Re: [ferret_users] Edges and @AVG

["Ansley C. Manke" <ansley.b.manke@xxxxxxxx>]

Hi Alexendar,

You're right. If the monthly data is meant to represent the first through end of the month, then the best way to recenter would be by using the @ASN transformation as you suggested.  That doesn't do any operation on the data itself.

Alternatively you can redefine the time axis within the session, which replaces the definition that Ferret/PyFerret is using within the session (but doesn't change the file).  It really comes down to the same thing.  Give it the right time-axis definition, probably using /EDGES to put the cell edges at the start/end of months and therefore the grid points at the middle,

yes? use my_data.nc
yes? define axis/edges/t[...]  `var,return=taxis`

You'd get a warning message like this,

Replacing definition of axis TIME
           *** NOTE: grid GDN1 used by data set my_data
           *** NOTE: Redefinition may alter apparent contents of data set

On 2/25/2020 2:15 PM, Alexander Audet wrote:

Afternoon Billy,

As far as I can tell (link) although the monthly means are timestamped at the first of the month and seem to be centered there, they actually represent the average value of the month instead 1/2 the last and 1/2 the named month. I am still confused about this, so I could be wrong. 

If so, by my understanding, it wouldn't be smoothing to recenter the data into the middle of the month?

Thank you for your comments! These are definitely details I need to think about!

Alexander Audet

On Tue, Feb 25, 2020 at 4:28 PM William S. Kessler <william.s.kessler@xxxxxxxx> wrote:

> On Feb 25, 2020, at 12:26 PM, Alexander Audet <alexander.c.audet@xxxxxxxxx> wrote:
>
> The full fix would probably be to first regrid it into proper 1-month time slots using the @ASN transformation as described in the above link, then regrid it a second time into the seasonal 3 months using the @AVG.

There's a danger here: The first step of averaging to put the data onto explicit months when the time points are centered on the first of the month is a smoothing operation before you start to do anything else. Won't that have a larger effect than the length-of-month problem?

It seems to me that "1/2 December, January, February, and 1/2 of March" is the right way (least smoothing) to define DJF. You could do this manually, or use @ave specifying dates; in that case if Ferret understands the calendar it will do this correctly.

Billy K