Forecasting Oceanic Cyclones at the NOAA Ocean Prediction Center Joseph M. Sienkiewicz, D. Scott Prosise, and Anthony Crutch NOAA/NWS/NCEP/Ocean Prediction Center Camp Springs, MD

1
Forecasting Oceanic Cyclones at the NOAA Ocean
Prediction CenterJoseph M. Sienkiewicz, D. Scott
Prosise, and Anthony CrutchNOAA/NWS/NCEP/Ocean
Prediction CenterCamp Springs, MD
2
Timeline
Sanders and Gyakum, 1980 Defined the
meteorological oceanic BOMB Anthes et al.,
1983 QE-II Storm major improvementsthrough
improved initial conditions, improved horizontal
and vertical resolution and changes in the
physical parameterization of surface fluxes and
latent heating. 1986-Genesis of Atlantic Lows
Experiment (GALE) Sanders, 1986 LFM is able
to capture essentials of the baroclinic process,
and that the amount of response to baroclinic
forcing remains intractable. 1988-89
Experiment on Rapidly Intensifying Cyclones over
the Atlantic (ERICA) Sanders (1992) the
state of the art in prediction of marine cyclones
has advanced substantially over the last 15
years Uccellini et al., 1999 the occurrence
and position of intense, fast moving oceanic
storms can be forecast in the day-4 time range.
3
OPC Graphical Forecast Verification
48 Hour Surface Forecast VT 1200 UTC 18 Mar 2004
Verifying Surface Analysis 1200 UTC 18 Mar 2004
Methodology For all observed cyclones Compare
forecast analyzed Position Central
pressure Wind warning category
96 Hour Surface Forecast VT 1200 UTC 18 Mar 2004
4
OOCL America Pacific - Jan 2000 350 containers
lost overboard 217 crushed or bent out of usable
condition.
Vessel lost power for a short period and took
several severe rolls.
5
October 2003 through March 2004
A-350
P-279
P-228
A-170
- Sanders, 1988-89 (A-Atlantic P-Pacific)
6
Categorical Position Errors
A96-350
P96-279
P96-284
P96-248
P48-228
A48-170
P96- Uccellini et al 1993-94 A,P48,96 Sanders
88-89
7
Categorical MSLP Errors
8
Warning Verification ( Correct)
9
Warning Verification
10
Conclusions
Remarkable progress in 25 years!!! 48 and 96
hour extratropical cyclone forecasts comparable
or exceed 10 year average track errors for
Atlantic Tropical Cyclones Track error less for
Atlantic than Pacific for all categories of
cyclones Hurricane Force track error
significantly higher than other categories of
strong cyclones for 96 hour forecasts Mean
pressure errors - tendency to underforecast
highest for extreme storms error larger for
Atlantic than Pacific (96 hours, 965 hPa or
less) Warnings Difficulty forecasting
intensity of the hurricane force event tend to
underforecast by one warning category NWP
Forecast winds NWP Pressure errors Short
lived events (24 hours or less)
11
Future Work
Improve day 4 forecasts of Hurricane Force
Cyclones Climatology of extreme
events Reanalysis data, cyclone phase
diagrams Are we missing something? SST,
small scale structure Ensembles - Improve
deterministic forecasts - Develop a suite of
ensemble based probabilistic forecasts Must
educate the customer Forecasts for 6 to 10 day
period