Re: [ferret_users] samplet_date

[Russ Fiedler <russell.fiedler@xxxxxxxx>]

Looks like a bug in samplet_date.

Something odd is going on with the time construction.

At one point in the code it looks like the mean year is 365.24 days long

date_axis(l1,1,1,1) = byr + bda/ 365.24  (line 362 in samplet_date.F)

but later it is 365.25 days long

timepoint = ayr + ada/ 365.25              (line 405)




Russ

On Fri, 2012-03-30 at 10:26 +1100, Carol Ladd wrote:
> I am trying to extract certain dates from a time series using the
> samplet_date function but it is not returning what I would expect.  It
> looks like the function is interpolating between the dates and I don’t
> know if it is a problem with the samplet_date function or a problem
> with how I have my time axis defined.  I have attached my time series
> to use as an example:
> 
>  
> 
> yes? use urot2.nc
> 
> yes? sh d
> 
> yes? list/t=9-sep-1999 urot2
> 
>              VARIABLE : U_ROT[GT=TQSCT@ASN]
> 
>              FILENAME : urot2.nc
> 
>              Z (m)    : 0
> 
>              TIME     : 09-SEP-1999 00:00
> 
>          -13.24
> 
>  
> 
> yes? list SAMPLET_DATE(UROT2,1999,9,9,0,0,0)
> 
>              VARIABLE : SAMPLET_DATE(UROT2[D=urot2],1999,9,9,0,0,0)
> 
>              Z (m)    : 0
> 
>              T        : 1
> 
>          -12.92
> 
>  
> 
> Why are they not the same?
> 
>  
> 
> Thanks.
> 
> Carol
> 
>  
> 
> ********************************************************
> 
> Carol Ladd
> 
> Oceanographer
> 
>  
> 
> NOAA/PMEL
> 
> 7600 Sand Point Way
> 
> Building 3      
> 
> Seattle, WA                    
> 
> 98115-6349                   
> 
>  
> 
> Phone: (206) 526-6024
> 
> Fax: (206) 526-6485
> 
>  
> 
> Email: carol.ladd@xxxxxxxx
> 
>  
> 
> ********************************************************
> 
>  
> 
>