Activities for week of November 2, 2015
- Talked with Carter to try to finally get the website to properly link www links to www2; nothing fixed yet
- Adding Hybrid regression test capability (not finished yet)
- Discovered potential problem with WRF 3.7.1 when setting up hybrid capability; discussing with Wei
- Additionally discovered segfault problem with gen_be_vertloc.exe
- Separately discovered a problem with WRF/WRFPLUS/WRFDA for compilation on AIX platforms; collaborating with Dave on a fix
- Finished script for automatically generating WRF ensembles for Hybrid tests
Activities for week of November 9, 2015
- Changes to trunk:
- Fixing segfault problem for gen_be_vertloc.exe
- Fixing WRF/WRFPLUS compilation problems on some AIX systems
- Finished adding Hybrid regression test capability
- New test is a case using the realtime domain with a subset of 6 ensemble members
- Added new AMSR2 regtest as well (not finished; still issues with HDF5 on yellowstone)
- Performing final tests on dynamic constraint branch
- Moved dynamic constraint calculations to their own separate subroutines, restored old "da_balance" code even though it is unused
- Discovered problem with indexing in dynamic constraint equation
- Made some minor updates/fixes to user's guide
- Got back to investigating incorrect output for print_detail_be...thought I had a solution, but there may still be issues that need addressing
Activities for week of November 16, 2015
- Contacted CISL about HDF5 problems; still investigating issues with code
- Fixing a whole bunch of issues that were causing regression tests to fail with debugging flags turned on
- Performing final tests on dynamic constraint branch
- Consolidated some dynamic constraint equation subroutines; most are too short to need their own subroutines
- Simplified do loops and removed unnecessary intermediate variables; ran many tests to ensure no change in results
Activities for week of November 23, 2015
- Holiday Thursday and Friday
- Changes to trunk:
- Fixing printout problems for print_detail_be....still think there are some issues with the lengthscale values!
- Fixing some problematic code with VARBC, radar routines that cause debug failures
- Fix a couple issues with recently modified surface diagnostics code
- Check for division by zero in wind speed/direction routines
- Performing final tests on dynamic constraint branch
- Investigating ways to properly handle boundary conditions for centered-differencing in adjoint
- Made final cleanup commits, will propose next week
- Investigating NETCDF4 long runtime issues
Activities for week of November 30, 2015
- Fixed compilation problem for WRFDA on Yellowstone with HDF5 support
- Heard back from CISL about HDF5 problems; still can only get it to work for intel but they are not very helpful about this problem!
- Cleaning up include statements in WRFDA code
- Reading HDF5 and WRFPLUS information from configure.wrf file
- Set variables so that there are no warnings for non-existent directories when HDF5/WRFPLUS are not set
- Fixing cv7 code which should have been fixed last year.
- Naming the control variables properly
- Skipping unnecessary calculations of regression coefficients for u and v
- Adding a whole bunch of if statements to ensure cv5 and cv6 calculations and variable names DO NOT change
- Fixing interface to WRFDA code so that variables are named correctly there as well; printouts now print the correct variable names for each option