[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[IDV #VGT-219571]: That pesky 0 Longitude



Hi Martha-

> Full Name: Martha
> Email Address: address@hidden
> Organization: Oregon State U.
> Package Version: 2.7u1 build date:2009-07-14 19:24 UTC
> Operating System: Windows XP
> Hardware: Java: home: C:\Program Files\IDV_2.7u1\jre version: 1.6.0_13 
> j3d:1.5.2 fcs (build4)
> Description of problem: Hi Don,
> 
> Crossing the 0 longitude (the area I'm working on is Lat -30 to -60; Lon 330 
> - 30 (or -30 to 30).
> 
> I've sent you a bundle but contend it would be better for you to just grab 
> the data yourself because it will show you how it's impossible to use the 
> region option to try things and how the bounding box shifts around.
> 
> If you access: Catalogue - NOAA Oceanographic Data - Coast Watch (2nd one) - 
> Satellite DataSets - Color - Chlorophyll-a, OrbView-2 SeaWIFs,Global - monthly
> you will find the gridded data I've been trying to cross the 0 longitude. 
> Just grabbing one time 8/16/2002 show enough.
> 
> I can only use the properties to set the area and have found that:
> Request longitude 330 - 30 will only provide 0-30
> 330 -  0 gives you 0 - 330
> 330 - 360 gives you 0 -  330
> -30 -  0 gives you 0 - 330
> 
> However, 330-359 works and requesting 0-30 filles in on the other side with 
> no data in the center.

The problem is that the data in the original dataset goes from 0-360 and the 
underlying netCDF-Java library does not allow subsetting across the grid 
bounds.  Thus, we can't make a request to satisfy what you want in the IDV.

However, it looks like the Live Access Server knows how to stitch together a 
cross seam request and can output a netCDF file that the IDV can read.  I got a 
file from:

http://las.pfeg.noaa.gov/oceanWatch/oceanwatch.php

and made the request to be -30 to 30.   Unfortunately, I haven't had similar
success trying to download a lat-lon region vs. time.  That doesn't seem to 
subset
correctly.

> I have had other various problems trying to cross the 0 and 180 and will send 
> you examples of those as they occur.

If you have some that use to work, but don't now, that would be helpful.  I made
a change the might fix those, but don't have a test case yet.

> Thanks so much for your attention to this issue. I'll be working with the 
> "spliced" sets for analysis but would love to have the capability of a 
> contiguous set for animations.

Until we have support in the underlying netCDF layer, I don't see this 
happening in the IDV side.  I'll talk to the developer again.

Don

Ticket Details
===================
Ticket ID: VGT-219571
Department: Support IDV
Priority: Normal
Status: Open