Mark Willis

1
Observations and Modeling of Ocean Waves in
Tropical Cyclones

• Mark Willis
• Meteorologist
• National Weather Service
• WFO Newport/Morehead City
• Acknowledgements Gene Iceman Chiellini, Joe
  Sienkiewicz, Victor Zlotnicki, Fabrice Collard

2
NWS Marine Serviceswww.weather.gov/marine
47 Local NWS offices provide coastal water marine
forecasts near shore (out 20 or 60 nm depending
on location.
Ocean Prediction Center, NHC Tropical Analysis
and Forecast Branch (TAFB), and Hawaii Forecast
Office cover the open Atlantic and Pacific oceans
between 35W and 160E.
3
Motivation

• 22 of all tropical cyclone related deaths are
  offshore or surf related (2nd only to inland
  flooding).

From Rappaport (2000)
4
Observations
www.ndbc.noaa.gov Buoy, C-MAN, Ship Obs
5
Wave tables on NDBC website
What if there are more wave groups than the
primary swell and wind waves? We cant ignore
them
6
Spectral Wave Observations
NE 4 _at_ 11s Primary Swell group
NW 8 _at_ 6s Primary Wind Wave, dominant spectral
component
3 _at_ 17s masked in observation tables!
7
Scroll Down on individual buoy pages to get wave
energy versus frequency (period) plots get an
idea for all the wave groups in the spectrum.
8
Spectral Density Plots
9
Hurricane Isabel Failure of Diamond Shoals Buoy
41025
Buoy 41025 Peaked at 45 _at_ 17 s before ripping
from its mooring.
Hurricane Isabel 2003
10
Wave Observations from Satellites

• Altimetry began with GEOS-3 (1975) by NASA. Near
  real time by 1977.
• Radar Altimeters onboard satellites measure
  distance between satellite and sea surface by
  measuring the round trip travel time of microwave
  pulses sent from the spacecraft to the sea
  surface.
• The details of the shape of the returned radar
  pulse gives information on wind speed and wave
  height.
• A calm sea with low waves returns a short,
  sharply defined pulse whereas a rough sea with
  high waves returns a stretched pulse because the
  energy begins to come back from the tops of the
  waves and continues until the pulse hits the
  valley of the wave.
• Several studies showing as good as buoys.
• Jason 1 USA (NASA) and France (CNES). Highest
  orbit 1336km, quickest repetitivity (9 days).
  Jason 2 launching in 2008. Follow up on the
  Topex/Poseidon.
• GFO US Navy, Envisat (follow up on ERS-2).

11
Altimeter as good as the buoywe need to use it
more!

• Gulf of Mexico.
• Hwang et al.1998. JGR-Oceans. P10,451.

12
Very good correlation to buoy data

• From Tolman et al, 2002. Weather Forecasting
  p311.

Essentiallywe can use it and have faith in it!
13
Jason-1 Track - 24 hrs(2 WNA per day)
Footprint small, but provides coverage in data
void areas anything helps!
14
Jason-1 Complete Cycle 10 DaysNearly full
oceanic coverage
15
Now the fun stuff

• Despite the relatively small footprint it can
  be helpful in wave analysis and forecasting!
• Lets see

16
Hurricane Flossie Approaching the Big Island of
Hawaii last year
17
Super Typhoon Man-Yi
18
Jason showing NE wind waves/swell interacting
with opposing gulf stream current. 14 ft observed
vs 6 ft model.
19
NNE Swell Against Opposing Gulf Stream Current
20
Hurricane Noel
Jason (36 ft) vs MWW3 (21 ft) During Noels ET.
21
Not just for Tropical Cyclones
The St. Patricks Day Super Swell 2008
22
The St. Patricks Day Super Swell 2008
23
St. Patricks Day SwellGFS vs. QuikSCAT
24
SAR Instrument on Envisat

• Radar Echo takes movie of sea surface
• From movie and well known responses of radar
  echo to swell slopes and velocities, can convert
  SAR signal into 2D wave spectra.
• This product from Boost-Technologies takes peak
  wavelength (period) and direction of each swell
  partition of the 2D spectra and propagates it in
  time using the deep water dispersion relation.

25
St. Patricks Day Super Swell in the Atlantic
Biggest EVER for Puerto Rico?
26
Where can we get this stuff?

• Jason Data available at TPC/OPC via N-AWIPS. Not
  in AWIPS yet lets push for it!
• Paul Wittmann at FNMOC has been ever so grateful
  to help us get WW3 vs
  
  
  Altimeter data back on https//www.fnmoc.navy.mi
  l/ww3_cgi/index.html
• Jason Data http//nereids.jpl.nasa.gov/cgibin/s
  sh.cgi?showjason1
• SAR wavelength animations http//www.boost-techn
  ologies.com/shom/nrt_atl.gif

27
Models used for Hurricane Waves

• Global NWW3 1.25 X 1 deg (GFS)
• Regional WW3 (WNA, ENP) .25 X .25 deg (GFS)
• These models now 2 way nested into MWW3 as of Nov
  2007.
• Hurricane WW3 (NAH, NPH) .25 X .25 deg (GFS
  GFDL winds, soon to be HWRF).
• UKMET and FNMOC global (UKMET and NOGAPS winds)
• Parametric Wave Models (Lyons and Wu) user
  inputs fetch length, duration, wind speeds and
  significant wave height/period are calculated
  based on various empirical methods.

28
WNA vs NAH WW3 Hurricane Debby
DEBBY Hs
29
WNA vs NAH WW3
Debby Peak Period
Island Shadowing Issues!
30
NAH vs WNA Hurricane Ivan (2004)
Hurricane Ivan Buoys 42007/42040
31
WW3 driven by NHC Advisory winds (Wittmann et al,
2006)
Available at www.nrlmry.navy.mil/atcf_web/wavewat
ch
32
Were almost overBear with me

• Hurricane Florence Case Study (Willis, 2007).
• Broke typical conceptual models of wave
  forecasting.
• Resulted in 2 deaths in Central Florida, numerous
  reports of coastal flooding and erosion.

33
Hurricane Florence ( Sept 2006)
NE swell peaking in Florida after NHC advisories
ceased
34
THE METEOROLOGY

• Large storm from origin.
• Strong high pressure over Canada built south
  11-13 Sept.
• This allowed significant NE fetch to build SW
  towards Florida.
• Led to peak of swell in FL as Florence seemingly
  was not a threat and moving out to sea.

NWS/NCEP Unified Surface Analysis 12Z 11 Sept 2006
35
Wave Model Verification
Swell Peaking in Florida when Florence was moving
out to sea and no longer tropical dangerous!
Wave heights offshore Central Florida
36
Wave Model Verification
Wave periods offshore Central Florida
37
Wave Model Verification
Wave heights offshore Cape Hatteras
38
Wave Model Verification
Wave periods offshore Cape Hatteras
39
Why were the NAH and WNA WW3s off? The wind
forcing!
GFDL Winds (used in NAH WW3) vs. Observations
40
(No Transcript)
41
Florence (Cat 1, 2006) vs. Luis (Cat 4, 1995)
Luis
Florence
42
Florence (Cat 1, 2006) vs. Luis (Cat 4, 1995)
FLORENCE MAX 70 KT
OR LUIS MAX 120 KT ?
Which storm created the largest swell offshore
Florida?
43
(No Transcript)
44
And the winner isTIE! Both swell events peaked
around 14 at 16s. Not unusual for Oahu, but very
unusual for Florida.
Saffir Simpson Scale not valid for swell events.
Outreach needed to teach the public this.
45
Photos from Florence Swell in Palm Beach, FL.
46
Thank You Questions? Comments?