Diapositive 1

1
SWICE South-West Indian ocean tropical
Cyclone Experiment
Frank ROUX (Laboratoire dAérologie) Matthieu PLU
Philippe CAROFF (Météo-France La Réunion)
2
Tropical Cyclones in the SWIO

• RSMC La Réunion area of responsibility
  0-40S , W of 90E

• Annually
• 8-15 Trop Storms (gt17 m/s)
• 1-9 Trop Cyclones (gt33 m/s)
• 0-3 Severe TCs (gt65 m/s)

3
Track Forecast
The TC track forecasts have regularly
improved in the past decades the 48-h forecasts
are now as good as the 24-h ones 10 years ago.
However, the errors remain too large even at
short term (gt100 km - roughly the diameter of the
TC core or the size of La Réunion - at 24-h) to
provide enough accuracy and guidance for
emergency management. Unlike track
prediction, intensity (i.e. wind rainfall)
forecast has shown little improvement
4
Intensity Analysis Forecast (1)

•  Dvorak Method 
• Current Intensity number (CI) is deduced from
  characteristic features in VIS IR images from
  geostationary satellites
• Wind (radius, height) Vmax(CI) x F(radius) x
  G(height)
• !! Vmax(CI) , F(radius) , G(height) calibrated
  validated over N Atlantic only (USAF NOAA
  missions)
• !! Uncertainties increase for
• rapid evolution cyclogenesis, eyewall
  replacement cycle, landfall,
• specific situations small / large storms,
  tropical / extra-tropical transitions,

5
Intensity Analysis Forecast (2)

• Microwave radiometers (AQUA, Coriolis, DMSP,
  TRMM, METOP, )
• Images at 37 (?h30 km) 85 GHz (?h10 km)
• identification of eyewall, external bands,
• structural changes ? Intensity changes
• estimate of rainfall (37 GHz rain, 85 GHz
  large ice)
• 4-6 images / day
• time resolution is slightly too large to
  correctly follow rapid changes

6
Intensity Analysis Forecast (3)

• Microwave sounders (NOAA series)
• AMSU-A (B)
• ?T in the eye ? Pmin (hydrostatic) , Vmax
  (thermal wind)
• But oblique sampling, poor space and time
  resolution, effect of clouds precipitations,
  validation,

7
RSMC Analysis Forecast
Satellite-Based TC Intensity and Size Estimates
Human expertise and (expert system)
8
SWICE proposed Field Experiment (1)

• South West Indian ocean tropical Cyclone
  Experiment
• January February 2011, from La Réunion
• Scientific Objectives
• To validate the empirical Dovrak laws VT ( CI ,
  r , z ) over the SWIO
• To participate to the Calibration-Validation of
  Megha-Tropiques satellite products
• To quantify the impact of  targeted
  observations  on cyclogenesis, track and
  intensity forecasts
• To analyze the influence of synoptic
  perturbations on cyclogenesis, storm intensity
  and track tropical waves (K, ER, MRG), MJO,
  upper-level perturbations, subtropical H L,
• To characterize the microphysical (? radiative)
  properties of iced hydrometeors in the upper
  tropical troposphere

9
SWICE proposed Field Experiment (2)
Jan-Feb 2011 ?
10
SWICE proposed Field Experiment (3)

• RADIOSOUNDING NETWORK
• (2 months at 00 12 UTC 4 / day during IOPs)
• 4 active upper-air stations Dar-es-Salaam
  (Tanzania) 6.52S, 39.12E Mahé
  (Seychelles) 4.41S, 55.32E
• Antananrivo (Madagascar) 18.48S, 47.19E
• Plaisance (Mauritius) 20.26S, 57.40E
• Dar es Salam (Tanzania), ? Diego Garcia
  (BIOT), Durban (South Africa)
• 2-3 additional stations in the intermediate zone
  
