[CF-metadata] How to describe nr. of samples & absolute/relative uncertainty

From: Antonio S. Cofiño <antonio.cofino>
Date: Thu, 22 Mar 2018 12:13:05 +0100

Dear Erik,

Please take a look at this thread in the mail list:

http://mailman.cgd.ucar.edu/pipermail/cf-metadata/2018/059879.html

The netcdf-U and uncertML could help you to encode the metadata you are
interested in.

Please come back with any progress you made.

Regards

Antonio

--
Antonio S. Cofi?o
Dep. de Matem?tica Aplicada y
         Ciencias de la Computaci?n
Universidad de Cantabria
http://www.meteo.unican.es
On 22/03/18 11:49, Erik Quaeghebeur wrote:
> Dear Eduardo,
> Dear list,
>
>
> I asked (among other things):
>>>
>>> I have 10-minute statistics [?] data for which I know the sampling 
>>> frequency and therefore the number of samples. [?]
>>>
>>> How do I encode the sampling frequency and number of samples in a 
>>> way that these are accessible in a standard way?
>>> I now use, e.g.,
>>> variable.cell_methods = "time: mean (interval: 10 minutes comment: 
>>> sampled at 1 Hz, so 600 samples)"
>>> but this puts the frequency information in an ?opaque? comment. [?]
>
> Eduardo replied (my thanks go out to him):
>>
>> I am not sure if I understood properly your issue, so I preferred to 
>> write directly to you instead of posting my answer to the mailing list.
>> I have the feeling that what you are looking for would be:
>> variable:cell_methods = "time: mean (interval: 1 second)"
>> then the length of the interval would be perfectly defined in the 
>> values of the time coordinate boundaries
>> time:bounds = "time_bnds" ;
>> so the number of observations could be obtained combining those two 
>> elements: the sampling frequency information within the "interval" 
>> bit of the cell_methods and the length of the sampling interval 
>> defined by the "time_bnds" values.
>
> And Eduardo is absolutely right about this. I had misunderstood the 
> definition of ?interval?. It would nevertheless still be of interest 
> to be able to put this information?sampling frequency and number of 
> samples?in separate /number-valued/ attributes. Now the cell_methods 
> field has to be parsed and the value obtained combined with an 
> interval length implicit in the time_bounds variable.
>
> Would you consider adding such attributes to the conventions?
>
> (I would be grateful for any replies to my questions about encoding 
> uncertainties.)
>
>
> Best,
>
> Erik
>

Received on Thu Mar 22 2018 - 05:13:05 GMT