Discussed in the following Daily reports:
Material from: Esti, Linus
Discussed in the following Daily reports: |
Storm Floris was a rare and unusually severe Atlantic storm for the season, affecting mainly Scotland, northern England, and parts of Norway. The storm brought exceptionally heavy rainfall, with averages of 20–40 mm and isolated parts of Scotland reaching up to 80 mm in a single event (reported in the media but not confirmed in our observations), along with powerful winds across Scotland and Norway. According to Met Office data, a gust of 59.9 m/s was recorded at the summit of Cairn Gorm in the Highlands on 4 August, at 1,245 m above sea level. The strongest gust outside mountainous areas was 36.7 m/s at South Uist (also not registered in the observations we received). In Norway, western coastal regions experienced gusts exceeding 20 m/s in exposed fjord areas, leading to weather warnings.
The most severe impacts were in Scotland: more than 50,000 homes lost power (around 22,000 still without electricity days later), there were 119 rail incidents (75 caused by fallen trees), and widespread cancellations affected trains, ferries, and flights (including an easyJet flight that had to turn back mid‑air). Authorities issued amber wind warnings signaling a “danger to life,” urging residents to remain indoors and secure loose items. Overall, Floris was described as “the most damaging summer storm in recent memory” in Scotland, where the combination of record gusts, heavy rain, and peak tourist season amplified the risk to life and infrastructure.
Here we will mainly focus on the 10-metre mean wind at 12UTC on 4 August in boxes centred on Southern Uist (57.4N, 7.4W)
The plots below show analyses of MSLP and 6 hour rainfall from 2 August 00UTC to 6 August 00UTC, every 12th hour.
The plots below show analyses of z500 and T850 from 1 August 00UTC to 6 August 00UTC, every 24th hour.
Observations and analysis for the event
Control forecast (IFS 9-km resolution)
DestinE (IFS 4.4km resolution)
AIFS-single (AIFSv1.0 ~0.25 resolution)
AIFS-CRPS ensemble (~0.25 degree resolution)
Below 24-h precipitation accumulated in the British Islands on the 4th August 2025 at different lead times with the ENS control (oper) and DestinE. 36-h maximum wind gusts in the operational 9 km forecast and DestinE at different lead times.
Below visible satellite simulated image from IFS oper and DestinE 4 km from forecast run 4 and 5 August 2025 00 UTC and valid at T+12h. Eumetsat 12 sat image valid at the same time.
EFI (based on IFS 9-km ensemble)
The plots below show EFI and SOT for 1-day maximum wind gusts on 4 August. The forecast from 2 August and earlier had the system shifted to the south.
Forecast Evolution plot
The plot below shows the forecast evolution for the mean wind in a 0.5x0.5 degree box centered on Southern Uist (see box above) valid 12UTC on 4 August. AIFS-ENS captured better the wind speed than IFS-ENS for this location in the forecast from 31 July and 1 August.
Legend:
Observation - green hourglass
Analysis - green dot
ENS CF - red dot
DestinE - dark-green dot
AIFS-single - cyan dot
ENS distribution - blue
AIFS-ENS distribution - pink
ENS m-climate - cyan
ENS m-climate maximum - black triangle
The plot below shows the forecast evolution for the maximum mean wind in a 5x5 degree box centered on Southern Uist (see plot below for box) valid 12UTC on 4 August. Here we see much higher maximum wind in the short-range forecasts from IFS compared to both AIFS and AIFS-ENS. Some members in IFS-ENS has very extreme winds. This should be compared to IFS cycle 50r1 once e-suite is available.
Below we have the ensemble stamps for MSLP for ENS control and ensemble members. Valid at 4 August 2025 12 UTC and T+36, T+84h and T+132, for ENS control and IFS ensemble members and AIFS single and AIFS-CRPS
MSLP errors (FC-AN) on the 4 August 2025 00 UTC forecast run valid at T+36h, T+84h and T+132h for various forecast models. 
500 hPa geopotential errors (FC-AN) on the 4 August 2025 00 UTC forecast run valid at T+72h and T+120h for various forecast models.
