Diapositive 1

1
Dynamical Sampling of the Mesoscale Convective
Systems by the Megha-Tropiques Instruments using
METEOSAT observations Preliminary results over
India
(from www.satmos.meteo.fr)
Rémy Roca, M. Capderou, M. Desbois, T. Fiolleau,
I. Jobard, and K. Ramage
Laboratoire de Météorologie Dynamique, Paris and
Palaiseau, France
2
Outline of the talk
1. Mesoscale Convective Systems (MCS) - Role in
the radiation budget - Observations from
geostationary satellites - Dynamical
Sampling 2. MCS over India during 15-22 July
2002 - Preliminary results for One Week in
Summer 2002 - Two case studies - a small well
organized cluster - a large continental weather
system - One week statistics of MCS and
Sampling 3. Conclusions Perspectives
3
Detection of the individual convective systems
(Houze et al., 1980 from Houze Betts1981)

• Readily observed by geostationary platforms
• Simple detection using IR threshold

4
Convective systems population during INDOEX
Whole Indian Ocean February March
Very Deep Tmin lt 210K
Deep 210Klt Tmin lt 240K

• The convective systems spans a wide spectrum of
  scales from 100 km2 to more than one million km2
• The Very Deep systems play an overwhelming role
  in the total upper level cloudiness

5
Convective systems radiative properties
LWCRF Fclear -F where F Instantaneous flux
averaged over the system cloudy flux (rather
than total) Fclear weekly average clear sky
flux over a rectangular region including the
convective system (336 images)
LWCRF exhibits strong scale dependence the two
deep convective classes contribute up to 70 to
the LWCRF
6
Dynamical Sampling
Sampling by the satellite instruments of
different regions, latitudinal bands, under
different viewing geometry has been detailed in
the previous presentation by Michel Capderou.
Dynamical sampling consists simply in adding
one more degree of freedom to the problem by
looking the sampling of the moving Mesoscale
Convective Systems along their life
cycle Preliminary results of our investigation
of such an issue using the ScaraB geometry.
7
Outline of the talk
1. Mesoscale Convective Systems (MCS) - Role in
the radiation budget - Observations from
geostationary satellites - Dynamical
Sampling 2. MCS over India during 15-22 July
2002 - Preliminary results for One Week in
Summer 2002 - Two case studies - a small well
organized cluster - a large continental weather
system - One week statistics of MCS and
Sampling 3. Conclusions Perspectives
8
MCS during the week of the 15-22 July 2002 using
METEOSAT-5

• Half-hourly full resolution IT imagery
• Thresholds of 213K and 233K
• Quite active week over the Bay of Bengal.
• Tracking is performed at LMD using the LTHE
  algorithm
• 5 classes of MCS depending upon duration only

9
Small MCS going from ocean to continent

• Tracking applied on threshold 213K
• No split-merge like 12h

METEOSAT-5 IR
17Jul-18h
17Jul-24h
18Jul-06h
18Jul-12h
104
From 17/7 21Z to 18/7 09Z
Size(km2)
102
TIME
How this system could have been sampled by MT ?
10
Small MCS going from ocean to continent

• Orbit starts at cluster initiation from 0E0N
• ScaRaB Geometry (swath 2200km)

First Orbit
11
Small MCS going from ocean to continent
Second Orbit
12
Small MCS going from ocean to continent
Third Orbit
13
Small MCS going from ocean to continent
4th Orbit
For this orbit, the system is not observed
14
Small MCS going from ocean to continent
From 18th -01Z to 18th- 09Z no observations of
the system
15
Small MCS going from ocean to continent
The growing phase is very well sampled
16
Large Continental Weather System

• Tracking applied on threshold 233K
• Two split-merge events
• Long lasting object 35h

17
Large Continental System

• One ellipse / 4 hours
• Two split-merge events
• Life 35h

e
2
s
1
104
Size(km2)
2
1
102
TIME
18
Large Continental System

• Orbit starts at cluster initiation from 0E0N
• ScaRaB Geometry (swath 2200km)

First Orbit
19
Large Continental System
20
Large Continental System
For these orbits, the system is not observed
21
Large Continental System
22
Large Continental System
23
Large Continental System
24
Large Continental System
25
Large Continental System
26
Large Continental System
27
Large Continental System
28
Large Continental System

• One ellipse / 4 hours
• Two split-merge events
• Life 35h

e
2
s
1
104
Size(km2)
2

• 8 samples
• 5 in a raw

1
102
TIME
29
Sampling statistics for the week under analysis
Number of ScaRaB obs. during One week over 5
long interval
MCS Occurrence during one week over Continental
India
Aliasing of the convective systems diurnal cycle
with the satellite sampling Can provide best case
or worst case along the precessing cycle
30
Outline of the talk
1. Mesoscale Convective Systems (MCS) - Role in
the radiation budget - Observations from
geostationary satellites - Dynamical
Sampling 2. MCS over India during 15-22 July
2002 - Preliminary results for One Week in
Summer 2002 - Two case studies - a small well
organized cluster - a large continental weather
system - One week statistics 3. Conclusions
Perspectives
31
Conclusions Perspectives
We have highlighted some cases of sampling of MCS
using MT. The preliminary results have been
focused on best cases for the ScaRaB
instrument on board MT which has the largest
swath. The preliminary results are nevertheless
very encouraging. We need to extend this
preliminary work to More cases over India
(Summer 1998-2005 to be processed
analyzed) More cases over Africa (Summer
1983-2005 to be analyzed) The MADRAS swath
too We need also to Provide more in-depth
analysis (fractional sampling of MCS
etc) Couple this MT sampling analysis to the
GPM era fleet simulations And to Think about
innovative system based level-3 products from
Megha-Tropiques