in the analysis:
On both these animations one sees a clear bimodal distribution in the cyclone tracks: some deep systems (the track to the N) and some weaker systems (the track to the S). Note also that some of the southern track systems disappear from the plots as full lysis takes place. Also note how the HRES in green is initially in the weaker group, but then jumps to the stronger group close to the event. This is thus a nice example of how users can benefit from making more use of the ENS, even at day 2-3. There is probably a dynamical explanation for the bimodal tracks, perhaps related to jet crossing, though I haven't looked at that.
3.4 Monthly forecasts
3.5 Comparison with other centres
4. Experience from general performance/other cases
5. Good and bad aspects of the forecasts for the event
6. Additional material
- Report from Christian Grams, ETH (ask Linus for access)