Today (Fri 26 July 2013), the last of the four Testbed weeks has ended. The current severe weather episode over France, the BeNeLux countries and Western Germany has made it a particularly exciting and intensive week, which is also the reason why there is no time for a longer blog entry today.

However, we promise to publish a wrap-up of the meteorological developments and our forecast activities for the high-impact severe weather situations that occur(red) yesterday (Thu 25 July 2013) and probably today (Fri 26 July 2013) and tomorrow (Sat 27 July 2013) as soon as possible, hopefully in the course of next week. As a teaser: the highest threat level 3 was issued for parts of France for today and tomorrow, meaning a greater than 15% chance of extremely severe weather (hail >5 cm, wind gusts >32 m/s and/or tornadoes rated F2 or stronger) within a radius of 40 km around a given point. Stay tuned!

Meanwhile, we thank all participants for four wonderful and productive weeks! We hope all of you enjoyed the time in Wiener Neustadt as much as we, the ESSL staff, did!

The models seem determined that we are heading towards an exceptional severe weather outbreak in the Western half of Europe in the next couple of days. Forecasting an approaching high-impact situation gets more and more exciting, but also more challenging.

Fig. 1 presents the day 1 forecast issued today (Thu 25 July 2013). The group on duty was working so hard that they ran out of time to issue an explanatory text along with it. 😉 However, a brief reasoning can be presented here, along with a handful of model maps. While the zone of widespread but weakly organized thunderstorms which has concerned us this week so far is slowly leaving the Paris domain to the East (a level 1 area for pulse storms with a main threat of excessive rain over N Italy and Slovenia is its remnant), the coming period of strongly organized convection is expected to be kick-started over SW France today in the evening.

Fig. 1: Day 1 forecast issued today, Thu 25 July 2013.

With the gradual eastward shift of the upper-level jet streak ahead of the Atlantic trough (Fig. 2), steep lapse rates from Iberia and rich low-level moisture accumulation along a convergence zone / sea breeze front are expected to coincide over SW France. This combination results in the first remarkable overlap of CAPE around 2000 J/kg and deep-layer shear between 15 and 20 m/s. Figures 3 and 4 show that GFS, ECMWF and COSMO-EU agree on this scenario despite their different model resolutions and the resulting different representation of the topography.

Fig. 2: ECMWF forecast for 500 hPa geopotential (black contour lines), temperature (colour shading) and wind (barbs) for 18 UTC Thu 25 July 2013.

Fig. 3a: GFS forecast for mid-level lapse rates (red contour lines, plotted above 6.5 K/km) and low-level moisture (colour shades) for 18 UTC Thu 25 July 2013.

Fig. 3b: As in Fig. 3a, but for ECMWF.

Fig. 3c: As in Fig. 3a, but for COSMO-EU.

Fig. 4a: GFS forecast for CAPE (colour shades) and 0-6 km vertical wind shear (red contour lines) for 18 UTC Thu 25 July 2013.

Fig. 4b: as in Fig. 4a, but for ECMWF.

Fig. 4c: As in Fig. 4a, but for COSMO-EU.

The convergence of the low-level wind due to sea breezes and upslope flows is thought strong enough to spark first thunderstorms towards evening, which have good chances to turn into supercells with a threat of large hail, severe wind gusts and locally excessive precipitation. A level 2 was issued for these threats. Even an isolated tornado is not ruled out along the sea breeze front, though it was agreed that the 0-1 km wind shear is rather borderline. Later tonight, the forecast models largely agree on the evolution of an MCS which shall travel northward across France, which is why a level 1 was extended northward for severe wind gusts and excessive precipitation.

Tomorrow (Fri 26 July 2013) looks even more threatening over France, BeNelux and W Germany. The day 2 forecast for tomorrow is presented in full detail in Fig. 4.

Fig. 5: Day 2 forecast issued on Thu 25 July 2013, valid for Fri 26 July 2013.

—–

Day 2 Forecast

Valid: Fri 26 Jul 2013 06:00 to Sat 27 Jul 2013 06:00 UTC

Issued: Thu 25 Jul 2013 09:15

Forecaster: ESSL TESTBED

A level 1 and a level 2 were issued for large parts of France, Belgium, the Netherlands, Luxemburg and Germany mainly for large hail, severe wind gusts and to a lesser extent for tornadoes and excessive precipitation.

A level 1 was issued for N Italy mainly for large hail and excessive rainfall.

France, Benelux, Germany:

Warm air advection overspreads W and Central Europe from the Southwest. Steep lapse rates originating from Spain, the Pyrenees and the Alps will be advected into France and Germany. Very rich low-level moisture is expected to build up along a wavy front from W France into Benelux. This results in CAPE values around 2000 J/kg, locally even up to 3000 J/kg.

Deep-layer shear is between 20 and 25 m/s beneath the Southwesterly jet streak. Any storm that forms can quickly turn into a supercell. Later towards evening, growth into one or more large MCSes is also possible. Large hail and severe wind gusts pose the main threat. A tornado is not ruled out in areas where plentiful low-level moisture and enhanced 0-1 km shear overlap.

In the absence of synoptic-scale lift, local convergence zones in the low-level wind field act as possible triggers for convection. These include upslope winds (e.g. over the Pyrenees), sea breeze fronts (e.g. over W France) or outflow boundaries from overnight´s convection (mainly over N France, Belgium and W Germany). A level 2 area was drawn where our confidence in convective initiation is best. We left the Netherlands and NW Germany under a level 1, because residual cloudiness from overnight´s convection may limit diurnal heating.

N Italy:

Very rich low-level moisture again results in CAPE around 1000 J/kg. Vertical wind shear remains weak. Pulse storms are expected especially over the mountains with a threat of large hail and excessive rain.

—–

This afternoon, evaluation of the various tested products (NowcastMix, COSMO-DE ensemble prediction system and the Mesocyclone Detection Algorithm) is the primary task, since nowcasting is not so interesting yet. Until now (14 UTC), first weakly organized storms have only formed over parts of Germany and – completely off our radar this week – over parts of Ireland, Sweden and Russia.

