[Thread Prev][Thread Next][Index]

Re: [ferret_users] out-of-range longitudes



Hi Ryo,

I think that this has to do with the logic of "subspan modulo" longitudes, which lets us work with axes shorter than the modulo length by introducing a fake "void point" to complete the wrapping around the globe, so that such an axis can be used along with a full modulo axis in expressions or refer to points using west longitude and so on.  Here the void point is at 45W. This shows it in a bit more detail,


yes? set data tmp.nc
yes? show axis `sst,return=xaxis`
 !-> sh axis COADSX56_60
 name       axis              # pts   start                end
 COADSX56_60 LONGITUDE          5mr   131E                 139E
   Axis span (to cell edges) = 10 (modulo length = 360)

yes? list sst[x=225w:215w], xbox[gx=sst,x=225w:215w], xboxlo[gx=sst,x=225w:215w], xboxhi[gx=sst,x=225w:215w]
             DATA SET: ./tmp.nc
             LONGITUDE: 135E to 145E
             LATITUDE: 19N
             TIME: 16-JAN 06:00
 Column  1: SST is SEA SURFACE TEMPERATURE (Deg C)
 Column  2: XBOX is XBOX (axis COADSX56_60)
 Column  3: XBOXLO is XBOXLO (axis COADSX56_60)
 Column  4: XBOXHI is XBOXHI (axis COADSX56_60)
              SST   XBOX  XBOXLO XBOXHI
135E   / 3:  26.13    2.0  134.0  136.0
137E   / 4:  26.31    2.0  136.0  138.0
139E   / 5:  26.39    2.0  138.0  140.0
45W    / 6:   ....  350.0  140.0  490.0

It seems the extra points are included when the requested range is outside the range given by the lower limit of the first grid point actually on the axis and the upper limit of the last point. 

As you suggest, these points should not be included in the listings or in the listings or the new file. 


Ansley


On 6/23/2023 1:17 AM, Ryo Furue wrote:
Dear Ferret users,

I was surprised that a construct like temp[x=120:150] refers to strange longitudes when 120 or 150 is out of the actual defined range.  A self-contained example is pasted below.

This is problematic when you save a subset of data because the longitude axis of the new dataset will include these strange longitude points unless you carefully specify the longitude range when saving it.

Is this an expected behavior or is it something that should be fixed in a future release?

Ryo
----
! PyFerret v7.63 (optimized)
! Darwin 19.6.0 - 12/10/21
yes? set data coads_climatology
yes? save/clobber/file=tmp.nc SST[x=130:140,y=20,L=1]
 LISTing to file tmp.nc
yes? set data tmp.nc
yes? list sst[x=120:135]
[ . . . ]
 45W    /  0:   ....
 131E   /  1:  25.68
 133E   /  2:  25.90
 135E   /  3:  26.13
yes? list sst[x=135:140]
[ . . . ]
 135E   / 3:  26.13
 137E   / 4:  26.31
 139E   / 5:  26.39
yes? list sst[x=135:145]
[ . . .]
 135E   / 3:  26.13
 137E   / 4:  26.31
 139E   / 5:  26.39
 45W    / 6:   ....
yes? save/file=tmp-new.nc sst[x=120:145] ! acquires strange longitudes.
yes? set data tmp-new.nc
yes? plot sst


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

Privacy Policy | Disclaimer | Accessibility Statement