Custom Query (964 matches)

Corrected minor typos and bugs:

• Corrected CPP directive PC02AIR_DATA to PCO2AIR_DATA in checkdefs.F.

• Corrected the reporting of Phase-1 initialization in driver jedi_roms.h.

• Corrected CPP directive in ad_set_avg.F, the if command was missing in

  #  if defined BULK_FLUXES && defined EMINUSP
  

The ESMF coupling was updated to enhance the NUOPC layer modules for ROMS and WRF. In addition, several new exchanged fields are now supported.

• WRF now exports surface air density (RhoAir), level 2 winds (Uwind_sbl, Vwind_sbl), and turbulent frictional wind magnitude (Wstar), so ROMS NUOPC cap file can compute the surface wind stress when BULK_FLUXES is off during coupling. In particular, we added an option to calculate the reference wind force (wind minus current) to compute the surface wind stress into the ocean when WIND_MINUS_CURRENT is activated.

The surface wind stress is computed as:

τ_x/ρ_w = ρ_a C_D |W_r| u_r

τ_y/ρ_w = ρ_a C_D |W_r| v_r

Where ρ_w is the mean seawater density, ρ_a is the surface density of air, τ_x and τ_y are the wind stress components, C_D is the transfer coefficient for momentum (drag coefficient), u_r and v_r are the wind components relative to the seawater (wind minus surface ocean current), and |W_r| is the relative wind magnitude.

Here, the drag coefficient is computed as:

C_D = W^*2 / |W_a|²

Where W^* is the frictional wind magnitude from similarity theory (WRF variable grid%ust) and W_a is the unreferenced surface wind magnitude, SQRT[(u_a)²+(v_a)²].

The coupling is incomplete because there is no feedback on the atmosphere. ROMS surface currents do not modify the near-surface wind in WRF. We are looking for ways to do so without changing WRF.

• ROMS now exports free-surface (SSH), barotropic momentum (ubar, vbar), surface 3D momentum (Usur, Vsur), and bathymetry (bath) to other coupled component. Recall that bathymetry can be time-dependent if sediment morphology is activated.

• ROMS NUOPC module now includes a generic ROMS_Rotate routine to rotate vector from computational grid to geographical EAST and NORTH directions and vice versa.

• As shown in Renault et al. (2016), the feedback from surface ocean currents to the atmosphere is an eddy killing effect that stabilizes the Gulf Stream separation at Cape Hatteras in WRF-ROMS fully coupled applications. The result is around 0.3 N/m² weaker surface wind stress in the core of the Gulf Stream

Below are the results for Hurricane Irene solutions with the wind stress computed in the WRF NUOPC cap module (solution A) versus the one calculated in the ROMS NUOPC cap module with the WIND_MINUS_CURRENT option (solution B). The plots show the differences in the wind stress component (B-A solutions).

The surface wind stress streamlines for solutions A and B are shown below, respectively:

Notice that the difference in the maximum values is 0.28 M/m² weaker in the solution with the reference wind relative to seawater (wind minus currents).

• The application IRENE input scripts were updated to compute the WIND_MINUS_CURRENT application. See coupling_esmf_atm_sbl_wmc.in input script.

Many thanks to John Wilkin for his help in coding and analyzing the results.

References:

Edson, J.B., V. Jampana, R.A. Weller, S.P. Bigorre, A.J. Plueddemann, C.W. Farall, S.D. Miller, L. Mahrt, D. Vickers, and H. Hersbach, 2013: On the Exchange of Momentum over the Open Ocean, J. Phys. Oceanog., 43, 1589-1610, doi:

Renault, L., M.J. Molemaker, J. Gula, S. Masson, and J.C. McWilliams, 2016: Control and Stabilization of the Gulf Stream by Oceanic Current Interaction with the Atmosphere, J. Phys. Oceanog., 46, 3439-3453, doi: 10.1175/JPO-D-16-0115.1

• Corrected a small bug in the implementation of the point sources as a volume vertical influx (LwSrc) in step3d_t.F and TLM, RPM, and ADM versions. We need to use the point source location (Isrc, Jsrc) when computing the area scale:

                      cff=dt(ng)*pm(Isrc,Jsrc)*pn(Isrc,Jsrc)
  

  Many thanks to John Warner for ​bringing this to my attention.

• Added the -liomp5 library to Compilers/Linux-ifort.mk configuration file. It is required by newer versions of ifort.

• Corrected missing arguments in convolve.F. Many thanks to Stefan Riha for ​reporting this problem.