john cannon

Water Level Prediction NomogramWeb Based Experimental Tool(Proposal for Portland Harbor and nearby locations/beaches) : Water Level Prediction Nomogram Web Based Experimental Tool (Proposal for Portland Harbor and nearby locations/beaches) NERACOOS Meeting April, 2007


Water Level Prediction : Water Level Prediction Water level forecast based on… Astronomical tides Storm surge guidance Prediction of waves (GoMOOS SWAN model?) Time (48 hours) Follow the bouncing ball Not limited to the stray 1-2 times a year coastal flood events!!!


Tidal Flooding: Top Ten Tides : Tidal Flooding: Top Ten Tides 1. 14.70’ - Feb 7, 1978 (Bliz. Of ’78) 2. 13.98’ - Jan 9, 1978 3. 13.40’ - Dec 4, 1990 4. 13.31’ – Mar 16, 1976 5. 13.29’ – Nov 30, 1944 5. 13.29’ – Nov 30, 1945 6. 13.18’ – Jan 2, 1987 6. 13.18’ – Oct 30, 1991 (Perfect Storm) 7. 13.09’ – Apr 7, 1978 7. 13.09’ – Dec 29, 1959 7. 13.09’ – Feb 19, 1972 8. 13.07’ – Jan 28, 1979 9. 12.80’ – Dec 3, 1986 10. 12.79’ – Apr 28, 1940 10. 12.79’ – Jan 20, 1961 Splash-over along a seawall courtesy Bruce Budd


Tidal Flooding: Top Ten Storm Surges : Tidal Flooding: Top Ten Storm Surges 1. 4.3’ – Mar 3, 1947 2. 4.1’ – Mar 1, 1914 3. 3.9’ - Dec 14, 1917 4. 3.6’ – Feb 19, 1972 5. 3.5’ – Nov 26, 1950 5. 3.5’ – Feb 7, 1978 (Blizzard of 78’) 5. 3.5’ – Oct 30, 1991 (Perfect Storm) 6. 3.3’ – Nov 30, 1945 6. 3.3’ – Aug 31, 1954 7. 3.2’ – Dec 2, 1942 8. 3.1’ – Mar 16, 1956 9. 3.0’ – Jan 15, 1940 9. 3.0’ – Feb 7, 1951 10. 2.9’ – Nov 13, 1925 courtesy Bruce Budd


Coastal flood talk? Not today… : Coastal flood talk? Not today… Just a few refresher slides… I have 100s of coastal flood slides if you choose to abuse yourself


Water Level Prediction Nomogram : Water Level Prediction Nomogram Users/Emergency managers (and others) could choose to use the interactive feature (“slider bars”) – on a web site Education/disclaimer would be needed Non-interactive display could be introduced first Should be used daily: not shelved for the twice a year events


Water Level and Waves : Water Level and Waves Late October, 2006 Storm


Patriot’s Day Storm : Patriot’s Day Storm


Patriot’s Day Storm : Patriot’s Day Storm


March 17, 2007 (St. Patty’s Day) : March 17, 2007 (St. Patty’s Day)


- - Storm Surge - - * Surge = observed tide - predicted tide * Mainly driven by wind…but there are other factors that numerical models do not include ……..Waves!!! : - - Storm Surge - - * Surge = observed tide - predicted tide * Mainly driven by wind…but there are other factors that numerical models do not include ……..Waves!!! http://www.nhc.noaa.gov/HAW2/english/storm_surge.shtml


We have two models… : We have two models…


Summary of ET Storm Surge Guidance : Summary of ET Storm Surge Guidance WAVES…not directly incorporated (pressure already is)


GoMOOS Water Level Prediction : GoMOOS Water Level Prediction Should use storm surge guidance Should be an ensemble? (use two operational storm surge models) How should this combination be applied? 60% weighted GFS model? 40% weighted NGM model? Why choose these initial numbers???...


Error of the GFS (MRPSSE Product) Portland Harbor : Error of the GFS (MRPSSE Product) Portland Harbor Data included an average of 2 storm surge products prior to coastal flood event Most back to back runs were nearly identical! The MRPSSE bias varied much more than the MRPECS product. The MRPSSE product under-predicted 7 surge events and over-predicted 4. Data range: -1.10’ - +0.61’; Average bias was -.3 feet. [GFS surge Guidance available at: http://www.weather.gov/mdl/etsurge/index.htm]


Error of the NGM (MRPECS Product) (Portland Harbor) : Error of the NGM (MRPECS Product) (Portland Harbor) Data limited to 15 coastal flood events using the MRPECS product Data included an avg. of 2 storm surge forecasts immediately prior to the coastal flood All back to back model runs (00Z and 12Z) were nearly identical (this is good!) 12 of 15 showed a model over-forecast surge bias (2 forecasts hit the mark) Data range: -.2’ to +1.2’; Average bias was +.38 feet.


Model Surge Bias Summary : Model Surge Bias Summary Limited dataset...but... MRPECS storm surge avg. overestimate ~ +0.4 ft MRPSSE storm surge avg. underestimate ~ -0.3 ft A blend of the guidance may be useful times Caution! The best forecast may be highly dependent on overall synoptic conditions shown by either the NGM or GFS (storm track, pressure gradient, wind direction etc.)