Fig. 6: Nowcast display at 1400 UTC Thu 25 July 2013: E-View satellite image, SYNOP reports and lightning detections within the last 15 minutes.

Everything is still quiet over the level 2 area, in line with our expected late initiation. Satellite imagery showed low stratus until 10 or 11 UTC, which then gave way to unimpeded insolation (Fig. 6). Dew points are around 20°C and temperatures well in excess of 30°C. We will continue to monitor when (and, as pessimists would say: if) storms will explode over the Pyrenees and SW France until we head out for our weekly joint dinner at 1630 UTC. I am sure some of us will even be intrigued enough to even resume monitoring between dinner and bedtime! 🙂

With the advance of a convergence zone and the associated increase of low-level moisture in the very warm air mass, a belt with daily widespread afternoon thunderstorms is slowly shifting eastward from Western into Central Europe these days. Due to still weak vertical wind shear, most of the storms continued to be poorly organized yesterday, but nonetheless they produced another round of excessive rain, large hail and severe downburst events scattered across France. The activity also started to spread into Western Germany, and the verification of yesterday’s day 1 forecast demonstrated that much of the severe weather events occurred further East than the issued level 1 areas (Fig. 1), similar to the day before. In the afternoon nowcasting session the same team struggled with the short-lived character and elusive nature of these pulse storms (Fig. 2).

Fig. 1: Verification of the day 1 forecast issued yesterday (Tue 23 July 2013).

Fig. 2a: Verification of the nowcast issued at 13 UTC Tue 23 July 2013, valid for the 13 to 15 UTC period.

Fig. 2b: Verification of the nowcast issued at 14 UTC Tue 23 July 2013, valid for the 14 to 16 UTC period.

Fig. 2c: Verification of the nowcast issued at 15 UTC Tue 23 July 2013, valid for 15 to 17 UTC period.

Now that Germany finally starts to get affected by thunderstorms again after several exceptionally calm weeks, more of our attention starts to shift back to those of our tested products which are confined to the German area, namely the high-resolution COSMO-DE ensemble prediction system, the NowcastMix product for an automatic warning guidance and the mesocyclone detection algorithm. All of them have been developed and provided by the German Weather Service (DWD).

Ongoing complaints that the forecast models tend to mis-estimate (and, in particular, undererstimate) the pulse storm activity of the current weather period led to a closer comparison of the CAPE and precipitation forecasts today, which provides a good opportunity to highlight the COSMO-DE ensemble prediction system (in the following: COSMO-DE EPS). Its individual runs receive their initial and boundary conditions from 4 different global models (ECMWF, GFS, GME and JMA) and are computed with 5 different model physics packages, resulting in a total of 20 ensemble members. The horizontal resolution of 2.8 kilometres is impressively fine.

The following images illustrate the panel of 4 different visualizations of COSMO-DE EPS that we have made available for our Testbed, using the forecasts of radar reflectivity for 14 UTC yesterday (Tue 23 July 2013) as an example (base time: 06 UTC). Fig. 3 presents the envelope maximum simulated radar reflectivity of all model runs, which can be interpreted as a “worst case scenario”. In Fig. 4, areas with forecast radar reflectivity greater than 40 dbZ are highlighted with various colours, each of which belongs to one particular EPS run; this display is particularly neat because the user can track these colour-coded features from one forecast time step to the next. In Fig. 5, this information from Fig. 4 is condensed to the percentage of model runs which simulate a radar reflectivity greater than 40 dbZ; here, the “personal identity” of the individual EPS runs gets lost, but on the other hand you get an impression of how many EPS members exceed the threshold, which could not be seen from Fig. 4. All in all, Testbed participants can largely agree that this visualization is very useful and that the various displays ideally complement one another.

Fig. 6 shows the radar reflectivity forecast of the deterministic COSMO-DE run for the same time. Comparing it to the Figures 3 to 5, you can easily see that the ensemble runs add valuable information by forecasting a possibility of convection also further North over Western and Central Germany, where indeed lots of storms with severe wind and rain events occurred yesterday (see the verification in Fig. 1 again). They reduced the under-forecasting of thunderstorm activity, which was lately perceived by the Testbed participants, and are a helpful vehicle which shall prevent forecasters to become over-confident!

Fig. 3: COSMO-DE forecast for 14 UTC Tue 23 July 2013: maximum simulated radar reflectivity of all EPS runs.

Fig. 4: COSMO-DE forecast for 14 UTC Tue 23 July 2013: areas of simulated radar reflectivity >40 dbZ in the individual EPS runs.

Fig. 5: COSMO-DE forecast for 14 UTC Tue 23 July 2013: percentage of EPS members with a simulated radar reflectivity >40 dbZ.

Fig. 6: COSMO-DE forecast for 14 UTC Tue 23 July 2013: simulated radar reflectivity according to the deterministic run.

Since the precipitation signals were even stronger for today (Wed 24 July 2013) in the eastward shifting belt across Central Europe, today’s day 1 team even issued a level 2 area for the Southern half of Germany and a large surrounding level 1 for the rest of Germany, E France and the entire Alpine region. Despite the weak shear, the expected large number of storms was expected to result in rather many hail, wind and rain reports. This slightly controversial decision along with a hint of actionism by Christoph, wo actively involved the other group participants without prior announcement, resulted in a particularly lively online weather briefing today.

Since many of the storms across Germany formed quite early in the day and since Thilo was particularly quick with retrieving severe weather reports, a preliminary verification of today’s forecast can already be done: a few handfuls of heavy rain, severe wind and marginally large hail events occurred mainly over Germany, but probably not enough for a level 2. The forecast models have finally stopped their recent under-forecasting. Darned if you follow them, darned if you don’t!

Yesterday’s pulse storms (see yesterday’s post) brought quite a lot of severe weather again. Excessive rain and localized severe wind gusts were scattered across France, Belgium, Southern Switzerland (Ticino) and Northern Italy (Lombardia and Alto Adige) in the afternoon and evening. The most stunning event was a fierce hailstorm between Bolzano and Trento in the Italian Alps: hailstones up to 4 cm and hurricane-force gusts managed to destroy 30.000 tons (!) of apples within minutes.

