francesc figuerola , montse aran , carme farnell , jordi mateo , …€¦ · francesc figuerola1,...

RESULTS AND VERIFICATION

An early severe weather warning system in Meteorological Service of Catalonia.Francesc Figuerola1, Montse Aran1, Carme Farnell1, Jordi Mateo1, Tomeu Rigo1, Santi Segalà1

1 Meteorological Service of Catalonia (SMC)francesc.figuerola@gencat.catECSS 2019. 10th European

Conference on Severe Storms

REFERENECESFarnell, C., T. Rigo, and N. Pineda, 2017: Lightning jump as a nowcast predictor: Aplication to severe weather events in Catalonia. Atmos. Res., 183, 130-141.Rigo, T., and M. C. Llasat, 2016: Forecasting hailfall using parameters for convective cells identified by radar. Atmos. Res., 169, 366-376.Williams. E., and Coauthors, 1999: The behavior of total lightninig activity in severe Florida thunderstorms. Atmos. Res., 51 (3), 245-265Willimas, Earle R: The electrification of severe storms. Severe Convective Storms. Amercian Meteorological Society, Boston, MA, 2001. 527-561.

MOTIVATIONSevere hail occurs 10 days peryear in average in Catalonia,according to a SMC internaldatabase (other severeweather phenomena has not been included).

MAIN GOALImplementation of an early severe weather warning based on aLightning Jump (LJ) algorithm adapted to Catalonia (Farnell et al2017) in order to forecast severe weather events.LJ is an abrupt increase of total lightning in a storm due to astrong updraft which favours the crash between ice particlesand higher separation of charge (Willimas et al, 1999,2001).

LIGHTNING JUMP: TECHNIQUE, RESULTS AND VERIFICATION

Period events POD FAR BIAS

2006-20131 49 0,73 0,11 0,70

20162 69 0,94 0,25 1,00

20172 109 0,82 0,22 1,01

20182 146 0,88 0,19 0,89

20191 14 0,82 0,25 1,09

LJ verification Lead Time: the difference between the LJobservation and the severe weather occurrence

LT>30’: 38 %LT>10’: 75 %Mean: -23’

LEAD TIME (2017-2019)

TECHNIQUE

Lightning data is analysed every minute. Flashes are grouped in geo-referenced densitymatrix of 1x1 km2 cells.

Cells properties, size andTotal Lightning (TL), for14 minutes continuouslyperiod is analysed.

TL is grouped and averaged in2 minutes period.Difference of TL rate betweenperiods is calculated (DFDRT).

longitude

lati

tud

e

After that standard deviation (σ) is calculated for the last 14 minutes.If DFDRT of the last period (DFDRT6, real-time) exceeds the 2*σ threshold then a lightning jump warning is triggered.

LJ

If DFDRT6 > σ*2 LJ

σ= {DFDRT1, DFDRT2, DFDRT3, DFDRT4, DFDRT5}

1:severe weather and small hail2:only severe weather

EXAMPLES

Heavy rainGust wind

17/07/2019

Large andvast hailHeavy rain

18/09/2019

A

B

A

Sant Cugat del Vallès

B

Sant Cugat del Vallès

B

0 25 50 km Arbúcies

0 25 50 km

Catalonia islocated in the NEof the IberianPeninsula. Thisarea covers32.000 km2.

Legend

AREA OF STUDY

THE SEVERE WEATHER WARNING SYSTEM OF THE SMC AND THE CATALAN EMERGENCY CENTRE

SMC has developed a specificsoftware to visualize LJ. It shows:- LJ level (with or without flash

multiplicity)- time and position of every LJ- a summary of whole electrical

evolution.

Once a LJ warning is triggered, the software:- estimates direction nowcasting of the

storm with radar data (Rigo and Llasat,2016). This nowcasting is overlaid at themap with concentric coloured circles.

Finally, forecaster analyses the differentprobability levels of occurrence in the nexttwo hours.

Municipality Flood warning for the Alt Empordà

WORKFLOW

1.Lightning Jump Alert 2. Decision making from SMC 3. Hazard warning issued 4. Emergency Centre 5. Risk warning issued

0 25 50 km