[Go-essp-tech] ok_min_mean_abs and ok_max_mean_abs

[Charles Doutriaux]

  Martin,

I will definitely do 2.

Also I will edit the tables to set these min/max for monthly tables only.

I will see if I can run the script for other frequencies as well.

C.

On 11/9/10 8:15 AM, martin.juckes at stfc.ac.uk wrote:
>
> Hello Charles,
>
> For some variables, e.g. meridional wind, the monthly mean values will 
> generally be much smaller than the daily values, and so won’t give a 
> good indication of what values to expect.
>
> I can think of two options – (1) do a scan of the AR4 archive at all 
> frequencies, (2) issue a guidance note saying this particular warning 
> should be ignored if there is reason to believe that the absolute 
> values of the monthly data are not a good guide.
>
> Cheers,
>
> Martin
>
> *From:* go-essp-tech-bounces at ucar.edu 
> [mailto:go-essp-tech-bounces at ucar.edu] *On Behalf Of *Charles Doutriaux
> *Sent:* 09 November 2010 16:07
> *To:* Karl Taylor
> *Cc:* Denis.Nadeau at nasa.gov; go-essp-tech at ucar.edu
> *Subject:* Re: [Go-essp-tech] ok_min_mean_abs and ok_max_mean_abs
>
> As Karl mentioned, I got the max/min from looping over 20c3m models 
> "monthly" in AR4.
>
> If anybody as better values please forward them to us we'll gladly put 
> them into the tables.
>
> If anybody's interested I can search the code I used to search through 
> the whole AR4 archive.
>
> C.
>
> On 11/8/10 2:07 PM, Karl Taylor wrote:
>
> Hi Denis,
>
> I'm copying Charles D. who I hope will check the following for 
> correctness.  As I recall:
>
> The mean_abs value is the temporal and spatial mean of the absolute 
> values of a field (based on latxlon grid cell values, but not area 
> weighting these).  The mean_abs value specified in the CMOR tables is 
> based on monthly means produced by CMIP3 models.  It basically gives 
> you the typical magnitude of the values of the field.  If a model has 
> a number much smaller or much larger than the CMOR table value, then 
> it may be unrealistic (or if off by an order of magnitude or more, it 
> probably means the units are wrong.)
>
> Charles can tell you how he got the "max" and the "min" that appear in 
> the table.
>
> regards,
> Karl
>
> On 11/8/10 11:43 AM, Denis Nadeau wrote:
>
> Hi,
>   
> Do you know how the mean_abs found in CMOR tables were calculated for
> some variables?  Is there some documentation about this?
> I get warnings from cmor2 API and would like to investigate further what
> they mean.
>   
> Thanks,
> Denis
>   
>
>
> -- 
> Scanned by iCritical.
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/go-essp-tech/attachments/20101109/2c98d910/attachment-0001.html