Turning to the second part of yesterday’s ouverture into a possibly memorable week, the hottest day since 2006 in England with temperatures up to 34°C was also followed by a night of widespread thunderstorms, as a short-wave trough provided some lift. These storms were accompanied by localized flooding and an amazing number of damaging lightning strokes, while hail and wind only played a minor role (non-severe hail in London and two cases of fallen trees near Stoke-on-Trent are known so far).

Fig. 1 presents the verification of yesterday’s day 1 forecast and the accompanying text. In general, the lightning areas were aligned pretty well with the detected lightning strokes (magenta dots), but the severe weather events occurred clearly more to the Northeast than expected (Belgium, N France, N Italy), which mainly resulted from an under-estimation of the severe weather potential associated with the pulse storms. The Western level 1, which was mainly issued due to the gradually increasing forcing and shear, verified better.

Fig. 1: Verification of the day 1 forecast valid for Mon 22 July 2013.

—–

Day 1 Forecast

Valid: Mon 22 Jul 2013 15:00 to Tue 23 Jul 2013 06:00 UTC

Issued: Mon 22 Jul 2013 14:55

Forecaster: ESSL TESTBED

A level 1 was issued for western France, northern Spain and southern England, Wales and surroundings for excessive rain, large hail and severe wind gusts.

Rich low-level moisture is overlapping with steep lapse rates from the east. CAPE will be around 1000 J/kg in the evening hours across northern France. A mid-level vort-max will move northward along the western French border spreading into the British Isles. Initiation of storm clusters is expected across France and England that move north/north-eastward. Main threat will be excessive precipitation due to rich moisture and due to storm movement parallel to the boundary.

Across France/Spain, larger CAPE enhances the potential of a few hail events. Additionally, vertical wind shear will be around 15 m/s, and a few organized storms are not ruled out. Bowing segments may than pose a threat of severe wind gusts.

Western Alps

Weak winds, high moisture and steep lapse rates are present, convergence in the complex terrain has lead to initiation. Pulse storm and stoirm cluster will be capable of producing isolated large hail or doenbursts. Main threat will be excessive rain due to the slow storm movement.

—–

Not so much change is expected for today (Tue 23 July 2013) yet, hence today’s day 1 forecast is looking pretty similar. The Western level 1 belt was shifted a bit to the East and the Alpine storms were left without a second level 1 area this time.

We are already having one eye at the development of the possible severe weather outbreak over Western Europe towards the end of the week, when GFS and ECMWF agree on an almost incredible weather pattern with a deep and strong SW flow that shall feature abundant CAPE and strong shear. We won’t see forecast maps like these too often in our lives. Fig. 2 shows a handful of them if you promise not to drool!

Fig. 2a: Today’s ECMWF forecast of 500 hPa geopotential (black contours), temperature (colour shades) and wind (barbs) for 18 UTC Sat 27 July 2013.

Fig. 2b: Today’s ECMWF forecast of mid-level lapse rates (red contour lines, plotted above 6.5 K/km) and low-level moisture (colour shades) for 18 UTC Sat 27 July 2013 (Paris domain).

Fig. 2c: Today’s ECMWF forecast of CAPE (colour shades) and deep-layer shear (contour lines) for 18 UTC Sat 27 July 2013 (Paris domain).

Today’s day 5 forecast is already covering Saturday, and it is not surprising that the team on duty used a strong wording – see Fig. 3 and text below.

Fig. 3: Day 5 forecast issued today, valid for Sat 27 July 2013.

—–

Day 5 Forecast

Valid: Sat 27 Jul 2013 06:00 to Sun 28 Jul 2013 06:00 UTC

Issued: Tue 23 Jul 2013 15:11

Forecaster: ESSL TESTBED

The severe weather situation of the previous day intensifies.

Severe weather is expected in France, the Benelux states and the northern part of Germany. With CAPE values up to 3500 J/kg and coincident 0-6 km wind shear of more than 20 m/s, well organized, persistent storm systems producing lightning, heavy rain, large hail, and strong windgusts can be expected.

In northern Italy and the south of Switzerland considerably high CAPE up to 3000 J/kg with weak wind and moderate 0-6 km wind shear less than 10 m/s can lead to slow moving convective cells with heavy rain, lightning, and possibly hail.

—–

Only fly in the ointment: it seems like the biggest storms day(s) will follow just after the end of the last Testbed week.

The 4th and last Testbed week 2013 has begun, and 13 new participants from Poland, the Czech Republic, Germany, Austria, Portugal, Finland, the USA and Hongkong (!) have arrived. Furthermore, also the usual ESSL team is complemented by Kathrin (our Deputy Director) and by Thilo (our ESWD Quality Control Manager), which means that the activity in our building is more bustling than ever before.

The weather seems to join in and there are convincing signs that it may approach top form by the end of the week. Fig. 1 presents the ECMWF 500 hPa forecasts in 24-hour time steps for today till Friday (26 July 2013) this week.

Fig. 1a: ECMWF forecast of 500 hPa geopotential (black contour lines), temperature (color shades) and wind (barbs) for 12 UTC Mon 22 July 2013.

Fig. 1b: As in Fig. 1a, but for 12 UTC Tue 23 July 2013.

Fig. 1c: As in Fig. 1a, but for 12 UTC Wed 24 July 2013.

Fig. 1d: As in Fig. 1a, but for 12 UTC Thu 25 July 2013.

Fig. 1e: As in Fig. 1a, but for 12 UTC Fri 26 July 2013.

The main feature this week is a broad upper-level ridge, which will very slowly move eastward and push an old cut-off low further into Russia. From the Atlantic, a new pronounced upper-level low will gradually come closer to Western Europe. Ahead of it, the Southwesterly flow is forecast to steadily increase during the week and advect hot air from Morocco and Iberia deeply into Central Europe, where temperatures within reach of record-breaking values can be expected by next weekend. Closer to the frontal zone, an extraordinary overlap of CAPE and shear may evolve from Spain via France to the BeNeLux countries and possibly into parts of Germany. This already smells like an outbreak of severe weather!

