Tropical Analysis and Forecast Branch (TAFB) Overview and Challenges
Tropical Analysis and Forecast Branch (TAFB) Overview and Challenges
Eric Blake
3-17-04
Eric Blake
3-17-04
TPC/WFO MFL OCT 4 , 2000TPC/WFO MFL OCT 4 , 2000
Florida Turnpike
Florida Int’l University
OUTLINE
• Brief TAFB duties and products overview
• Data and Model Sources
• Forecast Challenges
• Possible TC graphic and verification
• Future
TPC/NHCTPC/NHC
• Hurricane Specialists Unit • Tropical Analysis and Forecast
Branch (TAFB) • Technical Support Branch (TSB)
– Storm Surge Unit• Chief Aerial Reconnaissance
Coordination All Hurricanes (CARCAH)
TPC Employees Grouped by Branch
37%
27%
18%
11%
7% TAFB
TSB/Storm Surge
Specialists
Administrative
CARCAH
TAFB provides products which include marine forecasts and warnings, surface analyses, tropical weather discussions, and backup aviation forecasts and warnings (SIGMETs). The unit also issues satellite imagery interpretation and satellite rainfall estimates for the international community.
TAFB provides operational tropical cyclone support during the tropical seasons through manpower, forecast and media support, radar tracking, and tropical cyclone position and intensity estimates based on the Dvorak technique.
Tropical Analysis and Forecast Branch (TAFB)Tropical Analysis and
Forecast Branch (TAFB)
TAFB DutiesTAFB Duties
• Issue marine forecasts (text and graphical) for the tropical and sub-tropical Atlantic and Northeast and Southeast Pacific Oceans.
• Produce 3-hourly tropical surface analysis
• Tropical Weather Discussions• Estimate intensity and position of
tropical cyclones using the Dvorak technique
• Subtropical cyclone intensity estimates using Hebert/Poteat Technique
• Provide Satellite Rainfall Estimates • WSR-88D radar fixes for land-falling
tropical cyclones• Media support during tropical cyclone
events • Tropical cyclone forecast support• FEMA -Hurricane Liaison Team (HLT)
TAFB Duties TAFB Duties
TAFB Production Suite • Atlantic Desk
– Marine Weather Discussion (2)– High Seas - HSFAT2 (4) – Offshore Waters
• OFFNT3 (4) • OFFNT4 (4)
– Seastate Analysis (4)
– ATLC Wind/Wave Charts (8)
• Analyst/Forecaster– Tropical SFC Analysis (8)– TWDAT (4) – Wave Period/Swell (6)– PANAM Temps (2)
• Pacific Desk – Dvorak Classifications – (4 Synoptic/Intermediate)– TWDEP (4) – High Seas
• HSFEP2 (4)
• HSFEP3 (4)
– High Wind/Danger Area • Atlantic
• Pacific
– EPAC Wind/Wave Charts (11)
TAFB High Seas Forecast AreasTAFB High Seas Forecast Areas
Offshore Forecast AreaOffshore Forecast Area
Atlantic Surface Analysis
Atlantic Wind/Wave Forecast
Atlantic Wind/Wave Forecast
Atlantic Tropical Cyclone Hazard Area Graphic
1 June - 30 Nov
High Wind Area Graphic
• 12 Planet coordination with WFO’s
• 3 hourly surface analyses over tropical N. American region.
