NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
Hi Jeremy, NCSS actually already supports the subsetting-by-vertical-level operation that you describe. Here's an example: [1]. I don't know which version specifically added the capability, but I know that 4.3.15 (which is running on your server) doesn't have it, and 4.3.22 and later do. So you may need to upgrade. Lat-lon subsetting your example is difficult not only because the grid lines don't align, but also because the result must still be a rectangle. I think we did as well as you can hope for with our implementation, given the constraints, but we may consider allowing the user to specify arbitrary, polygonal bounds in the future. Cheers, Christian [1] http://thredds.ucar.edu/thredds/ncss/grib/NCEP/GFS/Global_0p5deg/GFS_Global_0p5deg_20140717_0000.grib2/GC/dataset.html On Tue, Jul 29, 2014 at 9:38 PM, Jeremy Cothran <jeremy.cothran@xxxxxxxxx> wrote: > I've been working with some ROMS model output at the below link, trying to > do some climatology oriented averages, a part of ncss I wished worked > better is that when I use ncss to get say temperature(temp) at the url > below there's no way to specify a single s_rho elevation like surface(s_rho > index = 35) or bottom(s_rho index = 0). I'm able to select out just a > particular s_rho afterwards using NCO, but would be nice if ncss allowed > subsetting on the all the netcdf file indexes the same as opendap. > > > http://omgarch1.meas.ncsu.edu:8080/thredds/ncss/grid/fmrc/sabgom/SABGOM_Forecast_Model_Run_Collection_best.ncd/dataset.html > > Also this model grid for the US Southeast is at approximately 45 degrees > to the equator, and ncss lat/lon bounding box seems to select parallel to > the model grid within the bounding box and not the full lat/long(so like a > smaller parallel box within the larger specified ncss bounding box) - can > still do what I need to, just guessing that for this to work correctly, > model grid domains need to be parallel to equator and not at an angle. > > Thanks > Jeremy Cothran > University of South Carolina > SECOORA/IOOS > > > _______________________________________________ > thredds mailing list > thredds@xxxxxxxxxxxxxxxx > For list information or to unsubscribe, visit: > http://www.unidata.ucar.edu/mailing_lists/ >
thredds
archives: