1. US GLOBEC-IV-B: Interannual Variability of the Circulation in the Gulf of Maine/Georges Bank: A Hindcast FVCOM Experiment for 1995-2006 C. Chen, R. C. Beardsley, C. Davis and G. Cowles Participators: Q. Xu, S. Hu, D. Stuebe, J. Qi, H. Huang, D. Limeburner, R. Ji

3. Regional domain MM5 (resolution: 30 km) Non-hydrostatic Wind stress, heat flux, air pressure, precipitation via evaporation Local domain MM5 (resolution: 10 km) ETA (32 km) Hydrostatic Air stations Buoy data Satellite data Local network SST (9 km) Local buoys Assimilation Nested The Gulf of Maine/Georges Bank MM5 Weather Hindcast/Forecast Model System Local domain Regional domain Sites of Surface Buoys 1.Hindcast assimilation for years 1978-2006 2.On-line forecast system

4. Phytoplankton P(i), i=1:np Light/temperatureAdvection/mixingRiver dischargesBoundary flux Y(i), i=1: ny (Auxiliary state variables) Nutrients N(i), i=1:nn Detritus D(i), i=1:nd Zooplankton Z(i), i=1:nz Bacteria B(i), i=1:nb DOM(i), i=1:nm

5. First Generation Second Generation Horizontal Resolution: 0.5-1.0 km in the coastal region 31-sigma levels in the vertical Horizontal Resolution: 0.3-1.0 km in the coastal region 31-sigma levels in the vertical

6. 256 Processors (Intel 3.4 GHz Pentium 4) 256 Gigabytes RAM, Infiniband High Speed Network 7 Terabytes disk space Third generation of GOM FVCOM Horizontal resolution: 10 -500 m in the coastal region

8. Boston harbor area

9. Pre-processing of forcing data for the Gulf of Maine Model Local forcing: meso-scale surface wind stress and heat flux; Remote-forcing: upstream open boundary on the Scotian Shelf; Assimilated with the satellite-derived SST, hydrographic and current data;

10. Short-wave Long-wave Latent Sensible

11. The TOGA/COARE (TC) heat flux algorithm developed by Fairall et al. (1996, 2001): More accurate skin water temperature; Better paramterization of surface roughness; More realistic vertical profiles for stable and unstable weather conditions Accountable wind gustiness

12. Observed MRF-TC

13. Years: 1995-1999

14. Years: 2000-2006

15. March 1995-2000 on the Scotian Shelf Boundary

16. March 2001-2006 on the Scotian Shelf Boundary

17. No boundary T/S forcewith boundary T/S force August 95

18. RMS [Data - OA] Miss Fit Objective Analysis of SST: GOM (1985-2000) Results for 1996 and 1997 Time series results at three locations:

19. SST OA time series Shelf Break Direct estimate of spatial correlation 31 31 grid cells = 270km 3 OI iteration-reach the convergence

20. SST OA time series Georges Bank Direct estimate of spatial correlation

21. SST OA time series Gulf of Maine Direct estimate of spatial correlation

23. Annual Net Heat Flux

24. Mean Net Heat Flux for January-March

26. March 1998 March 1999

27. March 1998 March 1999

28. http://fvcom.smast.umassd.edu/research_projects/GB/index.html

34. FVCOM Model Validation: Comparison with CMO Observations Time Frame: Aug 1996 to Jun 1997 Location: New England Shelf, SW of Nantucket # Moorings: 4 (Central, Alongshore, Inshore, Offshore) Oceanic Variables: velocity, temp, conductivity, bottom pressure, bottom stress. Atmospheric Variables: wind stress, heat flux, buoyancy flux. Steve Lentz’s data Cowles and Lentz’s work

35. Temperature (Celsius) during CMO Period at Central Site August 1996 - June 1997 Observed FVCOM

36. Subtidal Surface Current During CMO: Observed (red) FVCOM (black) Along Isobath (Above, ~east) and Cross Isobath (Below, ~north)

38. On-going Physical Model Works Conducted the assimilation experiments using the OI and Ensemble Kalman Filters: [Rizzoli (MIT), Beardsley (WHOI) and Chen (UMASS)] Reduced Rank Kalman Filter (RRKF) Ensemble Kalman Filter (EnKF) Ensemble Square-Root Kalman Filter (EnSKF) Ensemble Transform Kalman Filter (EnTKF) Upgrade the model to the third generation version and include the monthly volume transport on the Scotia slope: Generalized topographic coordinate FVCOM to have uniform-thickness lay ers in the upper mixed layer and bottom slope.

39. Test problem 1: Tidal oscillations in a flat-bottom circular basin Conditions: Linear, 2-D, no friction, shallow water R (radius) = 50 km Forcing: M 2 tidal elevation at the open boundary Case 1: Normal oscillation H= 10 m, M 2 tidal elevation at OB = 1 m Open boundary Case 2: Near-resonance oscillation H= 1 m, M 2 tidal elevation at OB = 1 mm

40. Sensitivity Experiments for Measurement Sites -5 0 5 (  10 4 ) e5 e4 e3 e2 e1 Surface elevation experiment (  10 4 ) 5 0 -5 -5 0 5 (  10 4 ) c10 c9 c8 c7 c6 Current experiment (  10 4 ) 5 0 -5

41. RRKF: RMS error analysis of the elevation experiment Forecast innovation based on previous analysisAnalysis innovation Full Forecast based on previous analysisFull Analysis

42. RRKF: RMS error analysis of the current experiment Forecast innovation based on previous analysisAnalysis innovation Full Forecast based on previous analysisFull Analysis

43. EnKF: Near-resonance case: 1 current meter mooring at #1 site Forecast error at the observational site Analysis error at the observational site Forecast RMS errorAnalysis RMS error

44. Ship-tracking (e.g. ADCP) True Model RMS