Ongoing analysis Material from: Linus, Lorenzo
Discussed in the following Daily reports:
1. Impact
During the last weekend of January massive precipitation hit southern Europe. Several cities in northern Italy were flooded (e.g. Pisa) and Croatia and Serbia experienced severe snowfall.
http://www.bbc.co.uk/news/world-europe-25992800
2. Description of the event
z500 and t850 from 12-hour forecasts |
|---|
| There are no images attached to this page. |
MSLP and percipitation from 12-hour forecasts |
|---|
| There are no images attached to this page. |
3. Predictability
3.1 Data assimilation
3.2 HRES
Precipitation verification and HRES forecast |
|---|
| There are no images attached to this page. |
3.3 ENS
EFI from 29 Jan 00UTC for 31 Jan to 5 Feb (day 1-5). We should produce the same to for the day before.
3.4 Monthly forecasts
Weekly precipitation anomalies |
|---|
| There are no images attached to this page. |
3.5 Comparison with other centres
3.6 EFAS flood forecasts
See below the forecasts of EPIC, the flash flood component of EFAS, issued on the 29, 30 and 31 January 2014 at 00 UTC. A rather large area in the Central and North of Italy showed a significant probability to exceed the three warning threshold, corresponding to 2, 5, and 20 years.
The forecast on the 31st also detected the severe event that caused widespread flooding in Rome, the Lazio and the Calabria region on the 2-3 February. (http://floodlist.com/europe/deadly-floods-italy)
For Italy, an EFAS partner exists only for the Po River Basin but not for other river basins. For this series of events, three EFAS flash flood watches were sent, one for the Oglio River on the 30/01/2014 (Northern Italy) and two for the Drava River (Southern Austria), on the 31/01/2014. See image below.
Discharge observations are available at some stations in the North of Italy. The comparison in the following figures show that forecasts in the Po River slightly underestimated the observed peak discharge (in red), though no flooding was caused by the river.
On the other hand, results show a clear underestimation in smaller river basins, partly due to model resolution issues. Indeed for EFAS hydrological forecasts, no flood alert is issued in river basins with upstream area lower than 4000 km2. See for example a comparison between EFAS forecasts and discharge observations in the Secchia River (1300 km2), where severe river flooding occurred.
4. Experience from general performance/other cases
5. Good and bad aspects of the forecasts for the event
6. Additional material