Merging of Storm Surge Models: An Ensemble : Merging of Storm Surge Models: An Ensemble Usually in meteorology…ensembles outcome provides the most likely solution May need additional verification Is there any other known verification out there at all????


Verification From St. Patty’s Day Storm… : Verification From St. Patty’s Day Storm… NGM Storm Surge for 2 computer runs prior to storm = 1.8 feet GFS Storm Surge for 2 computer runs prior to storm = 0.7 feet Avg. (NGM/GFS storm surge) = 1.25’ Actual storm surge = 1.24’ (1 hr after high tide)


Peak waves occurred near high tide (courtesy GoMOOS) : Peak waves occurred near high tide (courtesy GoMOOS)


Surge does not necessarily cause increased water levels: Blowout Tide–or Tide Levels Below MLLW could impact larger vessels : Surge does not necessarily cause increased water levels: Blowout Tide–or Tide Levels Below MLLW could impact larger vessels Mount Holly office


Other important reasons to describe tides/Large waves…Missing Lobsterman; Several capsized boats, Sept 13th 2006 (Florence) : Other important reasons to describe tides/Large waves… Missing Lobsterman; Several capsized boats, Sept 13th 2006 (Florence)


A couple slides describing waves and wave development… : A couple slides describing waves and wave development…


Energy Spectrum: The relationship between wind speed and wave height and energy : Energy Spectrum: The relationship between wind speed and wave height and energy


Statistical Wave Spectrum : Statistical Wave Spectrum Raleigh Distribution


Plunging Breakers & beach erosion(this often occurs along our beaches) : Plunging Breakers & beach erosion (this often occurs along our beaches)


The Role of “Wave Set-up” : The Role of “Wave Set-up” Here’s the kicker...”Wave setup” is the water rise at the coast due to breaking waves on the beach. Basically - - potential energy of large ocean waves is converted into the kinetic energy required to maintain an anomalous surge of water near the shoreline.


Wave Set-up : Wave Set-up Wave set-up (S) is breaker-height dependent, and hence, breaker-depth dependent as well. The Set-up magnitudes “S” = 10 to 15% of Shoaling Breaker Heights Courtesy COMET Shallow Water Waves Module, 2006


Other Secondary Effects: Breaker Run-up Assists Splash-over : Other Secondary Effects: Breaker Run-up Assists Splash-over On any beach there is a “spectrum” of waves… Most crashing waves run up the slope generally about the same distance. Periodically however, a wave will crash and run up the beach much higher than the average waves. This can cause “splash-over”. The highest 2 % breaker run-up can be mathematically computed (Water rise ~ 1/2 the predicted breaking wave height plus 50 cm) Breaker run-up Splash-over (York Beach, Maine)


Supporting documentation (NWS GYX coastal flood study)… : Supporting documentation (NWS GYX coastal flood study)… Suggests an empirical relationship between… Storm tide Large ocean waves Coastal flood or splash-over damage


Large waves are associated with major flooding : Large waves are associated with major flooding


GoMOOS Diagram : GoMOOS Diagram Water level prediction: Forecast Total tide from the NGM and GFS models (astronomical tide +/- surge) versus… Forecast offshore wave height (SWAN) - some adjustments may need to be made to forecast wave height at the beach


Display interactive nomogram logarithmically for best visual effect(color code “threat” regions) : Display interactive nomogram logarithmically for best visual effect (color code “threat” regions)


Slide39 : Severe Coastal Flooding Moderate Minor High Water Low Water Extremely Low Water * WATER LEVEL & WAVE PREDICTION Nomogram (Predicts Daily Water Level Forecast Waves including; Flooding/Splash-over, Beach Erosion) [* Experimental for demonstration purposes only] Minor Beach Erosion Moderate Beach Erosion Predicted Wave Height avg. at GoMOOS buoy “B & C” (ft) (SWAN Model) Predicted Storm Tide (ft) (astronomical tide +/- storm surge ensemble) 14’ -4’ 11’ 0’ 20’ Storm Tide Adjust Wave Height Adjusted by User Severe Beach Erosion Very Low Water Very Low Water Time Loop Hour 36 (Hour 00) = Splash-over = Minor Coastal Flooding = Moderate Coastal Flooding = Major Coastal Flooding Water Level is MLLW FS = 12’ 5’ 2’ 10’ 15’ Logarithmic Scale Splash-over Begins = Beach Erosion = No Problems Expected


Diagram : Diagram Should be… Dynamic, logarithmic in scale Visual Color coded for threat categories Help ship operators/beach goers etc…anyone who would need a water height Used daily by many users


Interactive features : Interactive features Should be… Small adjustments to forecast wave height 0 – 5 feet? Small adjustments to storm surge magnitude 0 – 1 foot?


GoMOOS Water Level Prediction : GoMOOS Water Level Prediction Put it out to the user community…ask for comments/suggestions/improvements Label as experimental