time parameter in the ocean.in file

[Chang]

Hello everyone
I am confused with the three parameter (dstart,tide_start,time_ref)in the ocean.in file.Could anybody there help me explain them for me?
Thanks in advance.

[kate]

In our realistic runs, we pick a time origin of 1, Jan, 1900. That becomes time_ref. dstart is how much later this particular run is going to start. tide_ref is the offset between the tidal time origin and your time_ref. It doesn't matter what time origin you pick, but you have to be consistent between all your forcing, boundary, initial files.

[Chang]

Hello kate:
I sincerely appreciate your help.
I post several pictures as follow which indicate how i defined the attribute of ocean_time,bry_time,forcing_time in my netcdf input files.As you said, i should make this several time be consistent. whether i define the 'units' of 'ocean_time' to 'second since 2016-04-18 00:00:00' or not?i think it's maybe a mistake due my careless.
Whether The 'base_date'(2016-01-01) in picture 'tide_base_date.jpg' is my tide origin time? if the answer is 'yes',i think i should define TIDE_START= -108.0d0 (tide base time - time_ref).
The values of time(including ocean_time,bry_time,forcing_time) are all set from 0. So i prepare to set DSTART=0.0d0 and TIME_REF = 20160418.0d0. I don't know whether my thought is right or not?
you said you picked a time origin of 1, Jan, 1900. Does that mean you will set TIME_REF to 19910101.0d0? I guess the value of time in your input files are not set from 0,does that right?
I hope my poor english won't confuse you.
Thanks in advance.

[kate]

Like much of modeling, you have to try it. It seems reasonable, does it work and give sensible results?

Yes, my dstart is often somewhere from 35000 to 40000 - number of days from 1900 to the start time.

[Chang]

hello Kate
thanks for your help.