Transcript Use of a Mesoscale Ensemble Weather Predictions to Improve

Integration of a Mesoscale
Ensemble System into National
Weather Service HydroMeteorological Operations
Operational Aspects of the
COMET Collaborative Project
Participants
• Stony Brook University:
– Brian Colle, Mike Erickson
• NWS New York City:
– Jeff Tongue, Nancy Furbush, Matt Sardi, Nelson Vaz,
Adrienne Leptich
• NWS Philadelphia/Mount Holly:
– Al Cope
• NWS Eastern Region Headquarters
– Josh Watson
• NWS MARFC:
– Joe Ostrowski
Objective
Integrate data from the Stony Brook
University Mesoscale Ensemble System
(MES)
1. Into AWIPS.
2. Into the Site Specific Hydrologic Predictor
(SSHP) application.
Why ?
• Allow forecasters to perform mesoscale
stochastic hydrology on small, fast
responding basins.
How??
• Provide WFO Forecasters with Mesoscale
Ensemble Data to assimilate via AWIPS
D2D (GFE) and the Web.
• Provide WFO Forecasters with Mesoscale
Ensemble Data to integrate into the Site
Specific Hydrologic Predictor.
Data
• Web
• AWIPS
• GFE
Web
http://chaos.msrc.sunysb.edu/
NEUS/nwp_graphics.html
http://chaos.msrc.sunysb.edu/NEUS/
nwp_graphics.html
Precip Plumes
AWIPS
•
•
•
•
00Z 12/36 km 14 Member Ensemble data
Available ~11 UTC
36 Hour Forecast
Volume Browser:
– Members
– Means/Standard Deviation
– Probability of Exceedence
AWIPS
AWIPS – QPF Members
AWIPS - Mean
AWIPS – Temperature Members
AWIPS – Wind Members
AWIPS - Concerns
•
•
•
•
Installation Instructions are Available.
Still needs some “clean-up.”
New “build” issues.
Stony Brook is not an Operational Center.
SSHP
Integrating Stony Brook ESP
into SSHP
1.Deterministically: From the
AWIPS GFE
2.Stochastically: From the
Stony Brook EPS
Accessing SSHP
1. GFE QPF - Deterministic
Convert 6 hour NDFD QPF to RR2 text product for use with SSHP
Note that GFE 6-hr Pcpn is evenly distributed
In SSHP
Basins:
–
–
–
–
–
–
–
Lodi - Saddle R.
Gardiner – Wallkill R.
HoHoKus – Hohokus Brook
Springfield – Rahway R.
Beacon Falls – Naugatuck R.
Southbury – Pomperaug R.
Yantic – Yantic R.
2. EPS into SSHP (under development)
• Experimenting with the Lodi Basin.
• Stony Brook University is producing hourly
resolution basin averaged precipitation.
– RR2 Product (ftp’ed):
• Mean
• Worst Case scenario
• Individual member data
Ensemble mean
accumulated 1 hr
forecast for hour 32
Ensemble mean for the
Lodi Basin
GFE QPF
SBU
Ensemble
Members
Mean Average
Worst case
Selecting Basin
QPF in SSHP
Example Worse
Case in SSHP
Future Work
• Automate the FTP of Data
• Train Staff
• Modify SSHP with help of OH to display
ensemble hydrographs
The Future in SSHP