Conference Meteo Alarm Budapest 14 – 15 December 2009 1 Forecasting severe convection Problems and some possible solutions photo: Karim Hamid.

Conference Meteo Alarm Budapest 14 – 15 December 2009 1

Forecasting severe convection Forecasting severe convection Problems and some possible solutionsProblems and some possible solutions

photo: Karim Hamid


Damage due to convection occurs over the whole year !Damage due to convection occurs over the whole year !


A distinction must been made A distinction must been made between warm and cold between warm and cold season :season :

In warm season situation, severe In warm season situation, severe convection is highly driven by convection is highly driven by (buoyant) instability, in (buoyant) instability, in cooperationcooperation with some windshearwith some windshearIn cold season situation, severe In cold season situation, severe convection is highly driven by convection is highly driven by dynamics and windshear, in dynamics and windshear, in cooperationcooperation with some with some (buoyant) (buoyant) iinstability (sometimes even neutral)nstability (sometimes even neutral)


Especially in cool season, damage due Especially in cool season, damage due to convection can be to convection can be underestimateunderestimate due to the (very) low instabilitydue to the (very) low instability

Some recent cases in Belgium show Some recent cases in Belgium show this…this…


Tornado case of 01/10/2006

photo: Karel Holvoet







photo: Karel HolvoetTubize


Tornado case of 21/01/2008 Grote Brogel

photo: Karim Hamid

A survey on the spot was made to investigate the damage


Tornado case of 05/02/2008 Dendermonde+Lier

photo: Karim Hamid

A survey on the spot was made to investigate the damage


At the RMI, we do efforts to better forecast At the RMI, we do efforts to better forecast situation with severe convection.situation with severe convection.

This is done by an “IThis is done by an “Ingredients-Based ngredients-Based MethodologMethodology, first puted foreward by Chuck y, first puted foreward by Chuck Doswell in 1987Doswell in 1987”.”.It says that It says that if some if some atmospheric ingredients atmospheric ingredients coincidecoincide in time and place, in time and place,the risk of severe the risk of severe convection increases, convection increases, wherever on the worldwherever on the world..


The most important ingredients are The most important ingredients are (buoyant) (buoyant) instabilityinstability, , windshearwindshear and and the the synoptic configurationsynoptic configuration..

An important point in all this is the An important point in all this is the problem of the problem of the ‘triggering’‘triggering’ of of convection’, which is not explicitly convection’, which is not explicitly assimilateassimilated in the whole storyd in the whole story..


instability

For the instability, For the instability, we found that the we found that the LI at 700 hPa LI at 700 hPa works better in works better in cool air masses cool air masses than the than the traditional LI.traditional LI.

LI 700LI 700

LI 500LI 500


The BRN Shear The BRN Shear could be use to could be use to disciminate disciminate between severe between severe convective convective situations and non-situations and non-severe situations.severe situations.

A median value for A median value for the BRN shear was the BRN shear was found of 22 m²sfound of 22 m²s-2-2 with no severe with no severe convection, and 80 convection, and 80 m²sm²s-2-2 with severe with severe convection !convection !

wind shear


The Te850 can The Te850 can be use to be use to outlineoutline the the regions where regions where MCSs are very MCSs are very improbableimprobable with with a threshold a threshold value around value around 50°50°

Forecasting Mesoscale Convective Systems (MCS)


One can outline regions where the risk on MCSc is One can outline regions where the risk on MCSc is elevatedelevated

Forecasting Mesoscale Convective Systems (MCS)

Te850Te850


Also some synoptic archetypical Also some synoptic archetypical configurations ate made, e.g for the jetstreakconfigurations ate made, e.g for the jetstreak

(nearly) all tornado’s are formed in storms in the (nearly) all tornado’s are formed in storms in the left exitleft exit of a jetstreak of a jetstreak

(based on 211 cases)


Also some synoptic archetypical Also some synoptic archetypical configurations ate made, e.g for the jetstreak configurations ate made, e.g for the jetstreak (isotachs 300 hPa)(isotachs 300 hPa)

A clear difference is visible between the two cases (left exit cyclonic curved jetstreak plays a key role)A clear difference is visible between the two cases (left exit cyclonic curved jetstreak plays a key role)

tornado cases (winter)tornado cases (winter) non tornado cases (winter)non tornado cases (winter)


An attempt is made to improve some existing An attempt is made to improve some existing severe weather indices (STP, SCP, …), with an severe weather indices (STP, SCP, …), with an adaption for our region and by using the adaption for our region and by using the threshold values and best working threshold values and best working parameters, found in our research.parameters, found in our research.


4540.

1500

2000.

²/²40.

/14

03.

/10

01.

2

5005.0

)500(CAPE

CAPE

m

MLLCL

sm

BRNshear

sm

kmshear

sm

kmshearVVVV

ZSTP

250.

1500

2000.

²/²40.

/15

03.

/12

01.

2

7001.0

)700(CAPE

CAPE

m

MLLCL

sm

BRNshear

sm

kmshear

sm

kmshearVVVV

WSTP

[If Lift Index < +1K][e850 > 45°C][shear01 8 m/s]

[If Lift Index 700 < +0.5K][e850 < 45°C][shear01 14 m/s]

model-convection

windshear humidity low levels

instability

example: iexample: improvement of the Significant Tornado mprovement of the Significant Tornado Parameter (STP)Parameter (STP), one for cool, and warm air masses, one for cool, and warm air masses

An attemp to assimilate the triggering of convection is done by inserting the vertical velocity in the model (1).

(1) Contribution to severe weather and multimodel ensemble forecasting in Belgium, PhD Thesis D. Dehenauw, 2006

warm air masses

cool air masses


A nice example of a unexpected tornado case:A nice example of a unexpected tornado case:

The Hautmont-tornado of August 3th, 2008The Hautmont-tornado of August 3th, 2008

At least At least 66 tornadoes tornadoes were were observed on observed on that day !that day !

The tornado The tornado in Hautmont in Hautmont was the most was the most intense one..intense one..

Source: European Severe Weather Database - http://essl.org/ESWD/

Hautmont

F2

F4

F2



The tornado of Hautmont of 03/08/’08

F4 damage F4 damage !!



A survey on the spot was made to investigate the damage: some pictures…

Photo: Karim Hamid



Photo: Karim Hamid



Photo: Karim Hamid



Photo: Karim Hamid



Photo: Karim Hamid



Photo: Karim Hamid



Photo: Karim Hamid


DWD analysis 24z


The situation was The situation was apparentlyapparently not not threateningthreatening, with only the , with only the passage of a stable frontal wave in a westerly flowpassage of a stable frontal wave in a westerly flow

Position tornado Hautmont



At upper levels, an important trough swept over Wester Europe…At upper levels, an important trough swept over Wester Europe…





Problem: ahead of the active wave, the cloud shield broke with as Problem: ahead of the active wave, the cloud shield broke with as a result warming of the boundary level and increasing instability…a result warming of the boundary level and increasing instability…

bright spells

18z



Instability was only Instability was only marginalmarginal to moderate with a CAPE around 700 to moderate with a CAPE around 700 J/kg…J/kg…

Source: KMI/D. Dehenauw



Even a larger problem: the windshear ahead of the wave was Even a larger problem: the windshear ahead of the wave was exeptional high!exeptional high!

A 0-1 km windshear near 20 A 0-1 km windshear near 20 m/s !m/s !

A 03SRH around 350 m²/s² !A 03SRH around 350 m²/s² !





It was a It was a combinationcombination of very strong forcing, of very strong forcing, exeptional high windshear (low level and exeptional high windshear (low level and deep level & storm-relative) and the deep level & storm-relative) and the precense of some instability which caused an precense of some instability which caused an exeptional violent tornado…exeptional violent tornado…

So, So, apparentlyapparently all the all the necessarynecessary ingredientsingredients were present at that place…were present at that place…



Was this situation well Was this situation well predicted ?????predicted ?????



Belgium: Belgium: orange color orange color for floodsfor floods

France: France: yellow color yellow color for floodsfor floods

Meteo alarm 18zMeteo alarm 18z



Despite the fact that the warning was valid for floods, a Despite the fact that the warning was valid for floods, a risk for risk for strong windgustsstrong windgusts was mentioned in the text at the was mentioned in the text at the RMI (RMI (based on the very high compouned parametersbased on the very high compouned parameters).).



Was this situation well Was this situation well predicted ?????predicted ?????

NO !!!!NO !!!!

……or (at least) not enoughor (at least) not enough



Was this situation Was this situation predictablepredictable

??????????



Very local things like tornadoes are Very local things like tornadoes are barelybarely predictable, because the very local situation predictable, because the very local situation just near a convective storm can’t be just near a convective storm can’t be predicted by models. Moreover, not predicted by models. Moreover, not everything is known concerning everything is known concerning tornadogenisis…tornadogenisis…



However, there were some However, there were some importantimportant signalssignals from the meso- from the meso-models:models:

Strong signalsStrong signals from the SCP, STP and original STP indices in ETA from the SCP, STP and original STP indices in ETA





- Abnormal highAbnormal high EEnergy nergy HHelicity elicity IIndexndex (EHI-black (EHI-black lines) for a lines) for a baroclinic situationbaroclinic situation

- - Very highVery high low low level SRH (colored level SRH (colored lines) up to 350 lines) up to 350 m²/s² wich is very m²/s² wich is very favorable for favorable for tornadogenesistornadogenesis




This situation This situation beautifulbeautifully ly showed the power of this showed the power of this compoundcompound indices indices


A 2the recent example: 25-26 May A 2the recent example: 25-26 May 20092009Passage of several very destructive Mesoscale Convective Passage of several very destructive Mesoscale Convective

Systems Systems

- Passage of a strong - Passage of a strong supercellsupercell over Belgium, followed by a over Belgium, followed by a very fast very fast Squall LineSquall Line

- Important Important wind damagewind damage

-Relative widespread Relative widespread hailhail damage with stones up to damage with stones up to 12 cm12 cm in in northern France and almost northern France and almost 10 cm10 cm in Belgium in Belgium


Hail case of 25-26/05/2009 in Belgium en N-France





Lapse rate of 19 °C between 800 hPa-600 hPa

This is 8.5 °C/km !!

