Status: Ongoing analysis Material from: Linus, Tim H., Bojan
1. Impact
Flooding affected a number of regions in Greece around 14 October. . The synoptic chart in the section below suggest Greece lay in a warm, moist, convective flow from the south, on the eastern flank of a Mediterranean low. Frontal rain may have been affecting the north at the time. Available observations show very large totals on Corfu and in Athens (the right marked observation is on the northern side of the Athens metropolis).
For more info see https://www.keeptalkinggreece.com/2021/10/15/corfu-destruction-stormy-weather/ and https://www.dw.com/en/flooding-brings-greece-to-a-standstill-forces-evacuations/a-59514192.
2. Description of the event
The plot below shows the synoptic chart for 12UTC on 14 October from Met Office.
The plots below show analyses of z500 (contour) and T850 (shade) from 11 October 00UTC to 15 October 00UTC, every 24 hour.
The plots below show analyses of MSLP and 6-hour forecasts of precipitation from 13 October 00UTC to 15 October 00UTC, every 12 hour.
3. Predictability
3.1 Data assimilation
3.2 HRES
The plots below show observations from SYNOP (first plot) and HRES forecasts of 24-hour precipitation valid 14 October 00UTC - 15 October 00UTC, from different initial dates. The black box marks Corfu, Greece.
3.3 ENS
The plots below show EFI for 1-day precipitation 14 October, from different initial times. Corfu is located on the north-western flank of the wet region. For longer lead times, the EFI signal was shifted to the east for the 14 October, and missed Corfu.
The plot below shows the forecast evolution plot for 24-hour precipitation valid 14 October over Curfu (0.5x0.5 degree box). HRES –red, ENS CF – purple, ENS blue box-and-whisker, Model climate – red box-and-whisker. Triangle marks the maximum in the model climate based on 1200 forecasts, and the green hourglass the mean of the observations..
The plot below shows the same as above but for Athens.
The plots below show EFI for 3-day precipitation 13-15 October, from different initial times. Also allowing for a longer time window, Corfu was still on the edge of the region with an early signal.
The plots below show forecasts of dominating flow pattern over the Mediterranean, based on Mastrantonas et al. (2021, https://doi.org/10.1002/joc.6985). Also early forecasts (from 4 October and onwards), predicted a cyclic pattern over central-eastern Mediterranean (Sicilian low, Balkan Low or Black Sea Low).
3.4 Monthly forecasts
The plots below show ensemble mean anomaly for precipitation on 11-17 October in extended-range forecasts. The forecasts started to pick up a wet anomaly on central-eastern Mediterranean from 4 October.
3.5 Comparison with other centres
24-hour precipitation from 14 October 00UTC for models participating in SEE-MHEWS-A. IFS (1st), Aladin (2nd), ICON (3rd) and NMMB (4th). All models are initialised from ECMWF analysis.
4. Experience from general performance/other cases
5. Good and bad aspects of the forecasts for the event
- Most rainfall over sea in short-range forecasts
- Medium-range forecasts ok for main parts of Greece, but missed Corfu if focusing on 14 October.
- Signal started to apperar ~10 days before the event.