• Mozambique channel Dzaoudzi (Mayotte) 12.48S,
  45.17E Main Development Zone Agalega
  (Mauritius) 10.26S, 56.45E Tromelin
  (France) 15.53S, 54.31E
  St-Brandon (Mauritius) 16.27S, 59.37E.
• Southern latidudes Tolonaro (Madag.) 25S, 45E
  

11
SWICE proposed Field Experiment (4)

• FR / Falcon-20 aircraft ( based in La Réunion or
  Mayotte other airports )
• (800 km/h, 3h30, 10 km) 10-15 dropwindsondes
  released every 100-200 km in the most sensitive
  regions around TCs or convective zones
• Microphysical probes (counters imagers)
  characteristics of ice particles
• Onboard O3, NO2, CO measurements troposphere
  stratosphere, mid latitudes tropics,
  continental oceanic air masses interactions
• RASTA 95 GHz cloud Doppler radar
• ? Electric field mills Electric field,
  lightnings
• synoptic surveillance vs. satellite
  validation missions

12
SWICE proposed Field Experiment (5)

• CNES-ISRO MEGHA-TROPIQUES (to be launched in
  April 2010 by India)
• 870 km, 20, ?2000 km swath? up to 5 overpaths
  per 24h (at 15N 15S)
• SCARAB SW LW radiative fluxes
• SAPHIR 183 GHz sounder ? humidity profiles
• MADRAS 18-157 GHz scanning radiometer ? wind,
  water vapour, liquid water, precipitation
  (liquid ice)

13
SWICE validation of MEGHA-TROPIQUES

• Oceanic tropical environment ( 2 others in
  Brasil Niger)
• 1/3 à 1/4 of the Falcon-20 flights colocated
  with M-T track (already done during AMMA / SOP-3
  for A-Train)
• Temperature Humidity soundings dropsondes
  vs. SAPHIR
• Microphysical characteristics of hydrometeors in
  the upper oceanic tropical troposphere
  Microphysical probes RALI vs. TBs (MADRAS)
• Eventually 3D reflectivity fields from the
  Meteo-France S- band Doppler radar in La Réunion
  vs. TBs rain
• Cost 1/4 à 1/3 of the Falcon-20 cost 150 -
  200 k

14
SWICE proposed Field Experiment (6)

• HIGH RESOLUTION MODELING in La Réunion
• Operational ECMWF T-799 (25 km) and
  ALADIN-Réunion at 10-km horizontal resolution
  with parameterized convection
• Research Meso-NH at, e.g., 2 / 8 / 32 km in
  nested domains coupling with the Ocean Mixed
  Layer
• ?? Future operational AROME-Réunion at 3-km
  with full physics ??

070224-00 ? 070224-19
15
SWICE proposed Field Experiment (7)

• ADAPTATIVE SAMPLING
• EUCOS EU IP / PREVIEW Data Targeting System
  experiments in 2008 (ECMWF UKMO, to be extended
  to Météo-France models)
• identify potential high-impact weather events,
  in particular cases with large uncertainty
• predictions of sensitive regions where
  additional observations are likely to have most
  impact in reducing the forecast uncertainty
• identify which additional observations to
  request
• metrics to assess the impact of changes in
  observational network
• Techniques Adjoint (Singular Vectors, Kalman
  Filter Sensitivity, ) Ensemble (Ensemble
  Transform Mono Observation, ET Kalman filter,
  )
• Open Questions non-linearity, horizontal
  resolution, humidity water cycle, ensemble
  members, numerical cost,
• Now optimized for Europe N Atlantic , HYMEX
  in 2011,

