Re: Fwd: month_to_season calendar error?

Hello,

Are the units associated with 'time' (ie, time@units) for either the
model or, more likely, the reanalysis dataset "months since .....".
If so, that is likely the problem. The deprecated function 'ut_calendar'
has a brief example of the issues associated with use of
"months since ..." units.

===
Cheers

On 2/17/14, 6:17 AM, Marston Johnston wrote:
> Hi again,
>
>
> I'm resending this post as I fear it may have fallen between the cracks and
> that is why I've not gotten any response so far. The NCL group is quite
> fast in responding so I suspect this email may have gotten lost.
>
> I'm trying to match two datasets: model and reanalysis in time and
> geographically. Each dataset has the dimensions (ntime,nlat,nlon). The time
> increments are monthly for a number of years. In matching the time, I also
> took into consideration that all years should have 12 months of data. My
> problem occurs when I have sliced the
> model data and apply the function: month_to_season; I get a shift in the
> year for a model that uses the 360_day calendar but not for models that use
> another type of calendar. Is this an error in the month_to_season function
> or am I not understanding the function correctly?
>
> There is a test code attached: test.ncl and I can provide the datasets upon
> request.
>
> The output is found in the log file (attached).
>
> Your help is greatly appreciated!
>
> Sincerely,
> Marston
>
>
>
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Mon Feb 17 08:28:13 2014