Happy Valentine's Day! Love, the netCDF Operators NCO version 4.9.2.
http://nco.sf.net (Homepage, Mailing lists, Help)
http://github.com/nco (Source Code, Issues, Releases, Developers)
What's new?
We have re-licensed NCO source code from GPL3 to 3-Clause BSD
starting with Version 4.9.2. Why? To encourage wider adoption
and support. Without institutional (i.e., federal agency) support,
NCO development would crawl to a standstill. We are acceding
to the preferences of our occasional sponsors (DOE, NASA, NSF)
for permissive OpenSource licenses.
Version 4.9.2 contains many regridding fixes and features:
Improved adoption has made the NCO weight-generator more robust.
Its signature strengths are speed and ultra-high resolution support.
It can create ~3 km global maps from in under an hour to and from
some (not all) grids including FV physics grids (ne1024pg2), MPAS
grids, and RLL grids. Problems remain, though ncremap also supports
ESMF and TempestRemap so take your pick. Moreover the map checker,
ncks --chk_map, produces improved metrics with fewer false alarms.
Work on NCO 4.9.3 has commenced and will improve NCO weight-generator
accuracy, reduce vertical interpolation memory use, and supply more
accurate weight generation options for rectangular lat-lon grids.
Work on NCO 5.0.0 has commenced "under the hood". A key leap in that
release will be support for netCDF4 user-defined types. Printing of
netCDF4 user-defined types ENUM and VLEN is ready now (though
unsupported) with the --udt flag. 5.0.0 will contain the finished
version of that, and include options for invoking mbtempest in place
of tempest.
Enjoy,
Charlie
NEW FEATURES (full details always in ChangeLog):
A. ncremap now supports vertical interpolation of pure-pressure
files to hybrid grids when the surface PS is included in the
pure-pressure data-file rather than the vertical grid-file.
Moreover, the PS may be time-varying. This feature permits
full decoupling of PS from the vertical grid-file which allows
for more elegant manipulation of vertical grid-files.
ncremap --vrt=vrt.nc ncep.nc hybrid.nc
http://nco.sf.net/nco.html#vrt
B. ncks --chk_map checks for and reports NaN weights.
Previously chk_map generally ignored NaNs in the weight array.
The format of the analysis report has also been improved, and gives
the locations of extrema, i.e., max/min for weights and errors.
The checker takes better account of masks and regional grids and
so now reports fewer false-positive map errors.
ncks --chk_map map.nc
ncks -D 2 --chk_map map.nc # Increased verbosity
http://nco.sf.net/nco.html#chk_map
C. ncremap and ncclimo both now export HDF5_USE_FILE_LOCKING='FALSE'
unless that environment variable already exists. This proves
useful at preventing mysterious crashes, especially in batch mode.
Let us know if this setting causes problems for your workflow.
http://nco.sf.net/nco.html#HDF5_USE_FILE_LOCKING
D. NCO's weight generator can now read unstructured SCRIP gridfiles
created by TempestRemap utilities that do not include the
(supposedly standard) "grid_dims" dimension when converting Exodus
files to SCRIP. The unstructured dimension size is used instead of
the contents of the "grid_dims" variable.
E. ncremap --devnull option controls whether internal (and sometimes
noisy) output from NCO weight-generator is sent to terminal or to
/dev/null. The default is to send internal messages to /dev/null.
This only works with the NCO weight-generator (--alg_typ=nco).
ncremap --devnull=Yes ... # Send internal messages to /dev/null
ncremap --devnull=No ... # Send internal messages to terminal
http://nco.sf.net/nco.html#devnull
http://nco.sf.net/nco.html#ncremap
F. ncpdq -U now always unpacks all variables, including packed
grid variables (like "lat"). Previously it would not unpack
variables with specially reserved grid names ("lat", "lon", etc.).
Thanks to Susan Rennie for reporting this issue.
This fixes a bug introduced in NCO 4.7.7. Solution is to use
earlier versions of NCO or to upgrade.
http://nco.sf.net/nco.html#ncpdq
BUG FIXES:
A. ncks --chk_map statistics no longer scrambled.
Statistics of "normal" global->global conservative maps are
unaffected. Some other types of map-files are.
There is no workaround. Solution is to upgrade.
http://nco.sf.net/nco.html#chk_map
B. ncremap --alg_typ=fv2fv now implements the same TempestRemap
options as --alg_typ=fv2fv_stt. The previous fv2fv options
were not actually as described. Solution is to upgrade.
http://nco.sf.net/nco.html#fv2fv
http://nco.sf.net/nco.html#fv2fv_stt
C. ncremap now passes all NCO options through to ncks.
Previously these options, such as file formats were not passed
and ncks used defaults instead.
There is no workaround. Solution is to upgrade.
Full release statement at http://nco.sf.net/ANNOUNCE
--
Charlie Zender, Earth System Sci. & Computer Sci.
University of California, Irvine 949-891-2429 )'(
