Doubts about Seagrid

General scientific issues regarding ROMS

Moderators: arango, robertson

Post Reply
Message
Author
chaos4u2
Posts: 26
Joined: Sun Oct 30, 2011 3:12 pm
Location: UPC

Doubts about Seagrid

#1 Post by chaos4u2 » Tue Mar 05, 2013 5:13 pm

I have a bathymetry file way too big, about 2 million data, and when I use the appliaction seagrid in matlab it almost blows up and goes very very slow.

My question is how would you do to have less data. I would take 1 of each 4 data value for example; but there must me a better way to do it. Isn't it?

And 2nd question: when I have my grid, how do you tell ROMS that in your domain everything is closed except 2 zones. I know there's the lateral boundary conditions, but they are north, south, etc... How do you concretise the coordenates of your domain where currents exists?


Thank you in advanced!!!

lanerolle
Posts: 157
Joined: Mon Apr 28, 2003 5:12 pm
Location: NOAA

Re: Doubts about Seagrid

#2 Post by lanerolle » Wed Mar 06, 2013 12:38 am

What we do is to generate a grid via Seagrid with a default/dummy bathymetric value.

Thereafter, we take the grid NetCDF file and write the bathymetry into it via say Matlab as a separate procedure. I have dealt with way more than 2 million points (may be x 10 or x 50 more than that) and loop over the wet-points of the ROMS grid and interpolate the bathymetry via an inverse distance squared algorithm which only looks at the soundings within a given radius around the grid point in question - you can do this Matlab via the inpolygon function.

Hope this helps.

chaos4u2
Posts: 26
Joined: Sun Oct 30, 2011 3:12 pm
Location: UPC

Re: Doubts about Seagrid

#3 Post by chaos4u2 » Sat Mar 09, 2013 8:31 pm

Thank you for responding!!

Another question: I have 2 files of coastline. Seagrid doesn't allow me to load both at the same time. When I charge one, the other disappears. So I put it together both in one file. But when I charge it, there's a line between the last point of the first file and the first point of the second file.

How can I charge both files without joining this 2 different figures?

miguel.solano
Posts: 11
Joined: Mon Jan 30, 2012 8:02 pm
Location: University of Texas at Dallas

Re: Doubts about Seagrid

#4 Post by miguel.solano » Sun Mar 10, 2013 5:46 am

I have never tried to load two coastlines at the same time in Seagrid, but I do know that coastline files are just a longitude, latitude pair where coastline segments are separated by Nans. So I would just try to separate file 1 and 2 by a pair of Nans so that Seagrid knows these are separated coastline segments.

Hope this helps

chaos4u2
Posts: 26
Joined: Sun Oct 30, 2011 3:12 pm
Location: UPC

Re: Doubts about Seagrid

#5 Post by chaos4u2 » Sun Mar 10, 2013 10:37 am

Thank you very much. It worked as you said with the NaNs.

User avatar
kate
Posts: 3780
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: Doubts about Seagrid

#6 Post by kate » Mon Mar 11, 2013 12:55 pm

chaos4u2 wrote:And 2nd question: when I have my grid, how do you tell ROMS that in your domain everything is closed except 2 zones. I know there's the lateral boundary conditions, but they are north, south, etc... How do you concretise the coordenates of your domain where currents exists?
If you have an open segment on a boundary that is otherwise closed, you tell it that it's open, then mask out some of the boundary points with the land mask.

Post Reply