Until then we have to still lower our sights a little bit, but the week is already starting more exciting than the previous ones. Today’s thunderstorm activity was expected to be centered over two areas, the first one over the Western Alps and the second one in a belt from Western France into England. In the absence of synoptic-scale forcing (apart from the British Isles, which see some support from a first Atlantic trough overnight), the main trigger mechanism is provided on local scales by upvalley/upslope winds and by sea breeze fronts, respectively. The 12 UTC soundings from Payerne (CH), Milano (IT) and Bordeaux (FR) in Fig. 2 exemplarily outline today’s environment in the Western half of Europe: rich low-level moisture and fairly steep lapse rates create locally quite high CAPE with an conquerably small cap, but vertical wind shear is weak to non-existent.

Fig. 2a.

Fig. 2b.

Fig. 2c.

The afternoon turned quite interesting, when a lot of pulse storms formed across France, Belgium, Luxemburg and the Western Alps – in fact more than expected also over hilly or even flat terrain. The rapid scan satellite images revealed that some of the convection was obviously triggered secondarily by outflow boundaries or fast-moving gravity waves, which made it particularly difficult to predict the next instances of convective initiation.

Fig. 3a: Meteosat E-View satellite image at 1330 UTC Mon 22 July 2013.

Fig. 3b: Meteosat E-View satellite image at 1430 UTC Mon 22 July 2013.

Fig. 3c: Meteosat E-View satellite image at 1530 UTC Mon 22 July 2013.

Fig. 3d: Meteosat E-View satellite image at 1630 UTC Mon 22 July 2013.

Since we don’t have anything to verify yet on the first day of a new Testbed week, this posting finally has to turn into some more or less random “image-dropping” again. Fig. 3 shows an hourly sequence of satellite images of the French storms at 1330, 1430, 1530 and 1630 UTC. Enjoy the particularly beautiful initiation of still indecisive storms along the sea breeze front over Western France and an isolated, powerful storm over the French North coast. The latter dumped a cloudburst over Caen, which resulted in a severe flash flood and a few broken trees. This was today’s first severe weather event we learned about, but certainly not the last one. Stay tuned for tomorrow’s verification!

Today (Fri 19 July 2013) well-organized thunderstorms occurred on the one hand in the Baltic States, Belarus and the Ukraine, and on the other hand again over parts of Spain and France. Not only the geographic location of these two areas, but also the character of the storms was diametrically opposed to each other.

Fig. 1 briefly presents today’s large-scale weather pattern in form of the 500 hPa ECMWF forecast for 12 UTC.

Fig. 1: ECMWF forecast of 500 hPa geopotential (black contour lines), temperature (color shades) and wind (barbs) for 12 UTC Fri 19 July 2013.

The Spanish and French convection again formed in the vicinity of the shallow but everlasting upper-level low, which provided the familiar environment of moderate to high instability and rather weak (albeit locally enhanced by sea breezes) vertical wind shear. Fig. 2 shows another display of well-organized, powerful storms which have formed over Northern Spain the 11th day in a row.

Fig. 2: Nowcast display at 1445 UTC Fri 22 July 2013: Meteosat E-view satellite image lightning detections within the last 15 minutes.

However, the arguably more interesting, because new, convection was the one over Northeastern Europe. As shall be shown below, this new environment of strong shear, strong forcing and limited CAPE provided bigger challenges for the Testbed participants than the x-th round of Spanish storms. Fig. 3 and 4 present the sequence of forecasts targeted on today.

Fig. 3a: Verification of the day 5 forecast issued on Mon 15 July 2013, valid for Fri 19 July 2013.

Fig. 3b: Verification of the day 4 forecast issued on Tue 16 July 2013, valid for Fri 19 July 2013.

Fig. 3c: Verification of the day 3 forecast issued on Wed 17 July 2013, valid for Fri 19 July 2013.

On Mon 15 July 2013, no threat area was issued yet for today (Fig. 3a). The forecast uncertainty was still too big, as it was also noted in Monday’s blog entry here: “Towards the end of the week, it looks like the active frontal zone over the far North will run another attempt to erupt towards Eastern Europe, but by then discrepancies between GFS (not shown) and ECMWF become quite high and a concurrence with unstably stratified air is still doubtful.”

The first threat area for the Baltic States for today was issued on Tue 16 July 2013 (Fig. 3b), along with the comment: “(…) Low probablilty of severe weather (cold front) – forecasters should keep an eye at this region.”

On Wed 17 July 2013, the issued threat area was quite similar (Fig. 3c) but the accompanying text became more assuring: ” (…) A strong PV perturbation moving very quickly from North to South (ECMWF) and NW to SE (GFS) will reach that area around the middle of the day. Large values of 0=6 Km wind shear (up to 20 m/s) are expected over the same area. CAPE values aorund 800 J/Kg over Lithuania and Latvia (ECMWF) or 1200 J/Kg over Lithuania and Western Russia (GFS).”

Fig. 4a: Verification of the day 2 forecast issued on Thu 18 July 2013, valid for Fri 19 July 2013.

Fig. 4b: Verification of the day 1 forecast issued on Fri 19 July 2013.

For the day 2 and day 1 forecasts targeted on today (Fri 19 July 2013), respectively, the Warsaw subdomain was chosen. The day 2 team on Thursday was confident enough to issue a (low-end) level 2 area centered over Belarus (Fig. 4a) along with the following text:

“A strong through is approaching from Scandinavia. At the eastern side of the through warm and moist air with steepening laps rates will be advected. This will lead to some CAPE around 800 J/kg. During the day and in the night a well definded cold front will cross the northeastern parts of the forecast area at the left exit reagion of a strong upper level jet. With the approaching cold front DLS will increase to values around 15 – 20 m/s. At the coldfront modells predict wide spread convection which will likely developing into a line because of the good environmental shear. At the line widespread severe wind gust and isolated large hail are possible. At evening hours LLS will also increas to values around 10 m/s. The stronger LLS and low LCL will support the chance for tornadoes in some embedded supercells, mostly in the level 2 area. Therefore a low end level 2 was issued.

A head of the coldfront some well organized multicells and some supercells could develop in a weak convergent flow. The main threat would be large hail and isolated downbursts.”

