Erorr when applying rivers

[javadi]

Dear users

I have three rivers in my domain. I have made my river runoff file but I have faced this error. I am not sure about my input file as I do not define v-shape; is it essential? Here are my runoff.nc and this error.

please help me with your interesting comments

[kate]

You could well be dividing by zero because it doesn't have the Vshape. Yes, you need it to be non-zero.

[javadi]

Dear Kate

Thank you so much for your response. Based on your comments, I conclude that I should make vshape variable. It should be the number of my vertical level. The sum of all values should be one. Here the river_transport variable is discharge(?) and v_shape divides river_transport into different layers. is it correct?

[kate]

Yes, exactly. How much do you want shallow and how much deep? For the coastal Gulf of Alaska, I use a linear profile, max at the surface, zero at the very bottom.

[javadi]

Dear Kate

I have changed my river_runoff.nc based on your suggestion. Here is the new one. However, I have faced this error which is related to river variable. Although I have input river, the model cannot find this variable. Please note that my ROMs is a new version.

[kate]

Here's what one of my rivers files looks like:

Code: Select all

netcdf JRA-1.4_Arctic_rivers_1980 {
dimensions:
	river_time = UNLIMITED ; // (366 currently)
	river = 18430 ;
	s_rho = 50 ;
variables:
	double river_time(river_time) ;
		river_time:units = "days since 1900-01-01 00:00:00" ;
		river_time:long_name = "river runoff time" ;
	int river(river) ;
		river:long_name = "river runoff identification number" ;
	double river_sign(river) ;
		river_sign:long_name = "river directional sign" ;
	int river_Xposition(river) ;
		river_Xposition:long_name = "river XI-position at RHO-points" ;
		river_Xposition:valid_min = 1 ;
		river_Xposition:valid_max = 689 ;
	int river_Eposition(river) ;
		river_Eposition:long_name = "river ETA-position at RHO-points" ;
		river_Eposition:valid_min = 1 ;
		river_Eposition:valid_max = 1089 ;
	int river_direction(river) ;
		river_direction:long_name = "river runoff direction" ;
	double river_Vshape(s_rho, river) ;
		river_Vshape:long_name = "river runoff mass transport vertical profile" ;
	double river_transport(river_time, river) ;
		river_transport:long_name = "river runoff vertically integrated mass transport" ;
		river_transport:units = "meter3 second-1" ;
		river_transport:time = "river_time" ;

I have the tracers in a second file that doesn't depend on the year. Here, the variable "river" is a "dimension variable" and is just a list of integers, 1 through river. The discharge is in the river_transport variable.

[wilkin]

There is a CDL template rivers file in the source code distribution at: Data/ROMS/CDL/frc_rivers.cdl

If you build your file like this it should work with the myroms.org code.

The fact that your file is named croco_runoff_monthly.nc makes me think you are using a different version of ROMS. If so, you should mention that in your post so that we aren't trying to solve version inconsistency issues for you.

[javadi]

Dear kate and wikin

Thank you so much for your suggestions. ROMS is sensitive to order in arguments and it was my problem. Now my model accepts my river file. I use the new version of ROMS(Rutgers). Now my model runs with my river files when I try to run the river_plume1 test case. However, I have faced a new issue in the result. The river does not come to the domain. The model does not show any action to rivers in outputs. I have revied the forum and I guess that my problem can be because of the definition of source inland(i and j). I do not understand some points. Firstly, about river_direction, I do not understand exactly how to define values for this. In my domain, I have three rivers where one of which is north-south and the others are west to east. Secondly, x-position and Eta_position, here I plot my grid file and the locations of rivers. For my rivers what are the values of direction, x-position, and E_position? I have marked the point source in my grid when I applied the model.

[wilkin]

Did you read this:
https://www.myroms.org/wiki/River_Runof ... _Eposition

When using option LuvSrc = T, river_Xposition and river_Eposition refer to the i,j index of the u-face or v-face the flow crosses - NOT the i,j index of the rho cell it flows into. The i,j values must follow ROMS Fortran numbering convention for the appropriate u-point or v-point on the ROMS staggered grid.

The ROMS FORTRAN numbering convention means rho points start from 0,0 in the "lower left corner", not from 1,1 (like Matlab would, e.g.).