search

NOAA-PMEL / LAS

Live Access Server
https://ferret.pmel.noaa.gov/LAS/
The Unlicense
13 stars 5 forks source link

ArcView gridded script must detect and deal with curvilinear data #428

Open karlmsmith opened 6 years ago

karlmsmith commented 6 years ago

Reported by steven.c.hankin on 26 Feb 2008 19:29 UTC Currently the arcView script fails with the Ferret script error message:

 **ERROR: command syntax: plot_var , xcoord, ycoord
Definition contains multiple expressions
DEFINE VARIABLE expr = plot_var , xcoord, ycoord

Since the arcView gridded format is not suitable for curvilinear data, the script should (minimally) detect this fact and give out a clear error message to this effect, such as

arcView gridded output cannot be created.
The arcView format cannot represent the curvlinear coordinate system of this dataset.

For the special case of tri-polar grids an improvement to this would be if, say, prior to issuing the error message it looked for a the existence of a property, say, $FER_TRI_POLAR_LIMIT=60

If the property does not exist, then a curvilinear dataset gives the error message just described. If it does exist, however, then the script checks to see if the northern limit of the requested region is less than the value of $FER_TRI_POLAR_LIMIT. If it is less, then the script alters the symbol definitions so the request can be treated as a standard rectilinear dataset. If it exceeds the tri-polar limit, then it issues a variation on the error message above, such as

arcView gridded output cannot be created for this data north of latitude 
($FER_TRI_POLAR_LIMIT).  Latitudes north of this limit utilize curvilinear 
coordinates that the arcView format cannot represent.

Jing -- I have for the moment assigned this ticket to you. Please see if you can knock it off as a part of the HYCOM server installation. If not, then pass it back to Ansley.

Migrated-From: http://dunkel.pmel.noaa.gov/trac/las/ticket/422

karlmsmith commented 6 years ago

Comment by @AnsleyManke on 26 Feb 2008 20:39 UTC The simple part of this is done.

I've fixed the arcView script so that it checks to see if the data is curvilinear. This is known by the scripts because that is part of initializing all datasets/variables.

If the data is on a curvilinear grid, the script puts out a message to std error and exits from Ferret.

For the idea of a part of the grid in which the coordinates are not curvilinear, and where we might be able to do arcView listings, there's a second issue to consider. arcView output is only for regularly-spaced data, and the tri-polar grid is irregularly spaced in latitude, with the grid finer near the equator. So we can't do arcView output for that reason either - I don't know if there are portions of the grid that are regularly spaced or not.

karlmsmith commented 6 years ago

Modified by @kevin-obrien on 5 Jan 2011 23:37 UTC