int2p returns some missing values when interpolating onto theta surface

I am using int2p to define the height of a theta surface but am getting
some missing values. This is a broader effort to make a routine that
will calculate Ertel's PV on an isentropic surface for stratosphereic
vortex positioning work.

The parameters being fed to int2p are an array of heights
(level,lat,lon) and a corresponding array of theta values. The third
parameter is a single theta value.

The program seems to work but returns random missing patches over the
plot domain. Further, missing patches do not remain the same for
different time periods. It is not a matter of the isentropic surface
intersecting the ground - when set to low thetas that happens and the
expected missing areas appear, however the missing areas still occur no
matter what the theta is set to. I have attached a number of images that
show a progression of isentropic surfaces up into the atmosphere - for
smaller values of theta the surface clearly intersects the ground, but
then missing patches appear that simply don't go away.

I verified that there are no missing values in the input data. I have
been using ind to on the missing resultant values to see if there are
any patterns to the height and theta input data, but I nothing obvious
has emerged. Variations of the linlog argument also do not have any effect.

I don't see that other int2p issues have cropped up on ncl-list so I
can't imagine that this is anything other than I am missing something
obvious.

If anyone wants to have a stab at it I have posted the program here:
<http://people.iarc.uaf.edu/~datkinson/NCL_question/IPV.ncl>

and its supporting data here:
<http://people.iarc.uaf.edu/~datkinson/NCL_question/air.1950.nc>
<http://people.iarc.uaf.edu/~datkinson/NCL_question/hgt.1950.nc>

Unfortunately the two netcdf files are large - 500MB each. They are
4xdaily NN Reanalysis 1 data.

Thanks,
David A.

-- 
David E. Atkinson, PhD
Assistant Professor of Atmospheric Sciences
International Arctic Research Center
930 Koyukuk Drive
P.O. Box 757335
University of Alaska Fairbanks
Fairbanks   AK  99775-7335
tel: +1.907.474.1126
fax: +1.907.474.2643
email: datkinson @ iarc.uaf.edu
_______________________________________________
ncl-talk mailing list
ncl-talk_at_ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk