Hosted WRF developers meeting on Friday Dec 11, made several commits:
Fixed compilation issue for HDF5
Cleaned up HDF5 and WRFPLUS include statements to avoid warning messages at compile time
Major bug fix for CV7: incorrect calculations were performed for several control variables in the initial implementation
Major cleanup of CV7: skipping unnecessary subroutines, renaming variables and file/directory names to reflect their contents, omitting junk data from be.dat
Regtest improvements:
Hybrid Regtest capability now works, have a hybrid test based on the realtime system.
Adding capabilities for HDF5 and NETCDF4 tests; still issues that need addressing
Running a lot of NETCDF4 tests to try to replicate the slowdown problems Rizvi has seen
Activities for week of December 14, 2015
One day of PTO
Working on further GEN_BE CV7 cleanup, including fixing a segmentation fault
Testing radar bug fixes sent by Ying Zhang
Working on some compilation problems reported with MADIS2LITTLER
Answering WRFHELP questions (3):
External user (Texas A&M University) reported above MADIS2LITTLER problems
External user (Zhou Lilong from CMA National Meteorological Center) reported bug in 4DVAR compilation for AIX
Helped user with questions about LITTLE_R and BUFR observations
Activities for week of December 21, 2015
Three days of PTO, one day Holiday
WRFHELP:
Answered one question about assimilating surface observations
Made 4 commits:
Finally got dynamic constraint code into trunk
Committed 4DVAR compilation fix for AIX from Zhou Lilong
Radar bug fixes from Ying Zhang
More CV7 cleanup changes, fixing segmentation fault
Activities for week of December 28, 2015
One day of PTO, one day Holiday
Made comments on AMSR2 paper
Added AMSR2 regression test (still only works for Intel on Yellowstone)
Still trying to replicate MADIS2LITTLER issue, need to hear back from user
Working on further cleanup of HDF5 compilation, eliminating warning messages from WRF/WRFPLUS compilation
WRFHELP questions (4):
Assimilating text and prepbufr obs at once
Viewing be.dat stats in real space (Thomas Schwitalla, University of Hohenheim)