What’s new in 2004Products
What’s new in 2004Products
• Sparse data - need for remote sensed data• Reliable guidance over the tropics (or lack
thereof)• Local effects (sea-breeze, topography, diurnal)• Small amplitude/scale associated with many
significant weather features• Reliability of ship observations (not quite as
good as land stations)
Challenges of analyzing and forecasting tropical weatherChallenges of analyzing and forecasting tropical weather
Typical 24 hour Composite of Atlantic Ship and Buoy Observations
• Numerical Guidance (GFS,ETA,NWW3,UKMET,NOGAPS, GFDL/GFDN, CMC, Ensemble probabilities)
• Surface Observations (metar, ship, buoys)
• Satellite
• Radar including Caribbean countries
• Upper Air Observations
• Weather reconnaissance data
Conventional DataConventional Data
Ensemble Probabilities
Spaghetti Diagram
Conventional Data Often is Not Enough
• TAFB relies on additional sources of data such as microwave and remote-sensing data in order to gain a better understanding of the current/forecast situation, especially in data sparse regions
• Satellite Derived Winds
• Quikscat
Additional Data used by TAFBAdditional Data used by TAFB
Quikscat
Quikscat
• Satellite Derived Winds
• Quikscat
• SSMI
Additional Data used by TAFBAdditional Data used by TAFB
SSMISSMI
• Satellite Derived Winds
• Quikscat
• SSMI
• ERS-2 Data
Additional Data used by TAFBAdditional Data used by TAFB
Forecast Challenges
• Wind “funnelling” through passages
• NWW3 biases
• Central Caribbean Sea strong trades
YUCATAN CHANNEL
FLORIDA STRAITSBAHAMA BANK
WINDWARD PASSAGE
MONA PASSAGE
ANEGADA PASSAGE
GULF OF HONDURAS
Gulf of Mexico Examples
• Recent examples of NWW3 under-forecasts along frontal boundary in the Gulf of Mexico and Western Atlantic.
• Next few slides show examples of ship and buoy observations (red) with the NWW3, significant wave height forecasts (black) valid at the time of the observations.
• Most observations in Gulf of Mexico and extreme Western Atlantic are moored buoys and are considered extremely reliable.
NWWIII under-forecasting seas behind
front in NW Gulf of Mexico6 h forecast
Similar example:9 h forecast
Example of an over-forecast of seas in central Gulf of Mexico well behind
the front.24 h forecast
NOAA Wavewatch III Biases• Does not develop seas fast enough behind strong
cold fronts, particularly in the Gulf of Mexico and W Atlantic, and is often slow in subsiding the seas beyond 24 hours after frontal passage.
• This bias likely due to improper handing of initial growth of seas, as well as under-forecasting of winds by AVN model.
• Under-forecast of seas in Caribbean during strong trade wind events (also largely due to under-forecasting of winds.
• Occasionally over-forecasts swell heights in TAFB forecast areas.
GFS 10-m wind forecast depictingfunneling winds in the Yucatan Channel
1135 UTC 29 November 2003 QUIKSCAT PASS depicting higher winds in Yucatan Channel
2319 UTC 11 November 2003 QUIKSCAT PASS depicting higher winds in Florida Straits
NWW3 wave height forecast depicting lower seas in the Bahama Bank
GOMEX Shelf Water Event25 January 2004 confirmed by
Quikscat
• This event was similar to the southerly Gulf Stream events off the mid-Atlantic and New England waters. The mid-Atlantic and New England events were documented by OPC and are characterized by warm southerly winds which do not mix down to the surface over the cool waters west of the Gulf Stream.
• Between about 0000 and 1200 UTC 25 January strong southerly winds occurred over portions of the western Gulf of Mexico.
• Offshore buoy over the warmer Gulf waters reported 28 to 33 kt sustained winds with gusts to 39 kt while winds over the cooler shelf waters were around 15 kt with gusts to 20 kt.
• Quikscat from 0050 UTC detected the weaker winds near shore. The gradient of weaker winds matched well with SST data from buoys and satellite.
Warm “Tongue”
Cold Shelf Waters
Quikscat from 0050 UTC 25 January. Surface observations from 0400 UTCSSTs in Red
Coordination
• Crucial part of operations due to overlap of OFF and CWF areas because of NAVTEX products.
• Means of Coordination– Hurricane Hotline - Large Synoptic scale systems
affecting multiple offices– Telephone call to individual offices– Marine Weather Discussion
• TAFB also coordinates with OPC and WFO Honolulu/CPHC.
