The Collaborative Effort Between Stony Brook University and the National Weather Service, Part 1 - Previous Results, Current Status and Future Plans

The Collaborative Effort Between Stony Brook University The Collaborative Effort Between Stony Brook University and the National Weather Service, Part 1 - Previous and the National Weather Service, Part 1 - Previous

Results, Current Status and Future PlansResults, Current Status and Future Plans

Dr. Brian A. Colle, Matthew JonesDr. Brian A. Colle, Matthew Jones

Institute for Terrestrial and Planetary SciencesInstitute for Terrestrial and Planetary Sciences

Stony Brook University / SUNYStony Brook University / SUNY

Jeffrey S. TongueJeffrey S. Tongue

NOAA/NWS Upton, NYNOAA/NWS Upton, NY

Learn about the benefits and weaknesses of high resolution versus ensemble modeling over the Northeast U.S.

Complete the first detailed multi-season verification over the Eastern U.S (see Colle et al. 2003a,b WAF).

Use the mesoscale model to understand various coastal weather phenomena (landfalling storms, sea breezes, convection).

Use the mesoscale model with other applications (wave modeling, storm surge, energy load forecasting, hydrological modeling).

For classroom education and forecasting.

Motivation

SUNY-SB Realtime MM5 Domains

Floyd (1999) Extratropical Transition

12-km MM5 at 2100 UTC 16 Sept 1999 (21 h)

10

12

18

22

26

oC

L

WSR-88D: 21 UTC 9/16/99

1.33 km MM5 (21 h)

See Colle (2003, MWR December)

SBU Storm Surge modeling around NYC using POM

Storm surge animation of Floyd (1999)

SBU Wave-modeling using MM5 and WW3 (48 h): Isabel (9/19/00Z)

L

PBL and wave effects on 30 December 2000 Nor-easter (21 UTC, h33)

L 990

CTL (MRF)

B-T PBL Eta PBL

MRF/WW3

Problem: Sea breezes move inland too fast

(At ISP, 50% early, 38% on time, and 16% late)

MM5 wind direction mean absolute errors (36 – 12 km)

Data sampling over Long Island Sound using the Port Jefferson to Bridgeport Ferry: http://www.stonybrook.edu/soundscience/

ObservedWinds

ModelWinds

17:00 GMT June 23rd, 2003

Port Jefferson 17 GMT departure

Bridgeport 18:10 GMT arrival N

360 0

= 1 m/s

23 JUN 1700 Z

Port Jefferson Bridgeport

Port Jefferson Bridgeport

48-h mean errors (at 44004) during the 2002-2003 cool season

SLP

TEMP

More negative slp errors More positive slp errors

NOV

JAN

DEC

FEB

24-h Energy Load Forecasts (SW Suffolk Co.) using 12-36 h MM5

36-km MM5 6-36 h Precip total for physics ensemble of Floyd (1999)

Development and Evaluation of a 18-Member Real-time Ensemble Forecast

System

A COMET collaboration betweenStony Brook University (SUNY)

Brian Colle and Matt JonesNOAA/National Weather Service offices (Upton, NY, Mt Holly, NJ, and Taunton,

MA), NE River Forecast Center, and Eastern Region Headquarters

NWS Vision 2005 Goals:“NWS will move towards adding probabilistic forecast products!”

Move from “subjective” model – to –model comparison to the use of more objective ensemble prediction systems

American Meteorological Society (AMS) Statement: “The AMS endorses probability forecasts and recommends their use be substantially increased.”

“Would allow user (not forecaster!) to make decisions based on quantified uncertainties with resulting economic and social benefits” ( e.g.,from taking umbrella along, to canceling a trip, to staging snow removal equipment, etc., etc.)

Work in Progess and Future Goals Objectively evaluate18-member ensemble, which became operational on 1 May 2003.

Provide more ensemble data to our forecast partners (AWIPS, etc…). Feedback from forecasters is critical!!

Add more ensemble members (Big question: How to construct an ensemble?) and graphics.

Use ensembles in various applications (hydrological modeling--RFC, storm surge and wave modeling, energy load forecasting).

Development of MM5 MOS and real-time verification page.

THE END