1. NOAA operational ocean modeling A joint NWS – NOS – IOOS enterprise Ming Ji NOAA/NCEP/OPC Frank Aikman IIINOAA/NOS/CSDL Hendrik L. TolmanNOAA/NCEP/EMC

2. Background  2004 Science Advisory Board and NOAA response A three-level system Global and basin scale operational Backbone Capability at NCEP (partnership with Navy) On NOAA operational super computers Coastal Operational Ocean Backbone Capability at NOS, integrated with NCEP effort Transitioning to NOAA operational super computers Operational local modeling at IOOS Regional Associations RA Operational modeling to adhere to NOAA standards

3. History  Louis Uccellini and Ming Ji at previous MARACOOS meeting (Fall 2012)

4.  Strategic Approach in Ocean Modeling Strategic alignment with Navy on Global Ocean model development Strategic alignment with NOS on coastal model development Strategic partnership with the IOOS community Linking backbone to local modeling for applications/services Standardization, realtime coastal/marine/biological data flow Community modeling and O2R Engage private sector through IOOS RAs for a broad range of local service delivery

5. From the whiteboard …  There are already many models in place, but particularly the interaction with the IOOS RAs still needs to be worked out in many details. SAB level: IIIIII AtmosphereGFS GEFS NAM SREF HRRR WRF ???? OceanRTOFS (HYCOM) CFS (MOM4) ROMS Adcirc FVCOM ….. Many WavesWAVEWATCH III (det. + ensemble) WAVEWATCH III SWAN NWPS Area:GlobalRegionalCoastal/Local

6. SystemCurrentQ2FY14Q1FY18 GEFS T254 (55 km) 0 to 8-dT574 SL (35 km) 0 to 16-dT1148 SL (17 km) 0 to 10-d T190 (70 km) 8 to 16-dT574 SL (35km) 10 to16-d T574 SL (35km) 16 to 32-d (new request) 20 Members @ 42 Vertical Levels 20 Members @ 64 Vertical Levels Semi-Lagrangian EnKF integrated with ETR for initial perturbations Full stochastic physics NEMS, coupled ocean MME (multi-model ensemble) Extended to week 3 & 4 Estimated compute factor 120 nodes for 1-hour2.5x – 300 nodes for 1-hour Note: uniform resolution for this configuration is to take advantage of Semi-Lagrangian model (saving computation cost for SL if the resolution is higher). 10x – 3000 nodes for 1-hour 400 nodes for 1 hour (new item) 1.Need additional resource to support for full resolution out to 16 days (20%) 2.Need additional resource to support higher vertical resolution (L128 – 2x) GEFS Evolution To 2018

7. SystemCurrentQ2FY14Q1FY18 NAEFS NCEP GEFS 21 members CMC GEFS 21 members FNMOC GEFS 20 members FNMOC GEFS 21 members 1.Exchange 90 variables 2.1*1 degree globally 3.Every 6 hrs, out to 16 days 1.Exchange ~100 variables 2.Additional 0.5*0.5 degree for selected variables, 3.Every 3 hours out to 8 days 1.Exchange 100+ variables 2.Exchange 0.25*0.25 degree for selected variables 3.Coupling with ocean, and extended to 32 days 4.Additional every 6 hours from 16 to 32 days 1.Bias correction 2.Downscaling 3.NAEFFS probabilistic products Expand more variables for 1.Bias correction 2.Downscaling 3.NAEFFS probabilistic products More statistical post- processing for week 3 & 4 forecast Estimated compute factor 10 nodes for 1-hour Time window – immediately follow GEFS finished 1.5x – 15 nodes for 1-hour8x – 120 nodes for 1-hour NAEFS Evolution To 2018

8. SystemCurrentQ2FY14Q1 FY18 SREF 21 members @ 16 km 35 levels Every 6 hours to 84 hr 20 members @ 12 km 35 levels Every 6 hours to 84 hr 20 members @ 12 km 50/60 levels with nests at 3km Some are 18-60 hr or 18-84 hr extensions of HRRRE members NAM 12 km North America parent to 84 hr Fixed nests CONUS, Alaska, HI, PR at 4/6/3/3 km run to 60 hr Placeable FireWx nest 1.33/1.5 km to 36 hr 12 km North America parent to 84hr Fixed nests CONUS, Alaska, HI, PR all run at 3 km Placeable/movable nest 1.5 km to 36 hr ENSEMBLE of runs: 12 km North America parent ENSEMBLE of 3 km runs CONUS, Alaska, HI, PR to Placeable/movable nest 1.5 km to 36 hr Every 6 hours to 36/60/84 hours Hourly to 18 hr with extensions for SREF & Every 6 hours control member extended to 36/60/84 hours for NAM HRRR/ HRRRE 3 km CONUSENSEMBLE of 3 km runs ARW for CONUS & Alaska NMMB for CONUS, Alaska, HI, PR & FireWx Hourly to 15 hoursHourly to 18 hours Mesoscale

