230 likes | 393 Views
RTOFS report. with a short diversion to other NCEP and NOAA activities. Hendrik L. Tolman Chief, Marine Modeling and Analysis Branch NOAA / NWS / NCEP / EMC Hendrik.Tolman@NOAA.gov. Almost out of scope. Other activities in NOAA relevant for GOVST GFDL: Climate scale ocean modeling:
E N D
RTOFS report with a short diversion to other NCEP and NOAA activities. Hendrik L. Tolman Chief, Marine Modeling and Analysis Branch NOAA / NWS / NCEP / EMC Hendrik.Tolman@NOAA.gov
Almost out of scope • Other activities in NOAA relevant for GOVST • GFDL: • Climate scale ocean modeling: • Going from MOM to GOLD. • Moving to coupled atmosphere-wave-ocean. • Appears to be some renewed interest in coupled hurricane modeling. • NOS: • Operational models transitioned to NCEP compute: • GL, Chesapeake, Delaware and Tampa Bay. • Northern Gulf, SF Bay, Columbia river next. • IOOS RA-s. • NOAA storm surge roadmap team lead by NOS • Strong link with USACE, Academia.
Almost out of scope • Ditto at NCEP: • CFS: • Probabilistic seasonal forecasting up to 9 months. • CFC v2 operation, v3 in design stages. • ENKF being set up for GODAS (MOM4). • Other relevant products …
RTOFS-Atlantic RTOFS-Global nested RTOFS coupled to HWRF RTOFS-NEMS RTOFS • Four major efforts: • Eddy resolving ocean modeling. • Eddy resolving ocean initialization. • Coupled modeling for hurricanes. • Coupled modeling for weather – CFS / NEMS. • All RTOFS models based on the community HYCOM model, MMAB is part of the HYCOM consortium. Operational 2005 Operational 10/25/2011 Live testing Early development
RTOFS-Atlantic • RTOFS-Atlantic • First full HYCOM model to become operational at NCEP. • Replacing previous POM model for East Coast (excluding Gulf of Mexico). • Lack of capability to get Gulf Stream to detach properly from Cap Hatteras. • Issues with model: • Left in limbo to get RTOFS-Global in fast. • Major boundary fixes this week. • To be nested in RTOFS-Global. • Expect to be fully up to speed in 6 months. • Next decision point: do we keep this model around? • Tides. • 4-6 km coastal resolution for US.
paradigm shift for NWS RTOFS-Global • Push to global model (vs. regional models): • Satisfy requirements from NOAA SAB. • Provide boundary data for regional models. • 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: full initialization at NCEP. • NOMADS as main distribution points (OpenDAP, NetCDF). • RTOFS-Atlantic as testbed for global. • Linking to NOS Coastal Ocean Modeling Framework. • No tides yet (unlike RTOFS-Atlantic)
RTOFS-Global Grid and layer 1 temperature snapshot: Conversion to GFS forcing “clean”, but we want to look in more detail at the Artic Ocean.
RTOFS-Global 00 Z 06 Z 12 Z 18 Z Forecast Step 1 using GFS from n-00 to f-48 hours Forecast Step 2 using GFS from f-49 to f-96 hours Forecast Step 3 using GFS from f-97 to f-144 hours Nowcast/Hindcast using GDAS from n-48 to n-00 hours Initialization file from NAVO at n-48 hours via TOC Post-processing: generation of NetCDF files after each step (adding selected GRIB2 later).
RTOFS-Global • Products: • Class I : Global netCDF files on native horizontal grid but interpolated to isolevels. Delivery via NOMADS, ftpprd and NODC archives. • Surface 3 hourly files (8 variables) ~ 120 GB per cycle • Volume 3D files daily (8 variables, 33 z levels) ~ 160 GB per cycle. • Target: General user; maximum flexibility for slicing/dicing data using NOMADS/OpenDAP servers (both GDS & TDS). • Class II: Sub-regional and basin GRIB2 files (Mercator grid). • Delivery via ftpprd and AWIPS. • Surface 3 hourly files (7 variables) ~ 5 GB per cycle • Target: Internal NWS needs to provide results on AWIPS or via FTP.
RTOFS-Global • Products cont’ed: • Class III: Regional (CONUS-East, CONUS-West, Alaska) netCDF files. • Delivery via NOMADS and NODC archives. • Volume 6 hourly files (u,v,T,S) • Target: Other centers within NCEP (TPC, OPC) and NOS applications. • Eventually tide-resolving output time intervals. • Requires NOMADS upgrades.
RTOFS-Global • Focus on producing GODAE metrics products • Examples follow: • Class 1: Differences with SSH fields fromindependent analyses. • Class 2: Drifts fromclimatology at selected WOCE sections. • Class 3: Location of Gulf Stream. • Class 4: Statistics on location of Gulf Stream; Daily comparisonwithindependent SST, SSH data. • Working on: • Class 3: Transports at more than 100 sections. • Class 4: Comparisons of forecasts/analysis with ARGO profiles.
Fukushima • EMC became US government lead on • ocean plume modeling for Fukushima Dai’ichi ocean contamination. • Collaboration in large Interagency work group. • Leveraging modeling from Navy, DTRA, NOS. • NCEP Navy model data distribution point. • CONOPS to rapidly generate actionable information for decision makers. • Product 1: Surface particle tracing to identify potentially contaminated areas. (April 6, 180 d) • Product 2: Contamination estimates using particle tracing + HYSPLIT atm. Deposits. (April 20, 180 d) • Product 3: Full dispersion model as passive decaying tracers, using direct release (NOS) and atm. deposit (NCEP) (dev., still much source issues)
Fukushima Not for distribution, internal use only
RTOFS-HWRF • Ongoing experiments • Real time parallel runs since 2009. • Switching over to RTOFS-Global. • Many lessons learned
RTOFS-HWRF Intensity Forecast for 6 TCs (186 cases): Gert07L, Irene09L, Katia12L, Maria14L, Ophelia16L, and Philippe17L HY11=HyHWRF2011; HWRF=operational HWRF • HyHWRF average intensity error and bias are the best among participant models, except degradation at 12h in average error and negative bias at 24h. • HyHWRF standard deviation is consistently the smallest, except 12 and 24 h. 20
RTOFS-HWRF Track Forecast for 6 TCs (186 cases): Gert07L, Irene09L, Katia12L, Maria14L, Ophelia16L, and Philippe17L HY11=HyHWRF2011; HWRF=operational HWRF • HyHWRF exhibits comparable performance with HWRF for average track error and standard deviation (STD), except STD outperformance at 96 -120 h. GFDL shows mixed comparison but the best STD. • HyHWRF track bias is the same northeastward as HWRF, but the bias magnitude is better than the HWRF. GFDL shows west/northwestward bias. 21
RTOFS-NEMS • Working on 0.25 degree HYCOM for coupling to weather models (GFS, CFS). • Model setup and free runs successful. • Moving to coupled test version with GFS in NEMS in coming year. Year 5 day 2 Year 5 day 180
Coupling • Some additional considerations on coupling • Moving to hurricane coupling including waves. • Many more exchange parameters, sea spay to constrain fluxes. • Ice model for weather time scale critical for successful weather time scale coupled ocean-atmosphere model. • Additional focus on wave-surge model. WAVEWATCH III – ADCIRC 1-way coupled NCEP – LSU – ND collaboration URI coupling approach (I Ginis)