Thunderstorm documentation, simulation and verification over Pune

1
Thunderstorm documentation, simulation and
verification over Pune

• P. Mukhopadhyay
• Forecasting Research division
• Indian Institute of Tropical Meteorology
• Pune-411008
• Emailmpartha_at_tropmet.res.in

2
Outline of talk

• Motivation for the study
• Documenting few cases
• Simulation with a mesoscale model
• Verification of the result
• Conclusion

3

• Thunderstorms are localized (Pune) mesoscale
  systems. Is it possible to document the life
  cycle of these events in the absence of any meso
  network?
• Whether such mesoscale systems can be simulated
  using numerical model?
• To what level of accuracy? (validation)

4

• Thunderstorms Events over Pune (73.85oE,
  18.53oN)
• 9 June 2003
• 10 June 2003
• Synoptic conditions of 9 June 2003
• According to IMDs Indian Daily Weather Report
  (IDWR) There was an off shore trough from
  Maharashtra to North Kerala on 8 June which
  further extends from Gujarat to Kerala coast on 9
  June. The trough had a vertical extension between
  3.1 to 4.5 km above sea level (a. s. l.).
• There was a cyclonic circulation over
  Pakistan and adjoining Jammu and Kashmir with
  vertical extension upto 4.5 km a.s.l. The surface
  chart of 9 June shows the west coast trough and
  the low pressure area over Pakistan very clearly.

5
(No Transcript)
6

• Synoptic conditions of 10 June 2003
• According to IMDs Indian Daily Weather Report
  (IDWR)
• The persistence of off shore trough from
  Gujarat to Kerala coast.
• One fresh cyclonic circulation is reported to
  have formed off Karnataka-Maharashtra coast
  between 2.1 and 4.5 km.

7
(No Transcript)
8

• Contrasting rainfall of the events
• For 9 June
• (IMD)s Agricultural Observatory ( CAgMO) site
  57.8 mm and at IMDs Central Training Institute
  (CTI) observatory at Pashan ( 6 km from IMD
  CAgMO site) was 65 mm
• For 10 June
• The rainfall recorded at IMD CAgMO site is 2.6 mm
  where as that recorded in CTI, Pashan was 24.2 mm
  
• Rainfall for both the events lasted for an hour
  1700-1800 IST

9

• a. 0830Z 090603 WV b.10 Z 090603 IR c. 11 Z
  090603 VIS d. 1230 Z 090603 VIS

e. 13 Z 090603 VIS
f. 14 Z 090603 IR
g. 15 Z 090603 IR
h. 1730 Z 090603 WV
10

• a. 0730Z 100603 WV b. 09Z 100603 VIS c.10Z
  100603 VIS d. 11Z 100603 VIS

f. 14Z 100603 IR
g.15Z 100603 IR
h. 1730 Z 100603 WV
e.1230Z 100603 VIS
11

• Remarks
• METEOSAT-5 Satellite images do capture different
  stages (genesis, maturity, dissipation) of the
  two localized thunderstorm events
• However it does not provide information about
  the internal structure of the storm
• As a consequence it is hard to diagnose the
  cause of contrasting rainfall by the two events
  just by satellite imageries

12
Simulation of 9 June 2003 event

• Model Regional Atmospheric Modelling System
  (RAMS) Cotton et al., 2003 Pielke et al.
  1992
• (www.atmet.com latest version RAMSv6.0b2)
• Some features of RAMS
• Nonhydrostatic, compressible
• Terrain following height coordinate Gal-chen
  Somerville, 1975 Clark, 1977
• Cumulus parameterization Kuo, KF
• Radiation Harrington, 1997
• Cloud microphysics Walko et al. 1995
• Land surface model- LEAF2 Walko et al. 2000
• Initialization Horizontally homogeneous, gridded
  anl

13

• Equations of motion
• Thermodynamic equation
• Water species mixing ratio continuity equation
• Mass continuity equation

14
Model set up for the simulation

• Nos. of Domains 2 outer domain hor. Res. 16 km
• inner
  domain hor. Res. 4km
• Nos. of Grid points- 68 x 68 x 36 outer domain
• 58 x 58 x 36
  inner domain
• Stretched vertical levels- 36 levels at
  1.11002000 (first level will have a thickness
  of 100 m, the next level will be of thickness
  110m and so on till the level reaches the
  thickness of 2000m. When
• thickness will be 2000 m, it will remain fixed
  for the rest of the levels.)
• Domain centered at- 73.85oE, 18.53oN
• Cu parameterization- kuo scheme
• Explicit microphysics prognostic equation of
  rain,
• pristine ice, snow, aggregates, graupel and hail.

15

• Data Used
• 6 hourly NCEP daily analyses
• RS/RW of Mumbai (76.3oE, 9.12oN)
• USGS topography and vegetation at 30 arc sec
  resolution
• Simulation experiments
• Exp-1
• 00 UTC 9 June 2003 initial condition
• Initial nudging 6 hr
• Forecast for 12 hr from 06 to 18 UTC
• Exp-2
• RS/RW data merged with NCEP anl and a mesoscale
  reanalyses at 16 km is made
• Remaining steps are same as previous expt-1.

16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
Hourly Total cloud condensate forecast Exp-1,
16-km domain
20
Hourly Total cloud condensate forecast Exp-2
21
Hourly Total cloud condensate forecast Exp-1, 4
-km domain
22
Hourly Total cloud condensate forecast Exp-2, 4
-km domain
23
Ver. Velocity at 850 hPa, Exp-1, 16-km domain
24
Ver. Velocity at 850 hPa, Exp-2, 16-km domain
25
Ver. Velocity at 850 hPa, Exp-1, 4-km domain
26
Ver. Velocity at 850 hPa, Exp-2, 4-km domain
27
(No Transcript)
28
Vert Vel (Red) at 850 hPa, Pcprate (blue) and
Temp in deg cent (Green) at a grid close to Pune
lat-lon
29
Forecast rain (left panel), CPC daily accumulated
rain (mm) right panel
30
Concluding remarks

• The life cycle of localized thunderstorms of 9
  and 10 June 2003 over Pune is reasonably well
  captured by the hourly Meteosat imageries.
• Mesoscale model RAMS could simulate the storm of
  9 June with weaker intensity as compared to
  observation
• To document further details of a localized
  thunderstorm and to prepare a better input for
  mesoscale simulation and for model verification,
  data from remote sensing devices such as WP/RASS
  will be highly beneficial

31
Thank You