9. SystemCurrentQ2FY14Q1FY18 CFS 9 month forecast Atmosphere: 100km, L64 Atmosphere: 35 km, L128 Ocean: 0.5°, 40 levels Ocean: 0.25°, 60 levels Waves: none Waves 0.5° 4 members per day (120 per month) ECMWF 7 month forecast Atmosphere: 50 km, L91 Ocean: 1°, 40 levels Waves: 0.5° 51 members per month (all run on first of month) Seasonal Climate

10. NCEP  Real Time Ocean Forecast System (RTOFS) Based on community HYCOM model. Strong collaboration with Navy. Adopting existing 1/12° model from NRL (NOPP). GFS forcing (including diurnal cycle). Timeline: Operational 10/25/2011 with NRL/NAVOCEANO (NCODA) initialization (daily feed from NAVO). FY2014-15: full initialization at NCEP. Developing agreement with NRL on NCODA.

11. RTOFS  Presently five major efforts: Eddy resolving ocean modeling. Eddy resolving ocean initialization. Coupled modeling for hurricanes. Coupled modeling for weather – CFS / NEMS. Episodic tracers (with shelf life) All RTOFS models presently based on HYCOM RTOFS represents line of products. RTOFS-Global RTOFS-Atlantic Operational 2005 Operational 10/25/2011 RTOFS-HWRFLive testing RTOFS-NEMSUnder development RTOFS-ET-Pac Operational 07/24/2012

12. RTOFS-Global

13. SubsystemQ2FY14Q1FY18 GlobalExtend forecast to 8 days, OSIP output, improved coastal forcing. Add DA on 00z cycle. Extend lead time, ice and wave coupling, improved DA (same resolution) Estimated compute factor 2x4x AtlanticRe-initialize, go to 3km gridImprove physics and assimilation, resource neutral, possible low-resolution ensemble. Estimated compute factor 1x of global1.5x of global East Pacific, West Pacific & Alaska New models, compatible with Atlantic Estimated compute factor 3 times 1.5x of global HWRFUp to 30% of coupled HWRF NEMSDepending on GFS/CFS approaches TracersWill discontinue early 2014Depending on disasters... Estimated compute factor 0x0.1x RTOFS (HYCOM) Evolution To 2018

14. NCEP  Future RTOFS (2015-2018) Summary Global model at 1/12 degree resolution Regional models for Arctic, North Atlantic, East Pacific and possibly West Pacific at 3km nominal resolution (based on NCEP user requirements) Intend to go to ensemble based systems by 2018 depending on computer resources (particularly for ice) Full coupling (Atmos-Ocean-Wave-Ice) particularly in ice zones

15. SystemCurrentQ2FY14Q1FY18 Bay & Coastal Ocean- Forecast Systems GLOFS, CBOFS, TBOFS, DBOFS, CREOFS, etc. Expand from 13 to 17 OFS's (San Francisco, NE and NW Gulf of Mexico, Cook Inlet) Expand from 17 to 22 OFS's (Gulf of Maine, Mississippi River, Huron/Erie, West Coast, East Coast) Estimated compute factor 2x4x ESTOFS Increase Atlantic nearshore resolution to 500 m and increase number of ensemble members (from 5 to 10) Increase Pacific nearshore resolution (to 500 m) and number of ensemble members (5 to 10 members) Estimated compute factor 2 nodes 120 min10x2x NOS Systems Evolution To 2018

16. IOOS  Partnership with IOOS at NWS & NOS Data standardization and management through IOOS: RTOFS models available in NetCDF on NOMADS This was a major change driven by IOOS requirements Critical Partner for real time coastal/marine/biological data flow Providing requirements for RTOFS E.g., resolution, accuracy, … NOS collaboration with the IOOS modeling community Develop, agree & adhere to community standards Includes frameworks, data flow, products Effectively leveraging Testbeds (e.g. IOOS COMT)

