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

[netCDF #OZU-387092]: [Question] Data.

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: [netCDF #OZU-387092]: [Question] Data.
  • Date: Wed, 05 Dec 2012 13:52:53 -0700
Hi,

> This is Seungwoo Jason Chang who is a pH.D student of University of
> Florida. I downloaded the NetCDF data in this website
> (http://geoport-dev.whoi.edu/thredds/ncss/grid/prism4/monthly/dataset.html),
> but I can't find the any directions about the data. For example, in my
> downloaded file, time variable varies like 13194
> 13225 13253 13284 13314 13345 13375 13406 13437 13467 ... . Is there
> any direction to figure out the meaning of these values or standard?
> 
> I asked Gibson who is the creator of these data
> (http://geoport-dev.whoi.edu/thredds/prism4.html?dataset=prism4/monthly),
> but he said the data was converted in your website, so he has no idea
> about it.
> 
> Appreciate if you could explain to me this or give some contacts I can
> ask.

According to the "Gridded Dataset Description", which is a link you can click 
on, 
the time coordinate is defined as follows:

  <axis name="time" shape="1403" type="float" axisType="Time">
  <attribute name="units" value="days since 1858-11-17 00:00:00"/>
  <attribute name="_ChunkSize" type="int" value="1"/>
  <attribute name="_CoordinateAxisType" value="Time"/>
  <values>
  13194.0 13225.0 13253.0 13284.0 13314.0 13345.0 13375.0 13406.0 13437.0 
13467.0 13498.0 13528.0 13559.0     
  13590.0 13619.0 13650.0 13680.0 13711.0 13741.0 13772.0 ...

so the units are days since 1858-11-17 00:00:00".  You can convert this to a 
date/time using various libraries, but here is a way using the netCDF utilities 
ncgen and ncdump.

First create a netCDF CDL file, "days.cdl" for example, which is just an ASCII 
file that could look like this:

netcdf days {
dimensions:
        time = UNLIMITED ; // (20 currently)
variables:
        float time(time) ;
                time:units = "days since 1858-11-17 00:00:00" ;
data:

 time = 13194, 13225, 13253, 13284, 13314, 13345, 13375, 13406, 13437, 13467, 
    13498, 13528, 13559, 13590, 13619, 13650, 13680, 13711, 13741, 13772 ;
}

Then use ncgen to make a binary netCDF file, "days.nc" out of days.cdl, like 
this:

  $ ncgen -b days.cdl

Finally, dump that out with ncdump using the "-t" option to translate times
to a human-readable form, like this:

  $ ncdump -t days.nc

for which the output is:

netcdf days {
dimensions:
        time = UNLIMITED ; // (20 currently)
variables:
        float time(time) ;
                time:units = "days since 1858-11-17 00:00:00" ;
data:

 time = "1895-01-01", "1895-02-01", "1895-03-01", "1895-04-01", "1895-05-01", 
    "1895-06-01", "1895-07-01", "1895-08-01", "1895-09-01", "1895-10-01", 
    "1895-11-01", "1895-12-01", "1896-01-01", "1896-02-01", "1896-03-01", 
    "1896-04-01", "1896-05-01", "1896-06-01", "1896-07-01", "1896-08-01" ;
}

So it looks like the times are just the first day of consecutive months ...

--Russ

Russ Rew                                         UCAR Unidata Program
address@hidden                      http://www.unidata.ucar.edu



Ticket Details
===================
Ticket ID: OZU-387092
Department: Support netCDF
Priority: Low
Status: Closed