Striking COSMO-E signal for 70 hour forecast

COSMO-E predicts a very high probability for extreme precip amounts in the Valais and Ticino area of southern Switzerland for Thursday late evening.

20160614 COSMO E 72_maxprecip12h

Maximum projection of all ensemble members showing maximum 12 h precip sums of locally more than 200 mm.

20160614 COSMO E 72_fracprecip12h

Point by point probability for rain sums larger than 70 mm: around 80 % in some places.

20160614 COSMO E 72_upsprecip12h

Fraction of members exceeding the threshold of 70 mm in a 40 by 40 km square. Nearly all members for this timestep predict excessive rainfall in south-central Switzerland.

Share

End of the first successful testbed week

On the last day of the testbed we again started with the verification of the last day, followed by an outlook for Day1 and Day2.

The participants of the group that forecasted the Day1 outlook decided two choose the western domain at least for two main reasons. A sharp short wave trough passing France during the day looks by far more dynamic than the storms that will for sure fire up over parts of Italy and the Balkan Peninsula. And second: The European Championship will start this evening in Paris.

Bild19

Downstream of an amplified short wave trough, southwesterly flow led to an advection of warm airmasses into France. Initially a lack of low level humidity was found due to the foehn winds from the Pyrenees. But corresponding to the short wave trough also a maximum of humidity is advected into Southern and Southwestern France later on. This will partly overlap with higher lapse rates downstream. That raises the possibility for at least a few hundred J/kg of CAPE. Lift is provided both by the approaching short wave trough and a well-defined surface convergence zone, respectively.

Day1-Outlook_4

What brought us to the decision of issuing a LVL1 for the area of interest was the overlap of a high likeliness of storms and DLS values of 15 to 20 m/s.

During the night this risk diminishes due to the diurnal cycle as well as a reduction of the amplitude of the trough not staying as pronounced as before.

The second group decided to choose the eastern domain which already is and will be influenced also on Day2 by a strong trough with a corresponding surface low. The highest values of BL moisture can be found in a narrow band along the cold front. Having lapse rates between 6 to 6.5 K/km the combination with humidity is forecasted to result in a few hundreds of J/kg of MLCAPE. At the same time DLS is predicted to be very high with up to 30 m/s. However, initiation is a bit questionable but wherever storms may fire up they have an enhanced threat for severe weather.

Bild22

Day2-Outlook_4Bild21

After forecasting and the online presentation of the results the testbed groups had another evaluation of the products. As an example we were looking on a case of 27th of May concerning storms over Southern Germany at 16 UTC. It is visible that several products really gave a good hint for severe weather that finally was reported with the large hail events that are visible in the images.

RotationTrack_160527_16 VILTrack_160527_16 Radar+Hail_160527_1600

It is Friday and so the first testbed week ends. All participants were leaving us happy and satisfied and we had a great time with really enthusiastic participants. Having a look to the upcoming testbed week it seems that a high-end weather situation will be present especially during the second half of the week. So… really looking forward.

Bild20

TEstbedWeek1

Share

Area of interest is shifting to the south

It is the third day of forecasting and the area of interest has shifted from Central and Western Europe to Italy and the Balkan area. Therefore the participants concentrated their Day1 and Day2 outlook to the southern domain.

Bild14

Bild15

Concerning Day 1:Short wave features influence most parts of the southern domain which are one of the trigger mechanism for today. Quite high values of moisture could be find especially near the coastlines of Italy and Croatia as well as Greece, whereas lapse rates were best more inland. Looking on the streamlines and 10 m convergence several regions could be find where onshore flow and converging streamlines should enhance the threat for excessive precipitation. Local models have shown quite high amounts of rain. Having only low values of DLS not much organization is expected, so hail and wind should not be that pronounced. The overall threat moves from Italy during the day until Croatia and Greece during the evening and night. The strongest signals for excessive precipitation due to convergent onshore flow and a visible mid level low were encircled with a LVL2 area.

Day1-Outlook_3

We had vivid discussions concerning the concrete placement of the high lightning probabability and LV1/LVL2. This reflects also the uncertainty that exists with the outlook.

For Day2 the area of interest is almost the same. Italy and Croatia as well as Greece are the areas of interest. Again features of the trough should be the source for lift on that day. For Adriatic west-coast conditions seem to be better (moisture, lapse rates and upslope wind) so a LVL1 was issued.

In todays online session Rich Thompson had a quite interesting presentation about storm modes.

Bild17

Bild18

Compared to yesterdays Day2 outlook this afternoon and evening seems to become much more interesting. After finishing the evaluation of lightning density and rotation tracks the teams will proceed with either Day3 to 5 outlook or Nowcasting of the area of interest.

Bild16

Day2-Outlook_3

Looking forward to an exciting afternoon session.

 

Share

Verification and Forecasting

Todays forecast session started with a verification of the Outlooks and Mesoscale Discussions of the day before. The 50% thunderstorm line of the Day1 outlook was quite well placed. It covered almost all the thunderstorm activity. Concerning the LVL2, many reports reached the ESWD database, whereas the most prominent was excessive precipitation. This was true for central and northern parts of the LVL2 while for the southern region only a few reports arrived until now. So it may happen that it finally will not verify.

VerificationDay1

LVL1 and LVL2 should have been extended more to the  northeast to also cover  the excessive rain events in the Ruhr area. The reason that it was not done is probably that it was right at the edge of the domain and the team concentrated more on Northeastern France and Belgium.

What was interesting: One nowcasting team issued a mesoscale discussion for Northern Germany also with the probability of tornadoes It finally happened that indeed a tornado occurred that led to a case of emergency in eastern parts of Hamburg. So it was quite a good discussion.

NowcastVerific

The tornado was probably a non-supercellular one, given low wind velocities in upper levels and a strong convergence in the surface flow. Also no rotation tracks or mesocyclone detections have been found.

Situation1

It was quite interesting that NowcastMix was able to cover both, the severe rainfall event over the Ruhr area as well as the tornadic storm near Hamburg. In both cases an extreme severe weather warning was shown.

NowcastMix2NowcastMix

After the verification the two teams prepared an outlook for Day1 and Day2, respectively. Today it was much more difficult as the day before. Concerning Day1 many clouds and even precipitation was present in the morning hours. So it was quite questionable if the atmosphere may recover troughout the day. Eventually, only a LVL1 was issued. Concerning day 2 no LVL area has been issued at all, due to the many uncertainties. However it might become neccasary that one LVL1 is needed when looking on the data tomorrow … especially  concerning the Tuscany area which shows quite a good low level convergence.

Day2-Outlook_2Day1-Outlook_2

Looking forward for the afternoon session and what indeed will happen in the area of interest.

Bild11 Bild12 Bild13 Bild10

Share

Afternoon nowcasting session

We had an interesting afternoon session with much convection affecting the area that was highlighted in the Day1 outlook. Having pulsating convection, the collision of different storms as well as several convergence lines that we analysed in the surface maps, it was quite challenging to tell about the further development of the convective situation, especially over Northeastern France.

P1010283

The main focus was on excessive precipitation events. 12 UTC soundings showed almost no wind speed throughout low- and mid levels. So it was quite likely that storm movement is quite small and driven by the dynamics of the storms (cold-pool).  In addition rather humid vertical profiles with steep lapse rates were visible.

2016060712.07145.skewt.parc

While team 1 was concentrating on the Western domain where we have issued the outlook for, team 2 was looking for northern Germany were storms organized along a pronounced convergence line. Having in mind the situation yesterday and the day before, the participants decided to include tornados in their mesoscale discussion. This finally was the right decision having a first report of a tornado near Hamburg (https://www.facebook.com/Wetterspiegel.de/videos/1108717095860456/).

Tomorrow will probably be another interesting day with short wave features that introduce a synoptic-scale lift mechanism.

Share

First outlook preparation

Today we have started to prepare the first outlooks of the week. Therefore the participants were divided into two groups.

Group 1 was working on the day 1 outlook. They first started to make a surface analysis to diagnose the current distribution of low level moisture and the separation of warm and cool airmasses. In addition they were looking for convergence zones that might become interesting throughout the day. Although there have been quite a few differences the hotspots were anticipated by all surface analysis.

Bild5 Bild6

With the help of the soundings from 00 UTC it was also found out, that in the area of interest there was quite a good overlap of steep mid-level lapse rates and humidity. Whereas humidity was enhanced to the southwest, lapse rates became steeper to the northeast.

 

After that, the model output and the high-resolution ensembles have been studied to find out which regions will also stay under the influence of these good ingredients in the afternoon hours. The following map shows what the outlook at the end looked like.

Day1-Outlook

The main hazard is excessive precipitation due to the high humidity values in combination with almost non-existing storm motion. The lack of deep layer shear mostly prevents from big hail. However it cannot be excluded for local mesoscale developments. Also a local wet downburst cannot be ruled out given a somewhat larger spread in low-layer as seen in some of the 00 UTC soundings. Finally, having the almost non-existent flow in mid-level (850 to 500 hPa) and the development of local convergence zones it is also possible that we may see one or two tornados (landspouts).

 

What was in addition interesting is the approaching vorticity maximum from the west that is already visible in water vapor satellite images. This may lead to ongoing convection throughout the whole night.

Bild8

Short wave trough(s) are also interesting for our day 2 forecast.  Since these features are accompanied with synoptic scale lift mechanisms that should enhance convective activity and in addition lead to better organized convection. Especially in the Po valley were the first short wave trough arrives in the afternoon hours also quite a lot of low level moisture will be present. Also the model output does enhance convective activity in that area. So this is why a LVL2 was issued in that region. The main threat again is excessive precipitation but also large hail is possible especially in the LVL2 area.

Bild7Day2-Outlook

So … finally an interesting setup. Looking forward to the nowcasting session during the afternoon.

 

 

 

Share

An impressive COSMO DE-EPS forecast

Last Saturday unorganized convection was dominating in a low-shear environment, but a few cells became severe and some of them produced large hail according to the ESWD. One outstanding development occured in the border region of the southernmost Netherlands and neighbouring Germany.

Interestingly, COSMO DE-EPS produced nearly perfect forecasts of this development in three consecutive initialization times (see example for 06 UTC model initiation):

20160604 8_maxsdi

Supercell Detection Index (maximum SDI of all ensemble members) for 14 UTC
20160604 8_upssdi

Fraction of 20 ensemble members exceeding a threshold for the SDI: more than 90 % in the correct 40×40 km box, while the rest of the domain was correctly modelled free of such an intense development.
20160604 Radar

High resolution visible satellite and radar overlay of reflectivity and Mesocyclone Detections at 14:05 UTC.

An interesting example for the potential of high resolution ensembles that we found in the very first hours of this years’ Testbed.

Share