Re: time units in ferret & netcdf

[Joerg <joerg@xxxxxxxxxxxxxxxxxxx>]

Thanks to Steve and Mark for their suggestsions concerning the 
time/netcdf stuff!

In another mail Steve clarified for me that the L index can be used
for averaging and so on:

<excerpt>
>> The only thing I'm (currently) afraid of when using real days is the
>> 
>> question what I have to do if I need to average data use calculate
>> means.
>> 
>> 
>> In Steve's suggestion:
>> 
>> 
>> <excerpt>   yes? show grid/l=1:5 t_
>> 
>> 	...
>> 
>>            L     T                 BOX_SIZE       TIME_STEP (DAYS)
>> 
>>            1>  16-JAN-1946 12:00:00  31             89864.5
>> 
>>            2>  15-FEB-1946 00:00:00  28             89894
>> 
>>            3>  16-MAR-1946 12:00:00  31             89923.5
>> 
>>            4>  16-APR-1946 00:00:00  30             89954
>> 
>> 
>> </excerpt>
>> 
>> does the L mean that I can average all Decembers by just referring to
>> the
>> 
>> L index, i.e. do something like mean(xxx where mod(l,12)=0) 
>> 
>> I'd hate to type in all the nice dates ;-)
>
>
>If you ask for, say
>
>	SHADE V[L=1:5@ave]
>
>you will get exactly that -- the average of the first 5 months. The
>documentation (labels) on the plot will tell you the exact date ranges
>used. NOTE:, however, that the average will be weighted -- each month
>weighted by the number of days in it.
>To get an unweighted arithmetic average use
>
>	SHADE V[L=1:5@SUM]/5
>
>The pseudo-variable "TBOX" will contain the box size (days in the month
>-- possibly in units of seconds -- check the Users Guide on that), if you
>want ever to use it directly. 
>
>	LIST/L=1:5 TBOX[G=my_variable]
>

<\excerpt>

Hope this might be useful.

Joerg

--
Jörg Kaduk                           phone:  +1 (650) 325 1521 ext 416
Carnegie Institution of Washington   fax:    +1 (650) 325 6857
Department of Plant Biology
260 Panama Street                    e-mail: joerg@jasper.stanford.edu   
USA - Stanford, CA 94305