17.  Summary The 3-level framework from SAB review Strategic partnerships are essential elements Navy, NOS, IOOS community NOAA-IOOS RAs paradigm toward service delivery

18. Issues/challenges for discussion R2O-O2R: Connecting Backbone to local models Define operations How to run models (CONOPS) How to exchange data and information E.g., Standards Coordinate operations and service delivery E.g., the Weather Enterprise

20. NOS  Operational Forecast Systems (OFS) for the Coastal and Estuarine Environment in NOAA’s National Ocean Service (NOS) OUTLINE The state of NOS’s Operational Forecast Systems Requirements and geographic coverage CBOFS, NGOFS and ESTOFS examples OFS products & dissemination The Coastal Ocean Modeling Framework (COMF) Objective: More efficient R&D, O&M Challenges and Applications Collaboration with the IOOS Ocean Modeling Community Coupled Model Systems

21. NOS  NOS Marine Modeling Requirements  Develop a national network of operational hydrodynamic models providing nowcasts and short-term  (0 – 48 hr.) forecasts for:  Support of safe & efficient navigation  Water levels for under-keel clearance  Currents for right-of-way, maneuverability  Emergency response  HAZMAT  Search & Rescue  Homeland Security  Environmentally sound management of the coastal zone  Ecosystem applications  Marine geospatial applications Salinity SST

23. Chesapeake Bay Operational Forecast System (CBOFS) CBOFS (2-D) Operational 08/01 CBOFS (3-D) Operational 04/11 CurrentsTempSalt

24. Northern Gulf of Mexico Operational Forecast System (NGOFS) NGOFS is the first NOS OFS to extend the model domain to US shelf waters to fill gaps between global/basin models and coastal/estuarine models. NGOFS covers 6 NOS PORTS

25. Computes surge with tides for forecasting and for coupling to WAVEWATCH III® Applies ADCIRC model – 254,565 nodes – Coastal resolution ≈ 3 km – Depth-averaged barotropic – 6-hr nowcast followed by 180-hr forecast Pacific model in development http://www.opc.ncep.noaa.gov/ estofs/estofs_surge_info.shtml http://slosh.nws.noaa.gov/etsurge_ESTOFS/ Atlantic Extratropical Surge and Tide Operational Forecast System (ESTOFS)

26. Graphics on CO-OPS Web Site  Time Series Plots (24 hour nowcast and 48 forecast) of water levels, currents, temperature, salinity, and surface winds at selected locations  Contour and vector map plots and animation of water levels, currents, temperature, salinity, and surface winds Model data sets on NOAA’s WOC and CO-OPS Thredds Server  Station/point netCDF files (6-minute output):  Time series at selected stations  Gridded model output netCDF files (hourly output) Model visualization via nowCOAST GIS-Based Web Mapping Portal (WMS) Users value usefulness, accessibility, reliability  Outputs used by USCG, NOS/OR&R, ASA, etc. in real-time  Web products used daily by coastal managers, maritime navigation and emergency response communities.  Reliability and timely delivery of products are most important. NOS Coastal Operational Forecast System Products

27. REAL-TIME DATA INGEST QA/QC (COMF) OPERATIONAL MODELS (NOAA’s HPC) FORECAST MODEL GUIDANCE (water level, water temp, currents, & salinity) PRODUCTS (web pages and digital pt. & gridded data) FOR USERS tidesandcurrents.noaa.gov QA/QC (CORMS) 24 x 7 NOS Coastal Ocean Modeling Framework (COMF) Individual Model systems Data Tank on WCOSS: Atmospheric ForcingAtmospheric Forcing Coastal Boundary ConditionsCoastal Boundary Conditions Riverine Fresh Water InputsRiverine Fresh Water Inputs Products and archives Linux Server in CO-OPS

28. NOAA/NOS Coastal Modeling Challenges Continued Collaboration with the IOOS Ocean Modeling Community Develop, agree & adhere to community standards (e.g. frameworks; data flow; products) Testbeds (e.g. the IOOS COMT) Coupled Model Systems Riverine-estuarine-regional/coastal-basin; Hydrodynamic-wave; hydrodynamic-sediment transport Physical-biogeochemical coupling (e.g. ecological; water quality; habitat) Forecast uncertainty estimation Probabilistic approach/Ensemble averaging Coastal data assimilation techniques HF Radar; coastal altimetry; IOOS data; etc.