Hurricane Isabel CoordinationEAST CENTRAL FLORIDA FORECAST DISCUSSION
NATIONAL WEATHER SERVICE MELBOURNE FL
3:30 AM EDT SAT SEP 13 2003
MARINE… COORDINATION CALL FROM TAFB INDICATED THAT INCREASING SWELL HEIGHTS FROM ISABEL COULD BE ARRIVING IN OUR OFFFSHORE WATERS… BEYOND 60 MILES OF SHORE… AS EARLY AS SUNDAY NIGHT OR MONDAY MORNING. HAVE MADE TIMING AND WAVE HEIGHTS ADJUSTMENTS TO THE COASTAL WATERS FORECAST. WILL MOST LIKELY NEED TO POST SMALL CRAFT EXERCISE CAUTION SUNDAY AND SMALL CRAFT ADVISORIES FOR SEAS BY MONDAY.
AREA FORECAST DISCUSSION
NATIONAL WEATHER SERVICE JACKSONVILLE FL
3:20 AM EDT SAT SEP 13 2003
MARINE…OBS INDICATING LONG PERIOD SWELLS ATTRIBUTED TO HURRICANE ISABEL WITH PERIODICITY ~ 14 SECONDS. WILL KEEP SEAS 3 TO 5 FEET THROUGH SUNDAY. OFFSHORE WATERS WILL LIKELY REACH SCEC CONDITIONS BY SUNDAY NIGHT DUE TO COMBINED SEAS…FOLLOWED BY SCA CONDITIONS ON MONDAY. THANKS FOR THE COORDINATION CALL TAFB.
Introduction of the term“Possible” Tropical Cyclone
Introduction of the term“Possible” Tropical Cyclone
• TAFB realized a better service could be provided to mariners by giving a “heads-up” to “possible” TC development.
• TAFB began experimenting with using the term “Possible” Tropical Cyclone in some graphic and text products during the 2001 tropical season.
“Possible” Tropical Cyclone Forecasts
• TAFB forecasters and duty hurricane specialist (HS) collaborate on areas of possible TC development
• HS and TAFB forecasters use satellite analysis (Dvorak) to determine the current strength and structure of a system
• Use numerical models to determine its likelihood of future development
“Possible” Tropical Cyclone Forecasts
• If it is determined that TC development is “possible” within 36 hours, the HS and TAFB forecaster will agree on the shape and size of the “possible” TC development area on the danger chart.
• TAFB forecaster will make sure that all subsequent text and graphic charts issued by TAFB indicate “possible” TC development.
NHC Tropical Weather Outlook530 PM EDT (2130 UTC) Sat 07 Sep 2002
A large area of cloudiness and showersextend from the Bahamas northeastwardfor several hundred miles. This activityis associated with an upper-level area of low pressure and a surface trough.There is a potential for tropical or subtropical cyclone developmentof this weather system.
Image courtesy FNMOC
Classifications 07/1800 UTCTAFB - ST=1.5SAB - TWTC
GFS 24 h fcst valid 1200 UTC 08 Sep 2002 MSL and 10m wind
GFS 24 h fcst valid 1200 UTC 08 Sep 2002 850 mb wind and relative vorticity
Corresponding Danger Graphic
Result24 h lead time
1500 UTC 08 Sep30 kt ST depression
Examples of the TAFB Danger Graphic and
JTWC Formation Alert.
• TAFB DG is issued every 6 hours. While the JTWC FA is issued as needed and updated at least every 24 h.
• There is coordination between the duty NHC HS, JTWC, and Navy Norfolk when required.
JTWC Formation Alertissued 1200 UTC 25 Sep
TAFB Danger Graphic issued 0900 UTC 25 Sep
Verification of “Possible” TC Forecasts
Verification of “Possible” TC Forecasts
• 2002-2003 forecasts were verified from the TAFB TC Danger Graphic.
• Events forecasts were verified.
• The probability of detection (POD), false alarm rate (FAR), and average lead time were computed.
Verification Results of Atlantic Event ForecastsVerification Results of
Atlantic Event Forecasts
2002 2003
Probability of Detection
79% (11/14)
83% (15/18)
False Alarm Rate
32% (6/19)
32% (7/22)
Avg. Lead Time 13.7 h 28.1h
Future Changes
• Microwave satellite fixes-2004 season
• In-house SHIP QC Program
• Highest wave height during TC event
• Altimeter data in N-AWIPS
• Seamless Surface Forecasts Maps coordinated with OPC/HPC/HFO