Using BRAMS in simulating mesoscaleprocesses in Argentina

Matilde Nicolini

Departamento de Ciencias de la Atmósfera y los

Océanos. UBA.

CIMA- Centro de Investigaciones del Mar y la

Atmósfera. CONICET-UBA

Focus of presentation in:

� Evaluate the capability of BRAMS to properly reproduce conditions and features at different scales prior and during severe storm events

� Improvement in detail analysis of physical interactions between moisture transport, precipitation and evolution of the South American LLJ, through downscaling and assimilation of field campaign observations

� Ultimate goal: Improve short range weather forecast in Argentina

SouthSouth AmericanAmerican lowlow--levellevel Jet Jet ExperimentExperiment

(SALLJEX)(SALLJEX)

� 15 November 2002 - 15 February 2003

Field components:

� Precipitation network

� NOAA-P3 Aircraft missions

� Upper-air sounding network (RAOBs and

pibals)

� SOP’s: 2 RAOBS daily (06 and 18 UTC)

� IOP’s: 3-4 RAOBS and/or 8 PAOBS/day

ParanaCórdoba

Chamical

Santiago del

Estero

Tostado

Resistencia

PampaSalta JVG Asunción

Mariscal

Rio Branco

Cruzeiro

Vilhena

Dourados

Piura

Pucallpa

Ica

Puno

Cobija

Roboré

Santa Cruz

Villamontes

Trinidad

Foz de Iguazu

Zipser et al., 2006

Extreme values ofvariables from7 years TRMM

database used as proxy to

identify intense convection

� Hailstorm in Mendoza (February 14th 2005)

� Tornado generating storm in Northeastern Argentina (April 28th 2001)

� Environmental characterization prior and during MCSs development during a SALLJ event using downscaling and data assimilation (February 5-7 2003)

� Mesoscale convective system (December 19 2002)

Hailstorm simulation in Mendoza

Matilde Nicolini,Yanina García Skabar, Graciela Ulke and Paola Salio, 2005

4 nested Domains

� Grell convection parameterization only at 50 km resolution� Microphysics: two moments scheme

� Initial and boundary conditions: GDAS analyses from NCEP

� Vertical coordinate SHAVED ETA

30 atmospheric levels -9 soil levels� Horizontal resolution:

50, 12.5, 3.125 and 0.78 km

� 12 hours simulation

17 UTC 20 UTC 22 UTC

+70 dbz

Vertically integrated

cloud water species

Radar reflectivity

(observations)

BRAMS

32.7º S - 21 UTC

33.27º S - 21:30 UTC

33.05º S - 21:30 UTC

BRAMS

Vertical cross sections

along NW-SE transect

Zr (shaded),

q(graupel+hail)contours,

BRAMS Tornado simulation

Matilde Nicolini, Marcela Torres Brizuela andYanina García Skabar,

2005

� Vertical coordinate

SHAVED ETA 30 atmospheric levels -9 soil levels� Horizontal resolution:

50, 12.5, 3.125 and 0.78 km

� 12 hours simulation

� Grell convection parameterization only at 50 km resolution� Microphysics: bulk water scheme�Initial and boundary conditions: GDAS analyses

from NCEP

4 nested Domains

19:30 UTC

BRAMSaccumulatedprecipitation18 -20 UTC

BRAMSClouds and

vertical velocity20 UTC

April 28th 20 UTC

P*

1725m -red

6660m -blue

W (solid

lines), wind

vectors,

p*(blue) at

95,4m

W and

Zr(shaded)

850 hPa

Vorticity

Cloud(shaded)

Hook-echo Max. cyclonic ζ

mesocyclone

28.19ºS - 20 UTC

Cloud and total precipitating categories

Cloud, motion and

vertical vorticity

vertical alignement

(tornadogenesis signal)

Precipitation

Mesoscale convective system simulation(December 19 2002)

Juan Ruiz, Celeste Saulo, Yanina García Skabarand Paola Salio, 2005

A

B

A B15 Km

Banda brillante

Región con ecos débiles

Superficie

IR RADAR

00

0612 18

� Vertical coordinate Sigma Z (RAMS 4.3)

30 atmospheric levels -9 soil levels� Horizontal resolution:

80 and 20 km� 48 hours simulation

�Microphysics: bulk water scheme

�Grell convection parameterization activated at both domains�Initial and boundary conditions: GDAS analyses

from NCEP

TMI RADAR PR RAMSRAMS

Vertically integrated total condensate (shaded)

Precipitation rate mm/hr (contours)

06 UTC Dec 19th 2002

24-hour total accumulatedprecipitation in mm

18 – 19 Dec 2002

RAMS

ObservedCMORPH algorithm estimates

Resistencia

Santiago del Estero

GDAS

(triangles), RAMS (circles) rawins(no mark)

06UTC 19 Dec

Characterization of environment prior to the development of a sequence of subtropical MCSs during a SALLJ event using downscaling and data assimilation

Borque, Salio,

Vidal, Garcia

Skabar

and Nicolini,

2006)

(February 5-7 2003)

Cross section at 25ºS

Feb 06 to 07 15UTC

θθθθ , wind barbs and V component

Observed daily accumulated precipitation (upper panels) and modeled by BRAMS (lower panels)

White

areas

indicate

lack of

data

Downscaling and data assimilation

� Vertical coordinate

SHAVED ETA 30 atmospheric levels - 9 soil levels� Horizontal resolution:

80 and 20 km� Temporal resolution 3 hours

�Grell convection parameterization activated at both domains�Operative network and special data from SALLJEX assimilated during a 3 months period (November 15th 2002 to February 15th 2003)�Initial and boundary conditions: GDAS analyses from NCEP

80 km

20 km

Lower RMSE when data are

assimilated2

.5 DWSC1-SIGMADWSC2-SIGMAGDAS

DWSC1-ETADWSC2-ETA

�Summary

�The promising results in a generating tornado storm and a hailstorm suggest that BRAMS is capable to simulate severe convective events over Argentina.

�This satisfactory performance of the model encourages a thorough analysis of other events to progress in the understanding of preconditioning and triggering mechanisms of convection.

�Downscaling and data assimilation provides a set of

enriched analyses to characterize different case studies and understanding of convective related phenomena.

�Current BRAMS implementation in real time at UBA will provide further evaluation in forecast mode under a variety of synoptic conditions and using a more general configuration.

Acknowledgments: This research is supported by UBA grant X266 and ANPCyT grant Nº PICT 07 – 14420 the collaborative program IAI-CRN 55. SALLJEX was supported by PACS-SONET network, FAPESP grant 01/13816-1 (Brazil), NSF ATM0106776, NASA NAG5-9717 NOAA PID-2207021 and NA03OAR4310096.