HPVCIQuarterly Meeting

1
HPVCIQuarterly Meeting

• Duanjun Lu
• Dr. Paul Croft Remata S. Reddy
• HPVCI JSU MP

2
Activities since last review

• Order the NOGAPS data for three hurricane cases,
  Allison, Dennis and Floyd.
• Dennis 08/2200/1999 -----
  09/0712/1999.
• Floyd 09/0800/1999 -----
  09/1812/1999.
• Allison 06/0200/2001 -----
  06/1812/2001.
• Created a program which convert GRIB format of
  ordered NOGAPS data into the format which COAMPS
  model required.

3
Activities review ---- contd

• Run hurricane Dennis and Floyd on SGI O2100 and
  SGI Onyx.
• Audited Scientific Visualization class hosted by
  computer department.
• Collaboratively work with MSU for demonstration.
  A program was made to convert COAMPS model output
  into the format which MetVR can read.
• An oral presentation for HPVCI summer seminar was
  given in JSU
• An post presentation was given on the AMS 18th
  Conference on Weather Analysis and Forecasting
  /14th Conference on Numerical Weather Prediction.
• "Data Assimilation and Model Evaluation
  for MM5 and COAMPS"
• Interviewed five applicants for the position of
  graduate assistant.

4
Simulations of two hurricane cases using COAMPS

• The features of hurricanes
• o long life ( Dennis15 days Floyd11day)
• o swipe wide area.(Dennis2,200km
  Floyd3,300km)
• Simulation design
• o high resolution spacailly and temporally
• o large number grids of domain (Dennis200
  Floyd300)
• o big memory requirement (210320Mw )
• o memory limit (160Mw)
• o no moveable domain option available for
  coamps

5
(No Transcript)
6
Simulations of two hurricane ---- contd

• simulation design --- contd
• o solution "manually move".
• 4 domain schemes for Dennis and 7
• schemes for Floyd.
• 12112130 nested two domains
• with 27km and 9km grid spacing for
  
• outer and inner domain.

7
(No Transcript)
8
Simulations of two hurricane ---- contd

• Validation of hurricane intensity and track.
  
• o Track error (RMSE137km)
• o Minimum sea level pressure (RMSE
  16mb)
• o Maximum surface wind speed
  (RMSE28knots)
• o Visualization of Dennis (preview of
  CAVE demo).

9
The center track of Dennis and Floyd

• O observation
  
• Dennis

• Simulation
• Floyd

10
Minimum sea level pressure
11
Maximum surface wind speed
12
A few problems

• Satellite and Radar data availability, and format
  for real-time operational model runs.
• NOAA/NESDIS, Data source assimilation, GIS/RS
  possibilities.
• Tcl-sh script doesnt work on SGI O2100.
• Staff of one.

13
Future plan

• Further analyze the results of two hurricanes.
• Finalize the preprint of NWA and AMS meeting.
• Collaboratively work on the proposal of COMET.
• Point-to-point comparisons between the
  simulations of COAMPS/MM5 and observations.
• Gempak software will be installed in the lab with
  the assistance of graduate student.
• LDM will be setup on SGI O2100 so that real-time
  observations can be assimilated by operational
  COAMPS and MM5.