How to properly set time stamp parameters

Discussion of how to use ROMS on different regional and basin scale applications.

Moderators: arango, robertson

Post Reply
Message
Author
lcbernardo
Posts: 88
Joined: Wed Oct 01, 2014 8:57 pm
Location: International Coastal Research Center

How to properly set time stamp parameters

#1 Unread post by lcbernardo »

Hi everyone,

I have been able to successfully run ROMS but have gotten a little bit confused with the time stamp variables DSTART, TIDE_START, and TIME_REF.

I created a boundary conditions file which runs from May 15 to September 30, 2013, with a daily time step. Suppose I want to start the model from say August 15 instead of the first data available in the boundary file, what would be the proper way to set the time variables? In my present configuration, I have:

DSTART = 106.0d0 ! days
TIDE_START = 106.0d0 ! days
TIME_REF = 20130815.00 ! yyyymmdd.dd

I set this to correspond to my preferred start time, which is Aug. 15, 2013 00:00, but upon comparing the model output with observed data such as tide levels, I have a large mismatch with a phase difference of a week or so. Can anyone help me identify how to properly set the time variables for this run?

Thanks,
Lawrence

User avatar
kate
Posts: 4088
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: How to properly set time stamp parameters

#2 Unread post by kate »

There are two time origins, one for the forcing/boundary files and one for the tides. If your boundary file has time zero on May 15, then how you have set TIME_REF is correct. Also DSTART. Now you need to investigate time in your tides file. Did you have the right phase when starting on May 15? What was your TIDE_START then? If things worked well for that situation, don't change your TIDE_START.

We have picked the start of 1900 as our time origin for all forcing/boundary files for all our domains. That way, we can use the same global winds, etc., for all of them. All that changes from one run to the next is DSTART, which can have numbers like 40,000. How the tides relate to that depends on your source and the software used to create the ROMS tides file.

lcbernardo
Posts: 88
Joined: Wed Oct 01, 2014 8:57 pm
Location: International Coastal Research Center

Re: How to properly set time stamp parameters

#3 Unread post by lcbernardo »

Thanks for the reply Kate! At least now I know that TIME_REF and DSTART are set properly.

As for the tides, I followed the instructions (with a few modifications due to the differences in matlab versions, etc.) described for OTPS tides in https://www.myroms.org/wiki/index.php/Tidal_Forcing. The main matlab script to generate the tidal forcing seems to be otps2frc_v3.m or otps2frc_v4.m. There's a part of this script that shows:

base_date=datenum(2013,5,1); % Reference Tidal Date for simulation
pred_date=datenum(2013,7,20); % Correction time

For a run that would start on May 1, 2013 running all the way to September, I assume this would be the correct specification. But if I were to start the run on Aug. 15, 2013, would this require modification? Or can the adjustment be made in the ocean.in file directly?

Post Reply