Plotting vector velocity u and w and bedload_coeff

General scientific issues regarding ROMS

Moderators: arango, robertson

Post Reply
Message
Author
luhputria
Posts: 14
Joined: Tue Mar 31, 2020 3:43 am
Location: The University of Western Australia

Plotting vector velocity u and w and bedload_coeff

#1 Unread post by luhputria »

Hi all,

I have some questions regarding ROMS:
1. I can make vector velocity of u-eastward and v_northward because they have same position. Do you have any idea how to plot vector velocity for u and w? Because they have different position, u (x_u,y_u,s_rho,ocean_time) and w (x_rho,y_rho,s_w,ocean_time). Can I interpolate the result in my output file (roms_his.nc)?
2. Just to make sure that I understand ROMS and read the code correctly, is it correct if I choose UV_LOGDRAG the bottom roughness (ZOB) will be constant as we stated in input file while if we choose SW_BBL/SG_BBL/MB_BBL, the bottom roughness total is Z0S+S0ST+S0BF?
3. In sediment input file, how we can define value of bedload_coeff (Bed load transport rate coefficient)? In several test case the value is 0.05. While when I check croco manual documents (https://croco-ocean.gitlabpages.inria.f ... _usgs.html), bedoad_coeff is set 0. Can I set bedload_coeff=0? If I set 0, how about the bedload because in ROMS/Nonlinear/Sediment/sed_bedload.F, bedload_coeff affect FX_r.
FX_r(i,j)=FX_r(i,j)*bedload_coeff(ng)*bed_frac(i,j,1,ised)
I also try to change this bedload_coeff with the reason, different researchers propose different equation for obtaining the bedload transport rate (from van Rijn, Bagnold, etc.) and change my sandwave migration (my study) significantly as shown in figure below.
thumbnail_image002.jpg
thumbnail_image002.jpg (14.57 KiB) Viewed 2174 times
PS: I use ROMS but sometimes I check croco manuals because as far as I know, croco is also developed from ROMS

jcwarner
Posts: 1172
Joined: Wed Dec 31, 2003 6:16 pm
Location: USGS, USA

Re: Plotting vector velocity u and w and bedload_coeff

#2 Unread post by jcwarner »

sorry for late reply.
1. i think u-east and v-north have been averaged to rho points (?). but normally the u- and v- points are on a staggered grid, see here:
https://www.myroms.org/wiki/Fractional_ ... %B7_space) and similar pages.
you can interpolate the velocities to the cell centers.
2. uv_logdrag imposes a constant zo, unless you also specify # define UV_DRAG_GRID, then you can vary the Zo.
for ssw_bbl etc, you need to specify the flags to have the roughnesses added. best for you to look in the code ROMS/Nonlinear/ssw_bbl.h and see:

#ifdef SSW_CALC_ZNOT
! Calculate components of roughness and sum: zo = zoN + zoST + zoBF
.....
zo=zoN(i,j)
# ifdef SSW_ZOBL
zo=zo+zoST(i,j)
# endif
# ifdef SSW_ZORIP
zo=zo+zoBF(i,j)
# endif
# ifdef SSW_ZOBIO
zo=zo+zoBIO(i,j)
# endif
#else
zo=zoDEF(i,j)
#endif

3. yes you should be able to set bedload coeff =0. There are many bedload formulations, this is not an exact science. there are many factors that influence sed transport such as variations in grain size and flow turbulence.

Post Reply