gst03_v1_201120 =================== Date: Dec/01/2020 Description: ------------- New Gulf Stream domain ROMS grid (dated 2020/Nov/20 --> 201120) for E-CESM runs, based on 1/10 degree POP grid. POP Grid -------- Reference 1/10 degree POP Grid from Brian: Cheyenne: /glade/work/jaisonk/WORK/Models/POP/run/g.e20.G.TL319_t13.control.001/ - domain.ocn.tx0.1v2.161014.nc - g.e20.G.TL319_t13.control.001.pop.h.nday1.0020-02-01.nc ROMS Grid --------- POP I-J box for the ROMS domain: I=285:553, J=1412:1717 Method: define ROMS RhO-grid points based on POP-grid-cell edges. Resolution: approximately 3-km. Each POP grid-cell is divided into 9 ROMS grid cells (ROMS_dx=POP_dx/3, ROMS_dy=POP_dy/3). Bathymetry: from SRTM30 --> which is qualitatively very similar to the POP bathymetry. At the 4-edges of the ROMS domain, the SRTM30 bathymetry smoothly blends into the POP bathymetry over a 0.75 degrees wide region (so that right at the edge, ROMS bathymetry is identical to that of POP and 0.75 degree away from the edge, the bathymetry is the original SRTM30). Bathymetry Smoothing: After the blending process, bathymetry has been smoothed to reduce the ROMS slope parameter rx0 to 0.25 and rx1 to 12. Coastline/Land-sea mask: Follows that in the POP 1/10 degree grid. Minimal editing has been done to make sure that the ROMS grid points exist only between active POP grid points. ROMS Vertical Coodinate parameters are the same as before: - Theta_s = 10 - Theta_b = 2 - hc = 400 - Tcline = 400 - N = 50 ROMS ocean.in file have some diffusivity coefficient like tunable values which are grid dependent. I typically use similar values for all ROMS configurations with similar spatial resolution. You can either use values from the old 3-km ocean.in file or I can help to get a new ocean.in file just for this grid. ROMS ini/bry ------------ POP Restart File used for preparing ROMS ini and bry: g.e20.G.TL319_t13.control.001_hfreq.pop.r.0053-01-01-00000.nc from Brian Time: Please note that the year 53 in POP-restart is the calendar year 2010. So all the ROMS input files are labelled as 2010 but the netcdf file time points to year 53. Ini: Interpolated above POP restart data on to ROMS grid. Bry: Extracted data along 4-edges from the above initial condition file and repeated it for few time values so that the boundary condition file can be used for making test runs. ISSUE: The original POP grid (see above) have missing values over land for coordinate variable like TLONG, TLAT, ULONG and ULAT. This is bit strange compared to previous version of POP configurations. Howver, the domain file have TLONG and TLAT for both ocean and land points, and i have used this data/variables to define ROMS grid. For proper interpolation from POP B-grid to ROMS grid, we need ULONG and ULAT of POP-grid with values over both ocean and land. Since it is not available, i have used the same TLONG and TLAT for all POP variables (basically treating it as an A-grid instead of a B-grid). This can be corrected in the future if ULONG and ULAT is available over land grid-points too. Files on Cheyenne ----------------- Dir: /glade/work/jaisonk/pub/E-CESM/ROMS_gst/ggs03_v1_201120 ROMS Grid file: gst03_v1_201120_grd.nc ROMS Ini file: gst03_v1_201120_ini_TL319_t13_Agrid_20100101.nc ROMS Bry file: gst03_v1_201120_bry_TL319_t13_Agrid_20100101.nc ROMS Domain Fig.: fig_gst.png