16
SWICE related Field Experiments (1)

• ATLANTIC OCEAN (incl. CARRIBBEAN SEA GULF OF
  MEXICO)
• Observations in TCs with airborne Doppler radar
  and onboard instruments, and around TCs with
  dropwindsondes started with P3s in 1982 by NOAA /
  HRD synoptic surveillance with Gulfstream
  IV-SP jet since 1997
• Burpee et al. (1996) 15-30 error reduction for
  12-60 hr track forecast, 20 reduction for
  intensity forecast, by incorporating these data
  in the initial analysis. But the effectiveness of
  adaptative observations is highly variable !
• NW PACIFIC OCEAN
• Dropwindsonde Observations for Typhoon
  Surveillance near the Taiwan Region (DOTSTAR,
  2003, 2004, ) with ASTRA SPX jet (750 km/h, 6
  h, 14 km) sondes released every 150-200 km in
  the most sensitive regions around TCs ? improved
  diagnosis, analysis forecasts BL wind
  measurements
• THORPEX Pacific-Asian Regional Campaign (T-PARC,
  2008) with enhanced RS network, driftsondes,
  dropsondes, instrumented aircraft, genesis
  and evolution of TCs, improved use of cloud-
  rain-affected satellite data, advanced
  assimilation techniques, adaptative sampling,
  high-resolution modeling,

17
SWICE related Field Experiments (2)
DOTSAR  Typhoon Initialization in a Mesoscale
Model Combination of the Bogused Vortex and the
Dropwindesonde data  K.-H. Chou C.-C. Wu (MWR
Mar08)

• BNDY the DLM winds agree much better with the
  observed values when the DWS data are assimilated
• BYDN the implantation of the Bogus Vortex
  intensifies the inner core structure
• BYDY better analyzed outer circulation

18
SWICE proposed Field Experiment (8)

• AEROCLIPPERs
• 5-10 m balloons, vertically stabilized at 60 m
  altitude, with an attached guide rope floating
  at the ocean surface ? GPS position wind, SST
  SSS, atmospheric pressure, temperature humidity
  bulk momentum, heat moisture fluxes at the
  air-sea interface
• 2007 two Aeroclippers survived after passing
  through TC Dora in Jan-Feb 07 during the VASCO
  campaign in the SWIO (! wind position only !)
• ? Coordination with TRIO?

19
SWICE proposed Field Experiment (9)
20
SWICE proposed Field Experiment (10)

• OPERATION CENTER
• RSMC Météo-France La Réunion responsibility for
  Mayotte weather center
• Operational forecasting capability,
  telecommunications facilities
• International Roland-Garros airport French air
  force base

• PARTICIPATING LABORATORIES
• LACY (Université de la Réunion CNRS
  Météo-France) RS network, forecast
• Laboratoire dAérologie (Toulouse) Dropsondes
  In situ chemistry
• CNRM (Toulouse) Targeting
• LAMP (Clermont-Ferrand) Microphysics
• LATMOS (Vélizy) MEGHA-TROPIQUES RASTA radar
• LMD (Paris) MEGHA-TROPIQUES Upper
  tropospheric features Aeroclippers
• LOCEAN (Paris) Ocean (TRIO)
• Indian Ocean Commission
• WMO / Tropical Meteorology Research Program /
  Tropical Cyclone Panel
• World Weather Research Program / THORPEX
• USA / NOAA Annual Hurr. Field Prog, Taiwan /
  DOTSTAR, Australia, India

21
SWICE proposed Field Experiment (11)
22
SWICE proposed Field Experiment (12)
Examples of virtual flight plans for successive
synoptic surveillance missions with Falcon-20
for TC Ivan on 13-19 February 2008. Each flight
will be about 3h30 min long, and 10-15 dropsondes
will be released. (flight plans must first be
approved by the pilots and by the Air Traffic
Control)
23
SWICE proposed Field Experiment

• Distinct maximum in January and February
• Relatively large inter-annual variability
• In a 2-month period, one can generally expect 1-2
  (up to 3-4) TSs and/or TCs in the area of
  investigation

24
SWICE proposed Field Experiment

• Estimated total cost 1 M
• radiosoundings 300 k,
• Falcon-20 opérations 350 k,
• dropsondes 250 k,
• shipping and fees 20 k,
• travel stay 80 k
• Not included
• new RS stations and cost of operations,
• Aeroclippers (TRIO)
• Pending demands
• ANR 300 k
• Région Réunion 100 k
• CNES 200 k
• Météo-France 200 k
• CNRS / INSU 200 k