Office of Coast Survey NOAA’s Storm Surge Modeling Capabilities Jesse C. Feyen Storm Surge Roadmap Portfolio Manager.

Office of Coast Survey

NOAA’s Storm Surge Modeling NOAA’s Storm Surge Modeling CapabilitiesCapabilities

Jesse C. FeyenStorm Surge Roadmap Portfolio Manager


NOAA’s Tropical Storm Surge ProductsSLOSH-based storm surge products


SLOSH: NWS’ Operational Surge Model

• The basis for tropical products and services• Internal hurricane model driven by storm track,

radius of maximum winds, and central pressure• Uses domains (basins) which vary in resolution

from several hundred meters near the coastline to a few kilometers offshore– Basins have sub-grid cell features to model barriers

(such as levees or roadways) and

Further info: http://slosh.nws.noaa.gov/sloshPub


Generating MEOWs and MOMs• Primary purpose for SLOSH is to estimate

potential surge for hurricane evacuation studies– NHC runs thousands of hypothetical storms through a

basin to create composites of potential surge in the form of MEOWs and MOMs

– More info: http://www.iwr.usace.army.mil/nhp/

MEOW = Maximum Envelopes of WaterWorst case snapshot for a particular storm category, forward speed, trajectory, and initial tide level, incorporating uncertainty in landfall location

MOM = Maximum of MaximumsCombines all MEOWs for a particular storm category


How a MEOW is Assembled


Predicting Imminent Events

• Real-time deterministic runs– National Hurricane Center (NHC) begins operational

SLOSH runs when a hurricane watch is issued (48 hours before landfall) based upon NHC’s official forecast

– Used by NHC and local Weather Forecast Offices to guide advisories and statements


Deterministic Output for Ike (2008)But the prediction from one run is misleading

because it ignores forecast uncertainty


Hurricane Advisory – Approximately 12 hr. before landfall

NHC TRACK ERROR 12 hr. OUTNHC TRACK ERROR 12 hr. OUT

133 mph, 933 mb.


Actual Hurricane Track 30 mi. E of 12 hr. Advisory Forecast Track

TRACK FORECAST

ACTUAL TRACK

133 mph, 933 mb.


Rmax=25 mi(forecast)

Surge Based on NHC 12 hr. Advisory


Rmax=40 mi

Surge Based on NHC Storm Best Track


Filling the Gap: P-Surge

• Probabilistic storm surge predictions• Ensemble based on NHC forecast advisory

– Perturbation of cross-track error, along-track error, intensity error, and size error

• Computes probabilities and exceedance heights– “Probability of water exceeding X feet” available on

NHC’s website: http://www.nhc.noaa.gov/index.shtml– “Height exceeded by X%” of ensemble available at

MDL: http://www.weather.gov/mdl/psurge/

• Available in .kml, .shp, GRIB2


Perturbing the Forecast• Include 90% of

possible cross track error (roughly 3 times the size of the cone of error)

• Similar perturbations applied to size, speed, and intensity

• Results in hundreds of simulations


Probability of > 5 feet of Storm Surge for Katrina Advisory 23

•Provided– Ensemble of

hypothetical storm runs

– Associated weights

•Can compute– Probability of

exceeding a certain height


Height Exceeded by 10% of the Ensemble for Katrina Advisory 23

•Provided– Ensemble of

hypothetical storm runs

– Associated weights

•Can compute– Height

exceeded by a certain probability


Pacific SLOSH Basins- MOMs and MEOWs haven’t been generated- Operational?


SLOSH Display Package (SDP)

• Allows for viewing and exploring SLOSH data– Displays basins, MOMs, MEOWs, and historical

storms for viewing and analysis– Can animate and explore native .rex files– Can output .shp and .kml

• SDP is intended to be used by trained emergency managers, FEMA personnel, and NWS forecasters

Download: http://slosh.nws.noaa.gov/sloshPub/disclaim.php


Extratropical Storm Surge (ETSS)

• Operational modeling system based on SLOSH– Uses a larger grid since it is a larger phenomena and

driven by the Global Forecast System (GFS) model– 96 hourly surge forecasts are made every 6 hours, 24

x 7– Predictions given at points around U.S. coasts:

http://www.nws.noaa.gov/mdl/etsurge/– Visualizations also available:

http://www.opc.ncep.noaa.gov/et_surge/et_surge_info.shtml


ET-Surge Website Combines Tide and Surge Forecasts


Visualization of Gridded ETSS Predictions


Coverage of ETSS


Atlantic Extratropical Surge and Tide Operational Forecast System (ESTOFS)• Purpose

– Second operational set of forecast guidance for extratropical storms that covers the Western North Atlantic basin, Gulf of Mexico, and Caribbean Sea

– Compute surge with tides, astronomical tides, and subtidal water levels (surge only)

• Pros– Incorporates the astronomical tides, lacking in current model– ADCIRC unstructured grid could improve coastal resolution

• Cons– No river inflows or seasonable water level adjustments– Insufficient grid resolution at rivers, inlets, and barrier islands


ESTOFS-Atlantic

• Applies widely validated ADCIRC model

• East Coast 2001 tidal database grid (EC2001)

• 254,565 nodes• Coastal resolution ≈ 3

km• Specify the tidal forcing

at 60o W


Operational Set-up

• 4 times per day running cycle– Time is 00z, 06z, 12z, and 18z

• Total 8 days simulation– 12-hr nowcast followed by 180-hr forecast

• Surface forcing from GFS – 10 m winds and sea level pressure every 3 hours

• Tidal forcing from OSU TPXO v6.2• Scheduled to be operational at NCO FY12 Q2


ESTOFS Output• Deliver three types of water level

– Combined Water Level (CWL): Surge + tides

– Harmonic Tidal Prediction (HTP): Astronomical tides

– Subtidal Water Level (SWL): SWL = CWL – HTP

• Provide both field and point output– 6 minute water level at points, hourly water level for fields

• Output on grid (NetCDF) and 2.5 km NDFD (GRIB2)

2009 Veteran’s Day Nor’Ida2009 Veteran’s Day Nor’Ida


Water Level as Fields

EC2001 grid (NetCDF) NDFD CONUS grid (GRIB2)


Water Level as Fields

EC2001 grid (NetCDF) NDFD CONUS grid (GRIB2)


Hindcast Skill Assessment(ESTOFS CWL vs ETSS + CO-OPS TP + Anomaly)

East Coast Gulf of Mexico


Ensemble Model Development• Uncertainty in initial

conditions• Evaluate the extent of

model predictability• Model uncertainty:

dynamics• Model uncertainty:

Physics

• Deterministic model only consider one IC

• Single model can not give us probability clue

• Only one numerical method for single model

• Only one set of physical parameterizations can be used


Office of Coast Survey NOAA’s Storm Surge Modeling Capabilities Jesse C. Feyen Storm Surge Roadmap Portfolio Manager.

Documents