While the confidence in severe thunderstorms had increased from day to day so far, today’s (Fri 19 July 2013) day 1 team performed a break and merely issued a level 1 for Belarus (Fig. 4b) and the following text:

“High probability of thunderstorms during the afternoon and early evening. Strong synoptic forcing is expected over the area with values of CAPE around 500 J/Kg but wind shear around 15-20 m/s in 0-6 Km, 10-15 in 0-3 Km and around 10 m/s in 0-1 Km. Toghether with the arrival of the cold front to the area a low level convergence line  will appear.”

Scanning the sequence of reasoning issued along with the forecasts gives the impression that little doubt was left about strong vertical wind shear and strong forcing, but that the amount of instability (expressed in terms of CAPE) became doubtful the closer the event came. In particular, low- to mid-level lapse rates did not seem to be too steep, which did not allow higher CAPE formation. The other “ingredient” for CAPE, enough low-level moisture, seemed to have been in place.

Another reason for the lower confidence in severe weather on day 1 was that the forecast models kept accelerating the cold front. Indeed, a convective line formed already early in the day across Latvia and Lithuania, which was accompanied by a pronounced wind shift (Fig. 5). This rapid development even made us issue an unscheduled nowcast with a wind watch valid from 09 till 11 UTC for areas ahead of the convective line (not shown).

Fig. 5: Nowcast display at 0900 UTC Fri 19 July 2013: Meteosat E-View satellite image, SYNOP reports and lightning detections within the last 15 minutes.

As the line continued to move southeastward into Belarus in the afternoon, it kept its sharpness but the lightning activity got increasingly isolated. Indeed it seems like a lack of instability and the resulting too low cloud tops were the limiting factor here for a better shear consumption.

Till the time of writing, no severe weather reports arrived. The SYNOP gust reports were mostly in the 12 to 15 m/s range, hence an isolated severe gust (>25 m/s) could not be ruled out but should be considered quite unlikely.

While it was therefore a good choice for today’s day 1 team to only issue a level 1 instead of a level 2, the lightning verification showed that quite a lot of storms occurred outside the drawn areas (Fig. 4b). The team on duty seemed to focus too strongly on the main area of interest, something that even experienced forecasters are not immune against. Or was the approaching weekend already too close? 😉

We wish the participants of the third Testbed week a safe trip home and the participants of the fourth and last week a safe journey to Wiener Neustadt!

Today even two day 1 teams worked parallel on forecasts for the Toulouse domain, where another round of pulse storms over Southern France and the Western Alps and better organized storms over Northern Spain are expected, and for the Helsinki domain, where the beginning southward shift of the main frontal zone finally starts to create a more dynamic situation (Fig. 1). These two day 1 forecast are presented below. Enjoy!

Fig. 1: ECMWF forecast of 500 hPa geopotential (black contour lines), temperature (color shading) and wind (barbs) for 12 UTC Thu 18 July 2013.

Fig. 2: Day 1 forecast valid for Thu 18 July 2013 for the Toulouse domain.

—–

Day 1 Forecast

Valid: Thu 18 Jul 2013 08:00 to Fri 19 Jul 2013 06:00 UTC

Issued: Thu 18 Jul 2013 09:31

Forecaster: ESSL TESTBED

A level 2 was issued for northeastern Spain mainly for large hail and severe wind gusts.

A level 1 was issued for southeastern France mainly for local excessive convective precipitation.

…other comments…

15 m/s 0-3 km shear is presented by most models over the northeastern edge of the Spanish Plateau. High dewpoints are returning into the Ebro Valley. Together with weak capping and deep convergence storms will be triggered and likely transform into supercell storms with a chance of large hail. High LCL heights > 1700m enhance both hail and severe wind gust threat. The coastal region will remain capped, tough.

France is under very weak shear conditions, limiting the organization and severe weather potential, but slow storm motion can lead to locally high precipitation accumulations.

Western France has high dewpoints and convergence and may trigger early storms.

The IPV feature moving into SW France at night seems not capable of triggering storms, no model produces precip there at night as CAPE has disappeared, possibly by advection of lower theta-e air from N Spain.

—–

Fig. 3: Day 1 forecast valid for Thu 18 July 2013 for the Helsinki domain.

—–

Day 1 Forecast

Valid: Thu 18 Jul 2013 08:00 to Fri 19 Jul 2013 06:00 UTC

Issued: Thu 18 Jul 2013 09:07

Forecaster: ESSL TESTBED

A level 1 was issued for NE Latvia, E Estonia and parts of NW Russia mainly for severe wind gusts, large hail, as well as a slight risk for tornados.

Embedded into a NW upper level flow is a short wave trough that is expected to lead to some severe weather in the regions indicated above.

Low level moisture is advected from the Baltic Sea leading to considerable CAPE levels around 1000J/kg. In the northern parts of the Level 1 area this overlaps with considerable low level shear and a low LCL which indicates a slight risk for tornados in this area. The strongest activity is expected already in the afternoon hours.

—–

In due time for the afternoon nowcasting, plenty of storms formed up over the Toulouse domain. The “Team A” was excelling itself in issuing large watch areas mainly for hail and wind in Spain and mainly for heavy rain in France and the Western Alps. A parallel “Team B” covered the isolated convection over Latvia and Lithuania. As the last action of the day, even a tornado watch was issued for the Estonian storm visible in Fig. 4, as it moved into a region of Easterly surface wind and hence even stronger low-level shear than the models predicted.

Fig. 4: Nowcast display at 1545 UTC Thu 18 July 2013: Meteosat E-View satellite image and lightning detections within the last 15 minutes.

Tomorrow (Fri 19 July 2013) will be even more interesting for Northeastern Europe: an even stronger cold front and a pronounced short-wave trough will cross the Baltics, Poland and Belarus. The day 2 team issued the so far largest level 2 area centered over Belarus for tomorrow.

Today saw a premiere: the first tornado watch during the Testbed 2013 was issued! It is presented below in Fig. 1, along with the very comprehensible discussion that was added by the nowcasting team on duty.

Fig. 1: Nowcast issued at 15 UTC Wed 17 July 2013, valid for the 15 to 17 UTC period.

