NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.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.

Re: [thredds] WMS time dimension different from OPeNDAP

Hi John, Patrick,
Quick sleuthing. Looking at the OPeNDAP HTML page I see that the units of the time axis are "months since 1901-01-01 00:00:00"
It's miraculous that the two interpretations differ by a mere 21 minutes out of a 50 year offset from T0, considering that "months" is not a standardized time unit. CF advises that "/...the unit month ... should be used with caution/". An understatement that might better be stated as (say) "/units of months should not be used in any case where the desired precision in the specification of the date is finer than 1 day/" (to be on the safe side). 

    - Steve

------------------------------------------------------------------------


On 4/11/2013 11:30 AM, John Caron wrote:

Hi Patrick:
So WMS datetimes are probably being done through ncWms, using Joda Time library, whereas the TDS opendap service will not interpret the dates, so it must be the ferret client, presumably using udunits C library.
Its hard to imagine 2 date libraries getting different answers, but since one is done on the server, and the other on the client, it must be the difference in the gravitational attraction of the moon between the two locations, causing a slight difference in the length of the month. 

Do you by any chance know what is the correct answer?

John


On 4/9/2013 7:08 PM, Patrick Brockmann wrote:

Hi all,
I have a strange behaviour with time dimension values exposed by the GetCapabilities 
and values get from an OPeNDAP access.
Could someone explain me what is the source of the problem ?
http://webportals.ipsl.jussieu.fr/thredds/wms/CARBON/CARBONES_PRIOR_orchidee_hadgem2-es_historical_nosoc_co2_npp_monthly_1951_2004_yearmonmean_3.nc?service=WMS&version=1.3.0&request=GetCapabilities 

---> 1951-07-16T12:12:04.339Z, ....

And from the OPeNDAP access with ferret client, I get

ferret
yes? use "http://webportals.ipsl.jussieu.fr/thredds/dodsC/CARBON/CARBONES_PRIOR_orchidee_hadgem2-es_historical_nosoc_co2_npp_monthly_1951_2004_yearmonmean_3.nc"; 
yes? sg/t/l=1:10 npp2
    GRID GTL2
 name       axis              # pts   start                end
 MYLON     LONGITUDE          720mr   0E                   0.5W
 LAT       LATITUDE           360 r   89.75S 89.75N
 normal    Z
TIME TIME 54 r 16-JUL-1951 12:33 15-JUL-2004 09:01 

       L     T                   TBOX      TBOXLO TSTEP (months)
1> 16-JUL-1951 12:33:59 12 14-JAN-1951 21:39:23 606.4526 2> 15-JUL-1952 18:23:11 12 15-JAN-1952 03:28:35 618.4526 3> 16-JUL-1953 00:12:23 12 14-JAN-1953 09:17:47 630.4526
--> Note the difference between 1951-07-16T12:12:04.339Z and 16-JUL-1951 12:33:59 

My TDS version is Version 4.3.15 - 20121218.1126

Regards
Patrick



_______________________________________________
thredds mailing list
thredds@xxxxxxxxxxxxxxxx
For list information or to unsubscribe, visit: http://www.unidata.ucar.edu/mailing_lists/
  • 2013 messages navigation, sorted by:
    1. Thread
    2. Subject
    3. Author
    4. Date
    5. ↑ Table Of Contents
  • Search the thredds archives: