Re: [ferret_users] inconsistent presentation of _FillValue

[Marco van Hulten <Marco.Hulten@xxxxxx>]

Ansley—
 
Sorry, I cannot reproduce my issue any more.  I don't know what is
different now.  If it happens again, I'll do some more testing (and
report again if it persists).
 
On 20 Mar 09:19 Ansley C. Manke wrote:
> Could you show the ncdump header information for these variables?
 
$ ncdump -h NOINYOC_tn21_10_ciso.micom.hbgcy.0596.nc | grep po4
	short po4(time, sigma, y, x) ;
		po4:_FillValue = -32768s ;
		po4:valid_range = -32767s, 32767s ;
		po4:actual_range = 4.4436999178049e-06, 0.00532026284518526 ;
		po4:scale_factor = 8.11154384787661e-08 ;
		po4:add_offset = 0.00266235327255153 ;
		po4:units = "mol P m-3" ;
		po4:long_name = "Phosphorus" ;
		po4:coordinates = "plon plat" ;
		po4:cell_measures = "area: parea" ;
 
Then, looking in the full ncdump:
 
po4 =
_, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, 
...
_, _, _, _, _, _, _, _, -10232, -10335, -10536, -10703, _, _, _, _, _, _, 
...
 
When I plot `po4' now in Ferret, it looks fine, but clearly these very
negative values (though not equal to _FillValue) here suggest that I
must investigate further!
 
—Marco