http www maretec mohid com ww 3 Who

http: //www. maretec. mohid. com/ww 3

Who forecasts waves in Portugal? 3 Project MOCASSIM (IH) 3 Operational wave model for the Portuguese coast (Pinto et al, 2004) 3 Wave. Watch III model predictions for the Iberian western coast (Pinto et al, 2003) 3 Tests with SWAN coupled with WW 3 3 Instituto de Meteorologia 3 24 hours forecast (no spatial resolution) 3 Sites de surf 3 Use coarse resolution U. S. model results 3 Outras aplicações de modelos de ondas 3 IST (Naval, Guedes Soares)

Wave Forecast system for the Portuguese Coast – Present state 3 Wave model – Wave Watch III 3 Most used model for meso-scale wave modelling 3 Wind models 3 GFS (NCAR - resolution = 1ºx 1º) 3 ARPS (Meteo. Galicia – resolution = 50 x 50 km)

Wave Watch III - v 2. 22 3 Wave Watch III (Tolman 1997, 1999 a) 3 Refraction 3 Source terms (2 packages: WAM model or Tolman and Chalikov (1996)) 3 Wave growth and decay due to wind 3 Nonlinear resonant interactions 3 “Whitecapping” (dissipation) 3 Bottom friction 3 Linear wave propagation 3 Sub-grid representation of unresolved islands (not used by us, but to be) 3 ANSI FORTRAN 90 modular, most computing options performed at compile level, Open. MP, MPI

Wave Forecast system for the Portuguese Coast – Present state 3 Model setup (Pablo Carracedo – Meteo. Galicia) 3 North Atlantic Ocean (0. 5ºx 0. 5º) 3 Iberian Peninsula (0. 25ºx 0. 25º) 3 Portuguese Coast (0. 05ºx 0. 05º) 3 72 hours forecast on a daily basis, updated around 7 a. m. GMT (~68 hours) 3 Operational mode running since May 2004

Operational application -Possibility to easily add new nestings to the North Atlantic Ocean Model -Runs without need for any manual work -Sends a log e-mail in the end of the simulation -Starts everyday at 3 a. m. -Ends normally at 4 h 45 a. m. -Tries until a determined hour to start -If it fails to launch, starts from best restart possible, which can only be done for 3 days. More than that it must be restarted completely – needs 3 days simulation to get reliable results

Operational application steps 3 Downloads and stores winds (~5 MB/day) 3 Runs preprocessing tools (wind interpolations, nesting files, restart files, update start/end dates, etc) 3 Launches model 3 Store results and copies nesting files(~130 MB/day) 3 Plots results using GRADS (also possible with Mohid Post. Processor or Mohid. GIS) 3 Converts to HDF 5 (results can be used by MOHID) 3 Publishes on the Internet

North Atlantic model

North Atlantic model - Hurricanes

Portuguese coast model

Model validation and reliability 3 No validation performed so far 3 Model results have been followed qualitatively against real-time buoy data (IH, Puertos del Estado) and other model results (Meteo. Galicia, IM, Puertos del Estado). Same quality as IM forecasts. 3 Operational system is reliable. Failures due to: 3 Internet is down 3 Failure in wind models

Validation – Leixões buoy Model Buoy

Validation – Sines buoy Model Buoy

Products from WW 3 3 Navigation warning system 3 Commercial ships 3 Recreational boats 3 Oil spill management system 3 Forecast for the general public (newspapers, TV, surf websites)

Future work with WW 3 - 1 3 Validate the Portuguese Coast model (Time series comparison with automatic buoys – ask/buy data from IH? ) 3 Improve parameterizations 3 Independency in terms of wind forcing using MM 5 -IST and downloading GFS (Extend to 120 or 144 h forecasts – winds are reliable) 3 Higher resolution models for Azores and Madeira 3 Improve speed (use a faster computer or use MPI version with several computers – dedicated cluster? )

Future work with WW 3 - 2 3 Real time validation 3 Results publication in an operational modelling website including 3 non-technical version 3 operational storm warning system for navigation 3 validation against satellite images 3 Couple with MOHID? – It’s possible

Steps to be taken - 1 3 Get someone to work on this? 31 person full time 3 Understand waves 3 Find knowledge and learn from who knows 3 Understand wave modelling (North Atlantic swells, wind waves, Portuguese coast) 3 Time depending on person’s knowledge 3 Validate the Portuguese coast model (WW 3) 3 Get buoy data and compare with available results (1 month) 3 Improve parameterization (2 -3 months)

Steps to be taken - 2 3 Independency from Meteo. Galicia winds 31 week 3 Higher resolution models for Azores and Madeira 31 week 3 Improve speed (MPI) 3 3 days

Steps to be taken - 3 3 Choose a model for near shore simulations 3 SWAN – Simulating WAves Nearshore (open source, Pablo from Meteo. Galicia has experience in implementation, used and accepted in scientific community, parallel processing) 3 GKSS – Wave model faster than SWAN, similar results but not very disseminated in the scientific community, possibility of cooperation 3 Others – STWAVE, REFDIF, … 3 Import knowledge and implementation technology 3 Pablo Carracedo (Meteo. Galicia) – 2 weeks

Steps to be taken - 4 3 Implement SWAN (or chosen model) in strategic places 3 Tagus (SIMTEJO), Costa do Estoril (SANEST), Zambujeira do Mar (ICREW), Óbidos (Águas do Oeste), Ericeira (SIMTEJO), Aveiro (SIMRIA), Algarve (Águas do Algarve) 3 Start with most urgent site (implementation – 2 months)

Steps to be taken - 5 3 Technology - Couple with MOHID (Convert to HDF 5) 3 Sediment transport (Done) 3 Done (Final Course Thesis) 3 Near shore currents (Done) 3 Mohid-STWave coupling iterative software (done) 3 Influence on turbulence and mixing 3 MSc/Ph. D (include/apply revised turbulence models) 3 Oil dispersion (Done)

Steps to be taken - 6 3 Create an application to easily implement a high resolution near shore model operationally (emergency plans) 3 Setup forecast system in 24 hours or less with full coupling with Wave Watch III, wind models and Mohid (EROCIPS) – 1 year 3 Data base 3 Coupling tools 3 Emergency plans 3 Useful in oil accidents – wave action on dispersion (EROCIPS) 3 Navigation warning system in the Tagus estuary (TRANSTEJO) or in other commercial ports in which circulation can be affected by waves (Aveiro, Leixões, Viana do Castelo, Sines, etc) 3 Needs high resolution wind model? 3 Is it profitable?