Hello David,
I absolutely agree --- for model data the distinction is only a formality without any practical relevance. Instead the relevance of having both alternatives lies precisely in the formal difference: some indices are defined to use a strict inequality and others are defined to use a non-strict one. For observational data with limited precision the distinction does matter, and historically the focus of both the ETCCDI and ET-SCI were on observational data. I imagine that is the reason for that the index definitions are as they are.
And yes, it would certainly be useful if the providers observational data could do as you suggest, and include the precision of the data as you suggest. One complication would be that the precision might (will likely) change over time as new instruments and data processing equipment is introduced, but that is probably another discussion.
Kind regards,
Lars
From: David Hassell [mailto:david.hassell at ncas.ac.uk]
Sent: den 29 mars 2017 15:51
To: B?rring Lars
Cc: cf-metadata at cgd.ucar.edu
Subject: Re: [CF-metadata] Request for new standard names for climatological statistics based on thresholds
Hello,
For model output in particular, which has a precision of much less than 0.1K, I would expect the "at" and "at_or_above" indices to be essentially identical - so which one would you choose, given that you might have both indices from observations?
To make the distinction more meaningful, you could quantise your model data and record that in the cell methods, perhaps:
time: maximum within days (rounded to 0.1 K) time: sum over days
?All the best,?
?David?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <
http://mailman.cgd.ucar.edu/pipermail/cf-metadata/attachments/20170329/33556467/attachment.html>
Received on Wed Mar 29 2017 - 08:18:33 BST