—–

Nowcast

Valid: Wed 17 Jul 2013 15:00 to Wed 17 Jul 2013 17:00 UTC

Issued: Wed 17 Jul 2013 14:43

Forecaster: TEAM GREEN ROOM

Watches for imminent severe weather in the next 120 minutes are issued for these areas:

…extreme northwest Turkey and the adjacent Black Sea…

A persistent convective band over the southwest Black Sea has been sending convective elements southwestward torward the coast in Northwest Turkey just west of Istanbul. This convective line occasionally erupts cirrus plumes at times, which are carried back toward the northeast, as the winds alot above the EL are very strong and of the opposite direction from below. This line also intersects a sea-breeze convergence line which is evident from visible satellite imagery.

Although there is very little supporting evidence of a within in the sprase surface data, and the boundary layer air over the Black Sea does not appear to be appreciably colder than the sea-surface temperature, the visible and local radar trends suggest that this persistent convective line is associated with a local convergence zone. The numerical models support CAPE values anywhere from 500-1000 J/kg, suggesting there may be adequate instability to stretch any possible vertical vorticity along the convergence zone into a brief non-mesocyclone tornado over the water (a waterspout). Any waterspouts that form close to the coast have the possibility to move onshore and cause weak tornadic damage.

—–

Fig. 2 shows the cumulus band over the Black Sea which prompted the team to issue this tornado watch, and Fig. 3 presents the 12 UTC Istanbul sounding. The rather cool upper levels in the vicinity of the weak cut-off low over the Balkans resulted in at least a little bit of uncapped CAPE over the Black Sea. The pronounced low-level Northeasterly winds and the upper-level Southwesterly winds can clearly be seen in the sounding. This bizarre wind profile resulted in well-organized showers with weak electrification, which very slowly moved to the Southwest while their anvils were carried away in the opposite direction. The intersection of this supposed convergence line with the usual “frictional” convergence along the coast (visible by small cumulus clouds in Fig. 2) was a suspicious setup for a concentration of low-level vorticity into a waterspout spin-up. We were particularly aware of this scenario, because yesterday two spouts were observed at the Romanian coast in a very similar setup of two intersecting convergence lines. Till the time of writing, no spout reports from the Istanbul area have arrived yet, though.

Fig. 2: Nowcast display at 1430 UTC Wed 17 July 2013: Meteosat E-View satellite image and lightning detections within the last 5 minutes.

Fig. 3: 12 UTC Wed 17 July 2013 sounding from Istanbul (Turkey).

Apart from that, it seems like the participants get a bit fed up with forecasting the almost same Spanish storms from day to day. The day 1 team this morning chose the Sofia domain (independent from the surprising Istanbul scene, but for a chance of organized storms under strong Southwesterly upper-level flow over the Turkish mainland, which did not really materialize), and the day 2 team even switched to the Helsinki domain. More on this foreseeable change over Northeastern Europe will be written tomorrow.

Before that, however, still the nowcasting was arguably today’s most interesting part: Yesterday’s storms over Northern Spain and Southern France did not only bring another regional outbreak of severe weather, but also provided a good opportunity to test the “NearCast” product developed by Ralph Peterson from the University of Wisconsin, who is also on site in Wiener Neustadt this week.

This product displays fields of the equivalent-potential temperature at 778 and 497 hPa derived from satellite retrievals. These fields are computed hourly (at least in areas where no obstructive cloudiness is present), and a Lagrangian advection with GFS forecast winds is computed for up to 9 hours into the future. Users can switch between a separate view of either of these fields, or of the resulting difference, which shows potentially stable or unstable areas. The NearCast product has been a matter of active discussions for the whole Testbed period, as not everybody agrees on the concept of potential instability: critics claim that it is rather a “proxy” for good or not good conditions for convection than it actually reflects the physics of (surface-based) convection. Though not everybody acknowledged the product’s added value in comparison to conventional satellite products, it was generally agreed that it performed well.

During yesterday’s treatment of the French storms, Ralph pointed out another, less obvious possibility to use the 778 hPa theta_e field alone. In a quiescent situation like the present one with an absence of pronounced horizontal temperature variations, its maxima and minima can chiefly be interpreted as local maxima or minima of moisture on top of the growing convective boundary layer. A moisture minimum at ~800 hPa means that vertical mixing will effectively reduce the near-surface moisture by means of strong upward mixing as soon as the convective boundary layer grows deep enough. Conversely, near-surface moisture is largely preserved if the layers around 800 hPa also show a high moisture content.

This interpretation was put to the acid test yesterday (Tue 16 July 2013) for the situation prior to the storms over Southern France. The surface map at 08 UTC showed a limited moisture gradient in this area with dew points around 19°C in the foreland of the Pyrenees and around 15°C further North (Fig. 4). The 08 UTC NearCast theta_e field confirmed this moisture gradient also at 778 hPa (Fig. 5). Would daytime mixing indeed act to further decrease the surface dew points to the North, or, conversely, would the reduced daytime mixing (along with further evaporation) manage to maintain the very moist boundary layer over Southermost France?

Fig. 4: 08 UTC Tue 16 July 2013 surface map.

Fig. 5: NearCast equivalent-potential temperature at 778 hPa at 08 UTC Tue 16 July 2013.

Et voilà, Fig. 6 shows a satellite display and the surface observations four hours later, at 12 UTC. Dew points have indeed lingered around 20°C close to the Pyrenees, where numerous storms were now firing up. On the other hand, they have dropped as low as 11 to 12°C on the other side of this “dryline” from Bordeaux to the Northeast. The 12 UTC Bordeaux sounding confirmed a deep and well-mixed boundary layer, in which the lack of moisture precluded any storm development (Fig. 7).

Fig. 6: Nowcast display at 1200 UTC Tue 16 July 2013: Meteosat E-view satellite image and SYNOP observations.

Fig. 7: 12 UTC Tue 16 July 2013 sounding from Bordeaux (France).

In the following hours, numerous pulse storms formed over Southern France with large hail, severe wind gusts and excessive rain in many places. Over Northern Spain, deep-layer shear was slightly enhanced to 15 m/s, which was sufficient for a better storm organization. The most significant storm was a bow echo which brought a 34 m/s wind gust in Miranda de Ebro and extensive hail and wind damage in the La Rioja region nearby.

