Re: [ferret_users] white space in latitude-time cross-section plot (Hovemuller plot)

[William Kessler <william.s.kessler@xxxxxxxx>]

Assuming that the data is just 12 months, then Ferret will not recognize the axis as modulo, and won't know how to extend it earlier than mid-January or later than mid-December.

You can force it to be modulo, then do as Ryo suggests:

let modvar = var_name[gt=month_irreg@mod]    ! now has a repeating monthly time axis

! first check that the regridding did what you expect:
plot/line/sym/y="pickanylatitude" var_name,modvar    ! are they the same? Then proceed

fill/set/l=0:13 modvar          ! extend the plot from the preceding Dec to the following Jan
ppl time,w000001010000,w000012312359    ! force the time axis to be from 1 Jan to 31 Dec
ppl txlint,1,0               ! omit the year axis label (would be year "0000")
ppl fill                  ! finally, draw the plot

It can also be visually useful to plot two successive years:
fill/set/l=0:25 modvar          ! extend the plot from the preceding Dec to the second Jan
ppl time,w000001010000,w000112312359    ! force the time axis to be from 1 Jan to 31 Dec
ppl txlint,1,0               ! omit the year axis label (would be year "0000")
ppl fill                  ! finally, draw the plot


> On May 8, 2020, at 7:01 AM, Oluwaseun Ilori <wilseun@xxxxxxxxx> wrote:
> 
> Thanks Ryo for the clarification and I will try it out, before then I think I should mention this here. The data is climatological monthly mean data gotten from cdo (ymonmean) operator on 55 years monthly data. The output (ymonmean_REMO.nc) now contains 12 months of data (Jan-Dec) that I needed to examine latitudinal variation from January to December.
> 
> Thanks once again for your help and I will try your suggestion and get back to you. 
> 
> 
> On Fri, May 8, 2020, 2:15 PM Ryo Furue <furue@xxxxxxxxxx> wrote:
> Dear Oluwaseun,
> 
> If your time axis is annual climatology (that is, if it repeats itself every year), it should be defined as "modulo".  Then, you can avoid these white bands by extending the data beyond the beginning and end of the year, as in
> 
> fill/hlimits=0:365/ . . . pr[t=-40:400, . . .]
> 
> if the underlying units of time is "days" and if your climatological year consists of 365 days.  Basically, you limit the region of plotting to one year, but ask Ferret to take values before and after the year into account when generating contour lines (shadings).  Having said that, I'm not too familiar with this issue and I would always need some trial and error before I get the right results.
> 
> To see what time axis your data has, you want to see the outputs of these commands:
> 
> yes? show grid pr
> yes? list t[gt=pr]
> 
> Hope this helps.
> 
> Ryo