[Thread Prev][Thread Next][Index]

Re: [ferret_users] append variables on time when time reference (since date) are different



Hi -
Ferret definitely handles regridding to a common axis correctly. The underlying units of the axes are translated into seconds since year 0, and the regridding is done in those units.

The issue here appending the variables with the same name amd with the two axes having different time-axis units and/or origins. Ferret should issue a warning at least. We've entered this on our to-do list.

Ansley

On 11/15/2011 1:14 AM, Hein Zelle wrote:
Dear Ansley, Patrick,

Ferret's graphics lets you view plots of variables on different time
axes together using /HLIM or /VLIM. I think this is causing some
confusion. To save them to a single dataset, appending them, you
would need to put them onto a common time axis, and it is up to the
user to make sure that the time origin and axis units match.
I've run into this issue as well.  Does Ferret deal correctly with
regridding to a common time axis?  i.e.

dataset 1: time axis starts at 01-01-1970, hourly data
dataset 2: time axis starts at 01-01-1990, hourly data

define new axis my_t starting at 01-01-1900, hourly

let var3 = var1[d=1,gt=my_t] + var2[d=2, gt=my_t]
save (/append) /file=test.nc   var3

Will that do what I expect/want ?  If so, it's a perfectly acceptable
workaround, but it would be very nice indeed if ferret issues a
warning or an error when you're incorrectly combining time axes with
append.


Kind regards,
      Hein



[Thread Prev][Thread Next][Index]
Contact Us
Dept of Commerce / NOAA / OAR / PMEL / Ferret

Privacy Policy | Disclaimer | Accessibility Statement