While the forecast models had agreed on convective initiation over Northern Spain, they had been much more reluctant with precipitation signals over Southern France. However, in yesterday’s day 1 forecast we had extended the 50% thunder line and the level 1 well into Southern France, since the observed high dew points and the probably reduced vertical mixing (see above) convinced us of a high CAPE buildup and an explosive release as soon as the outflow boundaries from the Pyrenees would move into the foreland. The other level 1 area in the Western Alps for a chance of excessive rain worked out as well. All in all, we could agree that this was the so far best forecast of the entire Testbed period, which is why it is presented in full detail in Fig. 8. The afternoon nowcasts were equally successful (not shown). And now some champagne, please!

Fig. 8: Verification of the day 1 forecast valid for Tue 16 July 2013.

—–

Day 1 Forecast

Valid: Tue 16 Jul 2013 08:00 to Wed 17 Jul 2013 06:00 UTC

Issued: Tue 16 Jul 2013 08:55

Forecaster: ESSL TESTBED

A level 1 was issued for N and E Spain and s France mainly for large hail and severe wind gusts.

A level 1 was issued for NW Italy mainly for heavy rainfall.

…other comments…

N Spain, S France:

Moderate CAPE (~1000 J/kg) can build where coastal moisture gets advected inland. 0-6 km shear wind shear is below 15 m/s. We expect some pulse storms. The strongest ones may bring localized hail and downbursts.

E Spain:

0-6 km wind shear increases up to 20 m/s further South and a travelling short-wave trough provides some lift, but it is still doubtful if the strong cap can be broken. If storms form, they can organize into strong multicells or isolated supercells with large hail and severe wind gusts.

Western Alps:

We expect afternoon and evening storms in an environment of weak vertical wind shear. CAPE is rather low on the French and Swiss side, but increases to ~1000 J/kg over the Po valley, where surface observations show more low-level moisture (dew points around 19C). With the arrival of a weak short-wave trough, convection can continue into the night hours and may grow into a larger cluster. Heavy precipitation is possible.

—–

Today’s verification session turned into an instance of “forensic meteorology”, as supercell storms occurred yesterday afternoon in two areas that we were not fully aware of.

The first area was Turkey, were the satellite loops indicated several suspicious storms with either strongly left- or right-moving tracks with respect to the Southwesterly upper-level flow. Fig. 1 shows the situation at 1450 UTC. from satellite view.

Fig. 1: Nowcast display at 1450 UTC Mon 15 July 2013 (zoom on Turkey): Meteosat E-View satellite image and detected lightning within the last 5 minutes.

Out of the scattered afternoon storms over Turkey, the two Northwesternmost ones in Fig. 1 were most interesting. They were tiny and did not even produce any lightning at that time. However, their V-shaped appearance already marked them as possible supercells. The different orientation of their anvils further indicated their strongly deviant motion, which was confirmed by the satellite loops: the Western supercell moved almost due North, and the Eastern one almost due East.

Fig. 2: 12 UTC Mon 15 July 2013 sounding from Ankara (Turkey).

The next step of the “forensic meteorology task force” was to find a proximity sounding. Indeed the 12 UTC Ankara sounding was launched close to the areas of convection (Fig. 2). It showed a typical situation for summertime Turkey with a very deep and dry boundary layer and neutral (i.e., moist-isentropic) lapse rates aloft (disregarding the obviously erroneous measurement at 570 hPa). This is typical for arid regions and strong summertime insolation, when the sensible heat flux is immediately converted into convective mixing. The more eye-catching feature was the rather strong Southwesterly upper-level flow with 20 m/s at 500 hPa. Even though the cloud-bearing layer was remarkably shallow with cloud bases near 3 km and cloud tops near 7 km, the effective vertical wind was indeed considered supportive of “high-based and low-topped supercells” (copyright Pieter).

Fig. 3a: 14 UTC Mon 15 July 2013 surface map (zoom onto Turkey).

Fig. 3b: 15 UTC Mon 15 July 2013 surface map (zoom onto Turkey).

The 14 and 15 UTC surface maps (Fig. 3a and 3b) also revealed the development of a very strong circulation around the daily thermal low over South-Central Turkey. Eskisehir station (LTBI), situated immediately downstream of the conspicuous left-moving storm in Fig. 1, reported NW wind with 15 m/s at 15 UTC. Further East, Ankara station (LTAC) was not far from the right-moving storm and showed E wind with 5 m/s at 14 UTC. Modifying the sounding in Fig. 2 with these low-level winds would indeed create plenty of negative storm-relative helicity and an enhanced chance for anticyclonic, left-moving supercells around Eskisehir, and vice versa around Ankara, the way it was indeed observed. Ankara also got a direct hit by one of the other storms between 14 and 15 UTC: the temperature dropped from 30 to 18°C, the dew point rose from 10 to 16°C and a maximum gust of 28 m/s was recorded. This confirmed the potential for severe downbursts in such an environment of considerable wind shear and at the same time very dry subcloud layers. [Side note: The allegedly even higher mean wind and wind gusts during fair weather conditions at 15 UTC in Izmir (LTBJ) at the West coast are a coding error – knots were accidentally coded as m/s.]

Fig. 4: Nowcast display at 1300 UTC Mon 15 July 2013: Meteosat E-View satellite image, SYNOP observations and lightning detections within the last 5 minutes.

The second area which saw a supercell yesterday (Mon 15 July 2013) was Eastern Hungary, which was crossed by a cold front from the Northwest. A prefrontal environment of a few hundred J/kg of CAPE and deep-layer shear between 10 and 15 m/s in the Northwesterly flow did not look all too conducive for organized storms. However, around noon surface observations indicated that a shallow low formation in the wake of the Slovakian High Tatra mountains made the wind turn back to Southerly directions over a confined region of Eastern Hungary which also still saw enough sunshine. The nowcast display at 1300 UTC (Fig. 4) shows an arc of small cumulus clouds along the cold front from Hungary across Western Slovakia into the Eastern Czech Republic. Ahead of it, the station of Debrecen (LHDC) reported weak Southerly wind. The supercell formed right at that moment in close vicinity and tracked Southeastward. Its long-lived rotation was confirmed by a doppler radar analyses of the Hungarian Meteorological Service (OMSZ) and by pictures taken by spotters. Small hail, strong wind gusts (the highest measurement was 20 m/s) and maybe marginally flooding occurred with this storm.

