Hi Glenn,
I'm filling in for Ansley manke, who is off on vacation.
When Ferret initializes a data set and sees a longitude axis it tries to
decide if the axis is full-globe and therefore "modulo". It sounds like
the rounding error is causing it to answer the question with a "no". If
that is the correct diagnosis, then the resolution is simple
yes? SET AXIS/MODULO axis_name
You can determine axis_name either with SHOW GRID var_name or insert
this in place of axis_name:
`var_name,RETURN=xaxis`
Note that as of Ferret V5.5 longitude axes are recognized as "sub-span
modulo" if they are less than 360 degrees. I suspect that you are using
an older version of Ferret. In V5.5+ your regridding request should
result in a thin band of missing values down the seam where the
precision error occurs at x=359.999
- steve
===============================
Glenn Carver wrote:
I've having trouble producing plots to span the 0E line.
My netcdf file consists of variables defined on a regular
grid from 0 -> 350 (say), spaced by 10 degrees.
I want to produce a plot over a limited area so I do this:
def axis/x=-10:10:1/units=degrees xx
def axis/y=30:50:1/units=degrees yy
def grid/x=xx/y=yy gg
fill ozone[g=gg]
This works fine EXCEPT for one set of netcdf files I have
in which the longitude coords in the netcdf file have a
small rounding error. e.g. if you list x[g=ozone] you get
0, 10, 20, ...... 339.999, 349.999
In this case, I only get data plotted from 0 -> 10. From
-10 to 0, I get nothing. If I correct the longitude axis
to have 330.0, 340.0, 350.0, it all works! I don't
understand why ferret is so sensitive to this small
rounding error and doesn't regrid onto my new grid if
this small rounding error is there?
Glenn
--
Steve Hankin, NOAA/PMEL -- Steven.C.Hankin@noaa.gov
7600 Sand Point Way NE, Seattle, WA 98115-0070
ph. (206) 526-6080, FAX (206) 526-6744