202209 - Rainfall - Italy

Status:Ongoing analysis Material from: Esti, Tim, Ivan

1. Impact

At least 11 people died after flash floods hit the Italian region of Marche on 15 September. Torrential rain caused rivers and streams to overflow and inundate coastal towns around the regional capital of Ancona. Around 400mm (16 inches) of rain - half a year's worth - were recorded in just a few hours. In addition to 11 dead, Italian newspapers report that around 50 people were being treated at hospitals for hypothermia and other injuries sustained in the floods. More than 180 firefighters assisted in the rescue efforts, evacuating people who overnight were forced to climb up trees or get onto their roofs to escape the rising water. Cantiano's streets turned into rivers as torrents of muddy water washed away cars and reached the first floors of houses. On Friday, many schools in Le Marche were closed.

2. Description of the event

A convective V-shape regenerating system in an incoming trough's very moist prefrontal part is probably the main cause of such an extreme event.

3. Predictability

3.1 Data assimilation

3.2 HRES

3.3 ENS

	raw ENS 6h	raw ENS 12h	ecPoint6h	ecPoint12h	COSMO raw	COSMO post	Merge	OBS (24h precipitation)
BT: 15 Sept 2022 00 UTC T+ 24h
BT: 14 Sept 2022 00 UTC T+ 48h
BT: 15 Sept 2022 00 UTC T+ 18h
BT: 15 Sept 2022 00 UTC T+ 30h

Some quick comments related to the plots below:

Observations show afternoon and even early evening temperatures which are high enough to trigger convection, based on the model soundings.

Model soundings show wetter lower tropospheric conditions in the centre and centre west of the country, and drying east of the Appenines. Note also the very high shear in the low to mid-troposphere, and steering winds of 50kts or even more.

The orography map shows that the short range HRES wettest point was in an HRES valley (not what I was expecting!) suggesting that maybe the HRES 50mm in 6h rainfall peak was due to surface-based triggering in that model run.

The model forecasts in the preceding two days were inconsistent and jumpy (for the said point at least), with switches in what was expected to be the wettest period, and also in how wet it could be - e.g. from virtually no rain predicted in HRES from DT 00Z 14th, to 50mm from a 36h later DT.

IFS soundings at T+0=12UTC 15th, from west to east coast (left to right), crossing into Marche. *Central point (42.94N,12.48E), which is near Perugia, was where the HRES max 6h precipitation T+0-6 = VT 12-18UTC 15th was located (about 50mm). See map on table below this one. Also, one sounding for T+6=18UTC.

West coast	West-Centre	Centre	East-Centre	East Coast

Meteograms for the "Central point"* (from the above table), from four consecutive data times.

Location of "Central Point"* wrt HRES orography (see pin)	00Z 14th	12Z 14th	00Z 15th	12Z 15th

3.4 Monthly forecasts

3.5 Comparison with other centres

4. Experience from general performance/other cases

5. Good and bad aspects of the forecasts for the event

Report from Italian Aeronautical Militare regarding the performance of our forecast for this event: 15 sept 2022 Marche.pdf

6. Additional material

Photos yesterday afternoon from Cantiano's streets (Italy).
Some first data obtained from the most affected region with accumulated precipitation on 15 September 2022 (but it continued for more hours).
Satellite observation of the storm in Italy.
Precipitation rate (mm/h) estimations from radar in Italy at 19:25 UTC yesterday.
500 hPa geopotential and 850 hPa temperature. Base time 15 September 00 UTC and valid at T+18h. Satellite imagery and ESWD reports in the last 3h - animation