Fig. 4a: GFS forecast sounding for 12 UTC Mon 15 July 2013 for Debrecen (Hungary).

Fig. 4b: As in Fig. 4a, but for ECMWF.

Fig. 4c: As in Fig. 4a, but for COSMO-EU.

It is worth to examine the 12 UTC forecast soundings for Debrecen from GFS, ECMWF and COSMO-EU models (Fig. 5). While the models agreed on a few hundred J/kg of CAPE, the near-surface wind field was forecast in a different way: GFS showed a smooth Northwesterly flow (and consequently a drier boundary layer), ECMWF already weakened the surface winds and COSMO-EU even had signs of a temporary Southerly surface wind, the way it actually occurred. It seems like the finer model resolution helped significantly in capturing this mesoscale low development. Inserting the 13 UTC Debrecen measurements (temperature of 24°C, dew point of 14°C and weak Southerly wind) in the 12 UTC Poprad sounding (Fig. 5), which was already behind the cold front but otherwise in vicinity, indeed yields substantial CAPE and deep layer shear of more than 20 m/s, even more than the most “optimistic” forecast by COSMO-EU had shown.

Fig. 5: 12 UTC Mon 15 July 2013 sounding from Poprad (Slovakia).

Both the Turkish and the Hungarian case show situations when mesoscale processes favourably modified the conditions for organized convection despite a rather inconspicuous large-scale setup. Greg mentioned the nice term “mesoscale accidents”, which is colloquially used in the United States for such happenings. The structured land/sea distribution and orography of Europe makes such “mesoscale accidents” in Europe much more common than in the United States.

Today, as many maps as never before were manually drawn at the ESSL testbed: Yesterday’s Hungarian case was so alluring that it was re-analyzed by hand. Further hand analysis were drawn today for the 06 UTC, 07 UTC and 12 UTC maps of the Toulouse domain, which was – despite yesterday’s surprises – again chosen for today’s day 1 forecast. Yesterday a lot of storms occurred there and positively verified our lightning forecast, but severe weather was neither reported from within our level 1 area (see yesterday’s post) nor from anywhere else.

Hooray for the third Testbed week! 10 new participants from Poland, Hungary, Austria, Germany, Spain and the United States arrived, but did not bring any new weather yet.

Fig 1: ECMWF forecasts of 500 hPa geopotential (black contours), temperature (colour shades) and wind (barbs) for the week from Mon 15 July 2013 to Fri 19 July 2013 (12 UTC each day).

Fig. 1 presents the daily sequence of the 500 hPa ECMWF forecasts. The primary interest of this week will still be on Spain and (hold on tight!) even Northern Algeria, where the subtropical branch of the jet stream along with opposed Easterly low-level flow will boost vertical wind shear to values in excess of 20 m/s. Up to moderate CAPE may form where rich low-level moisture gets advected inland. It is anticipated that a series of short-wave troughs will discriminate between days with only isolated storms that struggle against the strong cap, and other days with rather widespread activity over Spain.

Otherwise, weakly organized thunderstorms are expected in the range of slowly decaying upper-level lows over Eastern Europe and the Balkans, and – especially after midweek – over the orography of Southern France and the Western Alps despite weak upper-level ridging. Towards the end of the week, it looks like the active frontal zone over the far North will run another attempt to erupt towards Eastern Europe, but by then discrepancies between GFS (not shown) and ECMWF become quite high and a concurrence with unstably stratified air is still doubtful.

The day 1 forecast for the Toulouse domain which opened this week is inserted below in Fig.2.

Fig. 2: Day 1 forecast issued on Mon 15 July 2013.

—–

Day 1 Forecast

Valid: Mon 15 Jul 2013 15:00 to Tue 16 Jul 2013 06:00 UTC

Issued: Mon 15 Jul 2013 14:51

Forecaster: ESSL TESTBED

A level 1 was issued for central Iberian Peninsula mainly for large hail.

A level 1 was issued for northern Italy mainly for large hail.

There is a high chance of storms over parts of Spain and northern Italy as result of weakly capped CAPE and convergence resulting from orography. In combination with around 10 m/s deep layer shear multicell storms should be the mode. Over northeastern Italy and eastern Spain the 0-6 km shear reaches over 15 m/s which supports weak supercells with large hail as primary threat.

Isolated severe wind gusts are not as likely due to weak flow and rather moist mid levels with mediocre lapse rates over most of Spain and northern Italy with the exception of northwestern Italy where Milan 12Z sounding shows a very dry layer at 600 hPa.

The probability of supercellular tornadoes is very low because of absence of significant low level shear (0-1 km shear less than 7 m/s in most of the area)

Storms may remain active overnight over Spain and northern Italy.

—–

While we were still working on our day 1 and day 2 forecast, respectively, the sky darkened outside our windows and some passing thundery showers reminded us that we should not completely omit today’s weak cold front, which was leisurely moving Southeastward into Eastern Europe and across the Alps. In the Southern Alpine region deep-layer shear up to 20 m/s overlapped with non-zero CAPE in the forecast maps, the low-level flow was still Southerly, and last week’s experiences had taught us utmost care in such situations.The situation had a striking similarity with last Thursday (extensively covered in the last two blog entries), despite a lower degree of moisture, lapse rates, wind shear and everything else you can imagine or not. We awaited some better organized storms to form over the Southern Alpine region again.

However, due to active discussions and the far from trivial situation, the day 1 to 5 forecasts took longer than expected and no time was left for nowcasting any more. This was no serious matter, because not too many interesting storms had formed yet. The Spanish storms were only slowly gearing up, and in the Southern Alpine region a little bit happened over Slovenia and Croatia but Northern Italy was left out this time (at least till the time of writing). Midnight isn’t exactly the right time to find out why.