Typical Elevated Mixing Layer (Spanish Plume)



Squall Line@ 100 km/h !

00.30z

supercell



23.50z

Supercell with pendant shape / hook echo



Hail upto 12 cm near the French border (Raillencourt-Sainte-Olle) – Source : E. Wesolak & P. Mahieu – Observatoire Français des Tornades et des Orages Violants



La forêt de Flines, à la frontière franco-belge – Source : windstorm.111e-monsite.com



Some people collected the stones in the freezer, and I could recover Some people collected the stones in the freezer, and I could recover some of the biggest hail stones in the recent Belgian history…some of the biggest hail stones in the recent Belgian history…

Biggest stone recovered with a diameter of 9.2 cm



This case permited to make some cross sections of the biggest hail This case permited to make some cross sections of the biggest hail stones, with detailled photos as result with some stones, with detailled photos as result with some interesting interesting features …features …

Features like enclosed air bubbles, ring structure, embryo,…are very nice visible



Also, the lob structure and radial air channels are well distinguishable





What with the model indices?What with the model indices?



High STP/SCP valuesHigh STP/SCP values High EHI valuesHigh EHI values



Due to the model-convection, the model-fields Due to the model-convection, the model-fields were somewhat were somewhat disdisruptedrupted





Potential hailstone-diameter is estimated by the model, Potential hailstone-diameter is estimated by the model, based on the Fawbush-Miller algoritm (1). based on the Fawbush-Miller algoritm (1).

(1) Contribution to severe weather and multimodel ensemble forecasting in Belgium, PhD Thesis D. Dehenauw, 2006

18z18z 21z21z

Hail path



Conclusion: here the IConclusion: here the Ingredients-Based ngredients-Based MethodologMethodology y also also worked very wellworked very well




Model convection is integrated in the map by blue shading

Regions with SWI > 0.8 are shown on a large scale

Jetstreaks are also integrated on the map

Tests of some indices on a global model (GFS) for larger regionsTests of some indices on a global model (GFS) for larger regionsRecent example of the low topped storm of November 23the, 2009 in Belgium/Netherlands

No model convection foreseen: smaller risk



hail

15z


Tests of some indices on a global model (GFS) for larger regionsTests of some indices on a global model (GFS) for larger regions

In Belgium, hail stones up to 5 cm in length fell near Bruges and had a very irregular shape

Recent example of the low topped storm of November 23the, 2009 in Belgium/Netherlands



As with the hail storm of 25 May, I made some cross-section taken in transmitted light to retrieve internal details

I sent the full report of this hail case to the Meteo Alarm forum



Wind damage

18z



In the Netherlands, local important wind damage was registered



model convection integrated in the SWI index


General conclusions:General conclusions:

Despite the shortcomings of the IDespite the shortcomings of the Ingredients-ngredients-Based methodologyBased methodology using models using models (unrepresenative proxy soundings, (unrepresenative proxy soundings, shortcommings in the models, etc…), they shortcommings in the models, etc…), they can help the forecaster to show regions can help the forecaster to show regions where there could be a where there could be a threateningthreatening situation.situation.

Even in seeming harmless configurations Even in seeming harmless configurations (e.g tornado Hautmont), they can give (e.g tornado Hautmont), they can give important signals and then, the forecaster important signals and then, the forecaster can go in detail to find out if there realy is a can go in detail to find out if there realy is a threat.threat.



An attempt to forecast severe An attempt to forecast severe convection in a uniform and objective convection in a uniform and objective

way (the same methodology by all way (the same methodology by all the forecasters)the forecasters)

The ChecklistThe Checklist


Online CHECKLIST for SEVERE CONVECTIONForecasting (severe) convective storms

383/8 18-00z


Online CHECKLIST for SEVERE CONVECTION



Direct access to the model maps







Color indicates if a treshold is attained

A risk is given for thunderstorm in general and also for severe storms

Some facultative info can be given – proposition of e.g warning color

The results ar for a cool air mass situation


Some case studies I made from the past in Belgium are availble on Some case studies I made from the past in Belgium are availble on my website:my website:

http://users.fulladsl.be/spb4195/cursus.htmhttp://users.fulladsl.be/spb4195/cursus.htm

Conference Meteo Alarm Budapest 14 – 15 December 2009 1 Forecasting severe convection Problems and some possible solutions photo: Karim Hamid.

Documents

damage slide

neutral slide

karim hamid slide

karel holvoet slide

wind shear slide

tornado case

karel holvoet tubize

triggering of convection