Status:Finalised Material from: Linus, Mohamed, Fernando, Tim
Discussed in the following Daily reports:
http://intra.ecmwf.int/daily/d/dreport/2015/01/02/sc/
http://intra.ecmwf.int/daily/d/dreport/2015/01/06/sc/
http://intra.ecmwf.int/daily/d/dreport/2015/01/07/sc/
http://intra.ecmwf.int/daily/d/dreport/2015/01/08/sc/
http://intra.ecmwf.int/daily/d/dreport/2015/01/09/sc/
http://intra.ecmwf.int/daily/d/dreport/2015/01/12/sc/
On 9-10 January 3 severe cyclones hit north-western Europe. The first (named Elon -DE, Dagmar - DK) hit Scotland early on Friday 9 January with mean wind of 46 m/s and wind gusts up to 63 m/s in Cairngorm and later hit Denmark and Germany. |
The evaluation below is both for the Friday 9 January and Saturday 10 January.
Facts from: http://artikkel.yr.no/sa-sint-var-nina-1.12143639 and http://www.smhi.se/nyhetsarkiv/egon-sasongens-forsta-ordentliga-storm-1.83408
The plots below show observed 24-hour maximum wind gusts for 9, 10 and 11 January. For the 9 January the worst affected areas were Scotland together with Denmark and Germany. On 10 January the strongest winds occurred along the Norwegian west coast. Sweden was also hit by strong wind gusts late on the 10th. On 111 January the strongest gusts were found around the Baltic sea.
The plots below show satellite images from yr.no every third hour. In the image sequence one can follow the development of the three cyclones.
The cyclones formed in connection to a strong baroclinic zone. The plots below shows the analyses of z500 and t850 every 12th hour.
The plot below show the jet steam and MSLP (in polar projection) for the same analyses as above.
The plots below show 12-hour forecasts of MSLP and precipitation.
Looking back at satellite imagery sequences for both Elon and Felix, and in both there was some evidence, from evaporating fingers of cloud head particulate, of sting jet activity, with model profiles, at least in 2-3 day forecasts, also supportive. For Elon sting jet activity and duration look to have been limited; Felix looked a bit more clearcut. The sequence below is probably after a short sting jet phase for Elon, given that the cloud head tip is already advancing around the cyclone, in 'hammerhead' format. So any strong winds associated were probably due to the 'cold jet' that typically follows.
The plots below show the 24-hour maximum wind gusts from HRES valid for 9 January.
The plots below show the 24-hour maximum wind gusts from HRES valid for 10 January.
The plots below shows the error in MSLP (shades), forecast (black contours) and analysis (green contours) for the forecast issued 8 Jan 12z. As reference, the NCEP forecast is also included, which for the Saturday storm was almost perfect (not the Friday one). The plots suggests that the ECMWF developed the cyclone somewhat too far north-east.
The plots below show the cyclone feature plots for maximum wind speed at 1 km height inside a 600 km radius for 9 January 00z.
The plots below show the cyclone feature plots for maximum wind speed at 1 km height inside a 600 km radius for 10 January 12z.
The plots below show EFI and SOT for wind gusts valid 9 January.
The plots below show EFI and SOT for wind gusts valid 10 January.
CDF for Bergen, Norway valid 10 January (observed 37.7 m/s in town)
CDF for Lysekil, Sweden valid 10 January (observed 38 m/s, Private)
CDF for Vaxjo, Sweden valid 10 January (observed 29 m/s)
The plot below shows the ensemble forecast for 48-hour maximum wind gusts for 9-10 January inside the box 55-60N, 10-15E (Kategatt). The maximum observed gust was 40 m/s (Vaderoarna, Sweden). The forecast from 4 Jan 00z to 6 Jan 12z were for a median above 35 m/s and 90th percentiel above 40 m/s. The later forecasts got wrongly confident in lower wind gusts.
The plots below shows probabilistic forecasts and verification of a binary event defined as exceeding 2-year return period for 24-hour maximum wind gusts.
First plot: 8 January 06z - 9 January 06z. Second plot: 9 January 06z - 10 January 06z. Third plot: 10 January 06z - 11 January 06z.
The plot below shows a 2-d plot of projections into regime 1 (NAO+) and regime 2 (Scandinavian blocking), where each plot represents an analysis at 00z and the last point (lower right) the the 11 January. On 9-10 January the projection was very strong on positive NAO.
The plot below shows the ensemble forecast of projection to regime 1 where the x-axis represents the initial time of the forecast. Forecast from 1 January and onwards had a strong positive NAO signal, especially from the 4 January and onwards.
The plot below shows forecasts from different centres for probability for >25 m/s 10-metre mean wind from 6-hourly data, valid 10 January. The data are from the TIGGE archive.
The plots below are the same as above but for 10 January.
The figure below shows the scatter plot of observations and short forecasts of 24-hour maximum wind gusts for 9 January (blue), 10 January (red) and 11 January (green). Tim found that there is a maximum in the SYNOP code (99 knot) for wind gusts, which seems to have been hit by 6 stations during this event.