[ferret_users] holes in data from time axis regridding

[Martin Schmidt <martin.schmidt@xxxxxxxxxxxxxxxxx>]

Hi ferreters,

I need some help with strange ferret results. I would like to process 
some data. My output should be hourly.

I define the new time axis and grid

yes? define 
axis/t=1-dec-1999:31-jan-2000:1/unit=hours/t0=1-jan-1955/cal=julian tax
yes? define grid/t=tax tgrid

! Open the file with cloud data (a reduced version is attached) for 
interpolation onto the new time axis
yes? use clouds.nc

! Define a new hourly variable

yes? let c=clour[gt=tgrid@lin]

! list example points

yes? list/x=2/y=57/prec=5 c[l=9:14]

I get

01-DEC-1999 08:00 / 9: 0.7163
01-DEC-1999 09:00 / 10: 0.9058
01-DEC-1999 10:00 / 11: 0.9979
01-DEC-1999 11:00 / 12: 0.9987
01-DEC-1999 12:00 / 13: 0.9996
01-DEC-1999 13:00 / 14: 1.0000

looks nice.
But now:

yes? shade c[l=13]

no valid data

yes? list/x=2/y=57 c[l=13]
VARIABLE : CLOUR[GT=TGRID@LIN]
FILENAME : clouds.nc
LONGITUDE: 2E
LATITUDE : 57N
TIME : 01-DEC-1999 12:00 JULIAN
....
Nanu? For l=13 I hit exactly a time of the input data.

yes? list/x=2/y=57/prec=5 c[l=12]
VARIABLE : CLOUR[GT=TGRID@LIN]
FILENAME : clouds.nc
LONGITUDE: 2E
LATITUDE : 57N
TIME : 01-DEC-1999 11:00 JULIAN
0.99873
OK But for l=13 the result is still missing.

And now:
yes? shade c[l=12:13@ave]

gives a figure.

yes? list/x=2/y=57/prec=5 c[l=13]
VARIABLE : CLOUR[GT=TGRID@LIN]
FILENAME : clouds.nc
LONGITUDE: 2E
LATITUDE : 57N
TIME : 01-DEC-1999 12:00 JULIAN
0.99958

(From the cash). This is a strange finding. The final data are too large 
to be written in one step. I have to do this with a loop over l. Doing 
so I get
gaps in the data set, in the way as shown above.

Are there any errors in my commands or is this a bug? I am using 6.1, 
but an old 5.41 for AIX shows the same behaviour.

Regards,
Martin Schmidt

Attachment: clouds.nc
Description: Cdf file