1. Nowcasting and Short-term Forecasting of Thunderstorms and Severe Weather Using OSCER
Keith A. Brewster1 Jerry Brotzge1, Kevin W. Thomas1, Jidong Gao1, Ming Xue1,2 and Yunheng Wang1 1Center for Analysis and Prediction of Storms
2School of Meteorology University of Oklahoma
Oklahoma Supercomputing Symposium October 12, 2011

2. CASA & NEXRAD Radars
CASA NetRad NSF ERC: Collaborative Adaptive Sensing of the Atmosphere
X-Band Dual-Pol Radars
40 km nominal range
Collaborative, Adaptive Scanning
Fill-in below coverage of NEXRAD
Toward phased-array panels – low-cost!
NEXRAD
S-Band Radars
14 covering domain
Data used out to 230 km

3. CASA NetRad Network
Southwest Oklahoma

4. Spring 2007-2009 Near-Real Time Forecast Domain
Dx = 1 km
53 Levels
600x540
Radars Used:
4 CASA
14 NEXRAD
Plus Satellite & Surface Data
540 km
600 km

5. CASA Forecasting Workflow
Observations
CAPS Ingest Cluster Linux Server Observation Pre-Processing
Radar Data File Selection
Mesonet & Sfc Obs Processing
Input File Generation
Operational
Forecast Model Data
CAPS Ingest Cluster
OSCER Sooner Supercomputer Analysis & Data Assimilation
Graphics on WW Web
Radar Data QC and Remapping
3DVAR Analysis
ARPS Forecast Model
Model Interpolation
CYCLE
OSCER Sooner Supercomputer Run Forecast Model
PSC Mass Store &
CAPS Linux Cluster 3D Data File Archive
ARPS Forecast Model
Graphics Generation

6. Improving the MPI Efficiency of Radar Remapper
Radar data are converted from 3-D polar to 3-D Cartesian coordinates.
Original Strategy: Horizontal Domain Decomposition
Each processor finds solution on columns within its domain
nproc_y=5
nproc_x=5
Potentially uneven workload

7. Improving the MPI Efficiency of Radar Remapper
Radar data are converted from polar to Cartesian coordinates of model grid.
Improved Algorithm
For each radar
Within domain decomposition, determine columns having valid data
Collect columns with valid data in 1-D array
Distribute work for these columns uniformly among processors
Execute remapping algorithm MPI
Distribute results to original home processor for output.
nproc_y=5
nproc_x=5

8. Real-Time NWP Runs 2009 9 Weeks in Spring Season
6-hour 1-km resolution forecasts
Use Radar Reflectivity & Radial Velocity
3DVAR wind with ADAS cloud analysis
ARPS Model
Runs posted to Web in real-time
Run on Parallel Linux Boxes OU OSCER processors/2 runs at a time
Total Run Time 1.5 hours
Two Runs in Near Real-time
CASA & NEXRAD
No CASA Data

9. 2007-2009 Assimilation Strategy
40-min Assimilation
5.5-hour Forecast
IAU
IAU
IAU
IAU
0150
0200
0210
0220
0230 03 04 05 06 07 08
Manual on-demand model start-up for storms in the network.

10. Assimilation vs. Analysis Wind Speed/Vectors 500m AGL 0220 UTC
Chickasha Radar
Analysis Only
Forecast/Assimilation

11. Forecast temperature perturbation + Vort. at z =500m AGL
End of Data Assimilation Period
0220 UTC UTC
Movie
0240 UTC UTC

12. 2010 Nowcast Strategy
2125
2130
2200
IAU
2230
2300
2-hour Forecast
5-min Assim
2330
Domain size: 350 x 320 x 53. Total Run Time < 10 min
800 cores (100 dual-quad-core servers)
Forecast model run every 10-min whenever the radars were operating (during precipitation).

13. Sample: 10 May :40
From NWS Norman

14. T=05 min (assimilated state) 2140
2140 UTC Nowcast/Forecast
T=05 min (assimilated state) 2140

15. 2140 UTC Nowcast/Forecast
T=15 min 2150

16. 2140 UTC Nowcast/Forecast
T=25 min 2200

17. 2140 UTC Nowcast/Forecast
T=35 min 2210

18. 2140 UTC Nowcast/Forecast
T=45 min 2220

19. 2140 UTC Nowcast/Forecast
T=55 min 2230

20. Data Assimilation Accomplishments
Developed a very efficient real-time data assimilation, nowcasting and forecasting system
Demonstrated initial impacts of CASA data on cloud-scale analysis and forecasting
Advanced real-time storm-scale assimilation to where we can directly compare forecasted small-scale vorticity features to radar signatures
Major step towards “warn on forecast”

21. Ongoing Work Using CASA Data
Objective Verification of recent forecasts, to also include object-based methods.
Rainfall (using QPE field from NSSL)
Vorticity Centers
Methods to improve data assimilation
Improvements to current algorithms
More sophisticated, but expensive, algorithms
Acknowledgments: NSF Sponsors CASA ERC
Computing: OU OSCER

22. In 2012 moving the radars to the Dallas/Ft Worth Metro
Metro Area - This is the recommend placement of the first four sites.
Late winter 2012 – First 4 radars
Summer 2012 – next 2 radars
Fall 2012 – Next radar
Winter – 2013 CSU
The sites include UNT and UTA buildings, which have the best connectivity so far from all proposed sites and the FAA building, which is presumed to have good network access too.
Covers the metroplex with very high resolution data,
Covers the gap above the NEXRAD radar in Fort Worth.
The north south orientation will capture storms coming from the northwest and into the metroplex, while the existing NEXRAD can give warning for storms coming from the southwest.
Provides unprecdented low level coverage of the trinity river for flooding, and accurate precipitation estimates.
Also provide coverage of rivers, lakes to the north which impacts precipitation estimates.
Very fast updates (1 Min) for rapidly evolving severe weather in the Metroplex.
Disadvantages:
CASA will provide very low coverage, but some coverage provided by NEXRAD.
This system will focus more on real time data vs. forecasts.
Which airports are we covering in addition to DFW and Love Field.
In 2012 moving the radars to the Dallas/Ft Worth Metro

23. More radars will be added during the year.
Northeast Expansion - This configuration continues coverage over highly populated areas and gets into areas where there is more of a radar gap.
More radars will be added during the year.