Point source for hydrothermal venting

[JMette]

Hi all,
I'm trying to set up a point source for modelling the inflow of a hydrothermal vent.
Starting from the Riverplume example, I have some questions about the variables to set in ana_psource.h.
1)If

Code: Select all

Nsrc(ng)=Nsrc(ng)+10                		! Add rainfall location
          DO is=2,6
            SOURCES(ng)%Dsrc(is)=2.0_r8       	! vertical influx, LwSrc=T
            SOURCES(ng)%Isrc(is)=6            	! i = 6
            SOURCES(ng)%Jsrc(is)=is+25        	! j = 27 to 31

denotes a vertical downward influx from rainfall, how is it possible to set upward influx? Is it somehow possible with LwSrc?

2)How is it possible to set the depth of the point source, to have influx only from the bottom? The rainfall example has no specified vertical coordinate.
Would it be similar to the other coordinates:

Code: Select all

SOURCES(ng)%Ksrc(is)=SOMEVALUE  

3)

Code: Select all

#if defined RIVERPLUME1

denotes the case of source the model uses, right? Is the variable "RIVERPLUME1" set by the application name?

4)Are the variables in ana_psource.h still used if a netcdf input is specified in roms.in or is this file than irrelevant? Because the Riverplume1 example seems to use both.

5)Are analytical point sources still problematic because of parallel bugs as stated here: https://www.myroms.org/wiki/River_Runof ... _psource.h
So would you recommend using a netcdf file or the analytical input for an application with one point source with upward inflow?

Thanks in advance for your help!

[wilkin]

(1) The Dsrc = 2 option is not really a vertical flux. It is a flux divergence imposed at the rho-points cell center. So, as such it is neither upward nor downward, it just appears. This is what you want if your vent is adding buoyancy but not momentum.

(2) The fractional distribution of the mass flux into vertical cells is set by Qshape. Look for Qshape examples in ana_psource.h. If you want all the flow to go into the bottom-most cell (k = 1) then you want something like:

Code: Select all

          SOURCES(ng)%Qshape(Nsrc(ng),1)=1.0_r8
          DO k=2,N(ng)
            SOURCES(ng)%Qshape(Nsrc(ng),k)=0.0_r8
          END DO

(3) The test cases are set by options such as RIVERPLUME1 which would mean you have, in your build_roms.sh script ...

Code: Select all

ROMS_APPLICATION=RIVERPLUME1

which causes ROMS to activate all the options in Include/riverplume1.h

(4) ROMS reads sources from a netcdf file if you have #undef ANA_PSOURCE (and ignores ana_psource.h). You will see in the report at the end of the run that Functionals/ana_psource.h is not in the list of Functionals used.

(5) If you follow the examples in ana_psource.h you should be able to configure the analytical inputs correctly. If you get it wrong you will know because nothing will happen.

[JMette]

Thank you very much for the help. I will have a look in the NetCDF way of point sources and will then decide, which one I will use.