Thanks a lot. Actually I have some hyperspectral data on map projected coordinates say mercator projection. And at different wavelengths. I will try the solution suggested by you. And one more query, by which software we can test. Netcdf -java or anything else?
On Tue, Sep 17, 2019, 3:00 AM Andrew Barna <[email protected]> wrote: > Hi Ghansham, > > > > I think what it wants you to do is the following (but am not 100% sure): > > Say you have a profile variable like this: > > float abs(x,y,z) ; > > abs:standard_name = “atmosphere_absorption_optical_thickness_due_to” > > > > If you want to add multiple wavelengths to it, add a new dim: > > float abs(x,y,z,wl); > > abs:standard_name = “atmosphere_absorption_optical_thickness_due_to" > > Then add another variable: > > float wl(wl): > > wl:standard_name = “radiation_wavelength” > > > > If you really want to be explicit, add a coordiantes attribute to the abs: > > float abs(x,y,z,wl); > > abs:standard_name = “atmosphere_absorption_optical_thickness_due_to" > > abs:coordinates = “x y z wl” > > > > Since I also have some ocean profile data with CDOM at the seawifs > wavelengths, the above was how I was planning on storing them in netcdf > files (with some differences due to non standard z levels). > > > > I hope this helps (and I hope I’m not wrong). > > -Barna > > > > *From:* Discussion around CF Conventions <[email protected]> > *On Behalf Of *Ghansham Sangar > *Sent:* Sunday, September 15, 2019 10:34 PM > *To:* [email protected] > *Subject:* [CF-CONVENTIONS] coordinate variable standard_name > "radiation_wavelength" > > > > Hi. > > > > > > Following url mentions for some of the variable names starting with > "atmosphere_absorption_optical_thickness_due_to" > string to have > > a coordinate variable with standard name "radiation_wavelength" or > "radiation_frequency" > > > > > http://cfconventions.org/Data/cf-standard-names/66/src/cf-standard-name-table.xml > > > > > > Can you just elaborate how to do this. I could not find anywhere in the > CF1-.7 conventions document or even 1.8, that there can be coordinate > variables with such a standard name. > > I have hyperspectral data that I want to write with wavelength as one > dimensions. > > > > regards > > Ghansham >
