How well did MeteoGroup forecast the severe storm XAVIER? The cyclone hit northern Germany with gusts ranging up to 130 km/h wind speed. We look into the performance of our weather warnings, issued to the public through www.unwetterzentrale.de and our app AlertsPro.
October 05th 2017, the severe storm XAVIER – pressure systems in Germany are named by the Freie Universität Berlin – crossed northern Germany. In a wide corridor from the German Bight to the Polish border, XAVIER’s gales reached speeds of more than 100 km/h. These severe gusts caused heavy damage, public traffic broke down and seven people were fatally injured by falling trees or branches.
From a meteorological perspective, XAVIER was a special cyclone. Its dynamics developed not like typical low-pressure systems in the mid-latitudes1, but followed the so-called Shapiro-Keyser cyclogenesis scheme2 3. Such vortexes move fastly in an overall strong atmospheric jet and involve damaging winds southwest of their centre. In case you are interested in atmospherical dynamics, I recommend having a look into the course at http://www.eumetrain.org/resources/sting_jets_2012.html.
MeteoGroup's forecasts and alerts announcing XAVIER
On Tuesday morning (October 03rd), the first weather forecasting models indicated a chance of a storm on Thursday, and our MeteoGroup forecasters started to communicate the upcoming storm risk in Germany. With each new model's results arriving, the team of www.unwetterzentrale.de sharpened the forecast and at noon the upcoming risk of gales exceeding 100 km/h was first announced.
On Wednesday morning (October 04th), we issued pre-alerts for gale-forced winds for Thursday, expected to reach peaks up to 110 km/h, starting alongside the German Bight and moving quickly towards the Southeast, crossing Hamburg around noon, and reaching Berlin in the afternoon, see http://www.wetter24.de/news/detail/2017-10-04-sturmtief-xavier-zieht-auf/. The forecasters communicated the special risks and potential impacts of the storm.
Those who started the Thursday (October 05th) with a check of our Unwetterzentrale warnings found a red alert for big areas in Northern Germany:
The screenshot from www.unwetterzentrale.de shows storm alerts as they were in effect by Thursday at 07:00 CEST. Pre-alerts for the red area have been issued Wednesday morning, first active red alerts in the Northwest started the night before. Red alerts announce that gusts of more than 100 km/h speed are to be expected.
Did our forecasts and warnings hit the nail?
To warn as little as possible but as often and intensively as needed is a challenging tightrope walk. Intensive atmospheric phenomena like thunderstorms or a Keyser-Shapiro cyclone include sub-scale dynamics and chaotic behaviour. The development cannot be forecasted exactly, but only in terms of ‘probability’, ‘chance’ and ‘risk’. We want to warn all people which might potentially be impacted by the storm. On the other hand, when we warn too frequently of storms which end up not hitting a warned user, the awareness will dilute.
At MeteoGroup, we measure the quality of our severe weather warnings, issued by www.unwetterzentrale.de, on a daily basis. This allows us to provide immediate quality feedback to our forecasters. We can learn from failures and celebrate our forecasting successes. More information on MeteoGroup’s verification of severe weather events can be found at https://www.researchgate.net/publication/319313555_Operational_user-centric_verification_of_severe_weather_warnings.
Below, I compare the warnings in effect on Thursday 07:00 CEST (left) to the highest observed wind speed (right). The dots and flags indicate weather stations reporting gusts, either operated by the German national weather service DWD (primary and secondary network, a total of 300 stations in Germany) or operated by MeteoGroup. The MeteoGroup-owned station network includes 467 stations in Germany reporting wind gusts.
Red flags show, where on Thursday stations reported gales between 100 km/h and 130 km/h, which corresponds to our red warn level. Orange dots mark stations where gusts between 75 km/h and 100 km/h have been observed. This range corresponds with our orange level for gust warnings. Green dots mark stations which measured only windspeeds below 75 km/h.
Even a handful of violet flags can be spotted, all on very exposed stations. The leader is the Brocken mountaintop, 1134 metres above sea level, where 177 km/h have been measured.
The warnings of www.unwetterzentrale.de excellently cover the gusts as observed on Thursday. Southwest of XAVIER’s track, for a ribbon of ~100-150 km width, the wind did not reach a red level. Given the extraordinary strength of the cyclone and the involved risks we consider this over-warning as acceptable. MeteoGroup’s early announcement of the storm, the communication of uncertainty, and, finally, the alerts issued in good time to prepare proper mitigation can be considered as exemplary.
MeteoGroup’s support for insurance claims
We use the gust reports from more than 750 observing stations, operated in Germany by the DWD and by MeteoGroup itself, for a storm field analysis. The observation data, radar information and high-quality atmospheric models are combined to provide a high resolution grid or spatial shapes of the daily maximum gusts in storm fields. The results are accessible for private persons via www.meteoarchiv.de; insurance companies either use a direct access via website or ingest the data into their own systems.
For storm XAVIER the MeteoGroup storm field analysis results in this pattern:
Gusts above 100 km/h as observed on Thursday 05 October 2017
The below table lists all gusts above 100 km/h as observed by DWD stations or MeteoGroup owned stations on October 05th 2017:
|Station-ID||Max gust 05 Oct 2017||Station name||Latitude N||Longitude E||Elevation m above NN|
|6510390||136||Berlin-Wannsee (exposed location)||52.4312||13.1842||40|
See Bjerknes, J., and H. Solberg, 1921: Meteorological conditions for the formation of rain. Geofys. Publ., 2(3), 3-61. ↩
See Shapiro, M.A., and D. Keyser, 1990: Fronts, jet streams and the tropopause. Extratropical Cyclones: The Erik Palmen Memorial Volume, C. W. Newton and E. O. Holopainen, Eds., Amer. Meteor. Soc., 167-191. ↩
See Shapiro, M.A., and Coauthors, 1999: A planetary-scale to mesoscale perspective of the life cycles of extratropical cyclones: The bridge between theory and observations. The Life Cycles of Extratropical Cyclones, M. A. Shapiro and S. Gronas, Eds., Amer. Meteor. Soc., 139- 185. ↩