...
Status:
...
Finalised Material from: Tim, Mohamed, Linus
...
Discussed in the following Daily reports:
| Show If | ||
|---|---|---|
| ||
http://intra.ecmwf.int/daily/d/dreport/2015/12/21/sc/ http://intra.ecmwf.int/daily/d/dreport/2015/12/22/sc/ |
...
1. Impact
During the Christmas week the northern parts of the British Isles were hit by floods. The worst affected areas were in N England, from Lancashire across to Yorkshire, with the cities of Manchester, Leeds and York all featuring in reports, as well as southern Scotland.
http://www.bbc.co.uk/news/uk-35193682
2. Description of the event
...
The deepening of low B must have been associated with broadscale dynamical uplift, and that broadscale ascent, together with the close proximity of the frontal zone, the slow moving nature to the front due to multiple lows/wave moving along it, the warm airmass source region near Morocco, and some orographic enhancement seem to have all contrived, together, to deliver the high rainfall totals observed.
3. Predictability
3.1 Data assimilation
The animation below shows the observation rejection (low weight) charts for subsequent assimilation cycles. Wind observations with low weight is marked with symbols and the 250 hPa wind speed in the analysis is plotted with shades. The position of low A is marked with a black dot. On 23-24 December a cluster of observations were given low weight over U.S, probably connected to severe convection.
3.2 HRES
The plots below show the accumulated precipitation between 25 Dec 06z and 26 Dec 06z and MSLP valid 12z on 25 Dec. The forecasts from 24 December 12z and earlier the position of low A is too far too the west and the band of precipitation is shifted somewhat to the north.
...
The plots below show the accumulated precipitation between 26 Dec 06z and 27 Dec 06z. In the forecast from 24 December 12z and earlier the handling of low is poor.
O-suite:
E-suite:
3.3 ENS
The following animation shows the 3-day ppn total EFI, for Fri/Sat/Sun, at reducing lead times (we do not compute a 2-day rainfall EFI).
The next plot shows 24h Ppn CDFs for a location in central N England.
There is a modest EFI signal early on, which becomes weaker, before suddenly reasserting itself just before the event. This would have hindered rather than helped forecasters, particularly in the medium range where we are relied upon most for guidance. This impression is echoed somewhat by the CDF plot, which jumps to higher totals in the penultimate (Fri 12Z = purple) forecast.
The plot below shows Dalmatian charts from successive forecasts for VT 12Z Sat (26 Dec), showing cyclonic features, coloured in by MSLP. We see a clear ensemble jump between the 12Z Thu and 00Z Fri forecasts. Prior to the jump the vast majority of forecasts continued to deepen low A, and took this past NW Scotland, an evolution that would have markedly reduced the duration of rainfall over N England (even though intensity might have been the same, or for a short period higher). In these forecast low B didn't exist or was very muted. In the later batch of forecasts low B deepens, and takes over, and low A is sumsubmed by this development and loses its identity.
3.4 Monthly forecasts
The plots below show the weekly anomalies of precipitation and MSLP valid for the week 21-27 December. For all lead times a positive NAO pattern is visible in the MSLP and wet anomalies for north-western Europe.
(Plots have been lost...)
The plots below shows ensemble mean seasonal forecasts anomalies for December. The plots include precipitation (shade) and surface wind anomaly (arrows). The forecasts are from 1 Dec, 1 Nov, 1 Oct, 1 Sept and 1 Aug. All forecasts shows sign of wet anomalies for north-western Europe and stronger-than-normal westerly flow (positive NAO).
...
- The flooding earlier in December - 201512 - Rainfall - British Isles - Part 1
5. Good and bad aspects of the forecasts for the event
...