Re: Time axis

```Hi Bernadette,

o If your time intervals are hundreds of years, my off-the-cuff
suggestion is that you can avoid the problem altogether by issuing the
command
yes? CANCEL MODE CALENDAR
This will bypass the need to translate time step values into formatted
dates, and may give you just the plots you want.

---
o If you intend to do averages or derivatives  in time, then I suggest
you work-around the problem by defining a purely numerical time axis
based on the year number as in

yes? DEFINE AXIS/T=100:1000:100 my_taxis    ! no units
yes? LET new_var = var[gt=my_taxis@ASN]

Now use new_var instead of var.

---
o Digging into the details, I assume that the problem you refer to is
illustrated by these lines

yes? define axis/T/from/t0="1-jan-0000"/name=tt/units=year360
t[t=100:500:100]
yes? show axis/l=1 tt

name       axis              # pts   start                end
TT        TIME                 5 r   25-JUL-0098 00:00    27-OCT-0492
00:00
T0 = 1-JAN-0000

L     T                 BOX_SIZE       TIME_STEP (YEAR360)
1>  25-JUL-0098 00:00:00  100            100

The explanation is that Ferret V5.0 does not explicitly understand what
is meant by units of "year360". It thinks "year360" is a constant 365
day year, but formats the dates as pure Gregorian. (The V5.0
documentation recommends against "years" as a unit due to its
ambiguities.) We are making V5.1 more sophisticated in this regard -- it

understands "year360" and "year366" -- though more work remains to be
done on Ferret in this area.

- steve

==============

> Hallo,
>
> I have some netcdf files with axis x_t,y_t and Time.
> Time is in units: year360 since 0000:01:01
>
> ncdump shows me, that the times in the file are greater than 100.
>
> But plotting some data with ferret it writes a date earlier than
> 01.01.100 on the plot.
>
> Any suggestions what I am doing wrong?
>

--
Steve Hankin
NOAA/PMEL, 7600 Sand Point Way NE, Seattle, WA 98115-0070
ph. (206) 526-6080 -- FAX (206) 526-6744

```

• References: