|
| 1 | +GFS V16.1.4 RELEASE NOTES |
| 2 | + |
| 3 | + |
| 4 | +PRELUDE |
| 5 | + |
| 6 | + NOAA awarded Delivery Order 3 (DO-3) of its commercial radio occultation |
| 7 | + (RO) data purchase to Spire Global on August 12, 2021. This purchase |
| 8 | + covers 3000 occultations per day over a six month period with the data |
| 9 | + flow starting on September 16, 2021. On the same date, the GeoOptics |
| 10 | + data from Delivery Order 2 (DO-2) will be discontinued. |
| 11 | + |
| 12 | + GNSSRO bending angle observations from GeoOptics and Spire were initially |
| 13 | + evaluated as part of Delivery Order 1 (DO-1), covering a 30-day period |
| 14 | + starting on December 15, 2020. The data from both vendors were found to |
| 15 | + be of similar quality to other operationally assimilated RO data. Low |
| 16 | + resolution experiments informed the configuration of the quality control |
| 17 | + and observation errors for the following DO-2, which was awarded to |
| 18 | + GeoOptics only, covering an average of 1300 occultations per day over a |
| 19 | + six month period starting on March 17, 2021. These data began to be |
| 20 | + assimilated operationally with the implementation of GFSv16.1.1 on |
| 21 | + May 22, 2021 after two months of parallel testing at full resolution. |
| 22 | + |
| 23 | + Following the v16.1.1 implementation, the observation errors for |
| 24 | + commercial RO were tuned to better utilize the new data. Additionally, |
| 25 | + two fixes related to the handling of super-refractivity in the RO bending |
| 26 | + angle observation operator were added. These fixes resolve minimization |
| 27 | + issues that were related to the increased vertical resolution in v16. |
| 28 | + Therefore, the stricter gross check quality control measures originally |
| 29 | + put in place to mitigate minimization problems will now be relaxed. |
| 30 | + |
| 31 | + This implementation plans to: |
| 32 | + * Turn off the active assimilation of GeoOptics data |
| 33 | + * Turn on the active assimilation of Spire data |
| 34 | + * Adjust the observation errors for the Spire data |
| 35 | + * Address two issues within the RO bending angle observation operator |
| 36 | + * Relax the strict gross check added because of those issues. |
| 37 | + |
| 38 | + These changes only affect two files within the GSI tag of the global workflow. |
| 39 | + |
| 40 | + In addition, a small bug fix is required to correct the ingest of new BUFR |
| 41 | + observations from ships that was previously causing erroneous observations |
| 42 | + to be assimilated. This alters one additional file in the GSI tag. |
| 43 | + |
| 44 | +IMPLEMENTATION INSTRUCTIONS |
| 45 | + |
| 46 | + The NOAA VLab and both the NOAA-EMC and NCAR organization spaces on GitHub.com |
| 47 | + are used to manage the GFS.v16.1.4 code. The SPA(s) handling the GFS.v16.1.4 |
| 48 | + implementation need to have permissions to clone VLab gerrit repositories and |
| 49 | + the private NCAR UPP_GTG repository. All NOAA-EMC organization repositories are |
| 50 | + publicly readable and do not require access permissions. Please follow the |
| 51 | + following steps to install the package on WCOSS-Dell: |
| 52 | + |
| 53 | + 1) cd $NWROOTp3 |
| 54 | + |
| 55 | + 2) mkdir gfs.v16.1.4 |
| 56 | + |
| 57 | + 3) cd gfs.v16.1.4 |
| 58 | + |
| 59 | + 4) git clone -b EMC-v16.1.4 https://github.com/NOAA-EMC/global-workflow.git . |
| 60 | + |
| 61 | + 5) cd sorc |
| 62 | + |
| 63 | + 6) ./checkout.sh -o |
| 64 | + * This script extracts the following GFS components: |
| 65 | + MODEL tag GFS.v16.0.16 Jun.Wang@noaa.gov |
| 66 | + GSI tag gfsda.v16.1.4 Russ.Treadon@noaa.gov |
| 67 | + GLDAS tag gldas_gfsv16_release.v1.12.0 Helin.Wei@noaa.gov |
| 68 | + UFS_UTILS tag ops-gfsv16.0.0 George.Gayno@noaa.gov |
| 69 | + POST tag upp_gfsv16_release.v1.1.4 Wen.Meng@noaa.gov |
| 70 | + WAFS tag gfs_wafs.v6.0.22 Yali.Mao@noaa.gov |
| 71 | + |
| 72 | + 7) ./build_all.sh |
| 73 | + * This script compiles all GFS components. Runtime output from the build for |
| 74 | + each package is written to log files in directory logs. To build an |
| 75 | + individual program, for instance, gsi, use build_gsi.sh. |
| 76 | + |
| 77 | + 8) ./link_fv3gfs.sh nco dell |
| 78 | + |
| 79 | + |
| 80 | +SORC CHANGES |
| 81 | + |
| 82 | +* sorc/ |
| 83 | + * checkout.sh will checkout the following changed model tags: |
| 84 | + * GSI; tag gfsda.v16.1.4 |
| 85 | + * src/gsi/setupbend.f90: Add two fixes for handling of |
| 86 | + super-refractivity layer, change the observation errors |
| 87 | + of the commercial data, relax stricter gross check back to |
| 88 | + original values |
| 89 | + * src/read_nsstbufr.f90: Changes to handle NC001013, NC001101 |
| 90 | + and NC001113 marine bufr subsets. |
| 91 | + |
| 92 | + |
| 93 | +FIX CHANGES |
| 94 | + |
| 95 | +* fix/fix_gsi: |
| 96 | + * global_convinfo.txt: Discontinue assimilating GeoOptics and |
| 97 | + begin assimilating Spire, relax stricter gross check back to |
| 98 | + original values. |
| 99 | + * gfsv16_historical/: Add fix files for retrospective parallels. |
| 100 | + Does not impact operations. |
| 101 | + |
| 102 | + |
| 103 | +PARM/CONFIG CHANGES |
| 104 | + |
| 105 | +* parm/config/config.anal: Add historical fix file entries. Does |
| 106 | + not impact operations. |
| 107 | + |
| 108 | + |
| 109 | +JOBS CHANGES |
| 110 | + |
| 111 | +* No change from GFS v16.1.3 |
| 112 | + |
| 113 | + |
| 114 | +SCRIPT CHANGES |
| 115 | + |
| 116 | +* No change from GFS v16.1.3 |
| 117 | + |
| 118 | + |
| 119 | +CHANGES TO RESOURCES AND FILE SIZES |
| 120 | + |
| 121 | + There should be no change in analysis runtime nor cnvstat file size |
| 122 | + greater than the normal cycle to cycle variation. |
| 123 | + |
| 124 | + |
| 125 | +PRE-IMPLEMENTATION TESTING REQUIREMENTS |
| 126 | + |
| 127 | +* Which production jobs should be tested as part of this implementation? |
| 128 | + * The entire GFS v16.1.4 package needs to be installed and tested. |
| 129 | + |
| 130 | +* Does this change require a 30-day evaluation? |
| 131 | + * No. |
| 132 | + |
| 133 | + |
| 134 | +DISSEMINATION INFORMATION |
| 135 | + |
| 136 | +* Where should this output be sent? |
| 137 | + * No change from GFS v16.1.3 |
| 138 | + |
| 139 | +* Who are the users? |
| 140 | + * No change from GFS v16.1.3 |
| 141 | + |
| 142 | +* Which output files should be transferred from PROD WCOSS to DEV WCOSS? |
| 143 | + * No change from GFS v16.1.3 |
| 144 | + |
| 145 | +* Directory changes |
| 146 | + * No change from GFS v16.1.3 |
| 147 | + |
| 148 | +* File changes |
| 149 | + * No change from GFS v16.1.3 |
| 150 | + |
| 151 | + |
| 152 | +HPSS ARCHIVE |
| 153 | + |
| 154 | +* No change from GFS v16.1.3 |
| 155 | + |
| 156 | + |
| 157 | +JOB DEPENDENCIES AND FLOW DIAGRAM |
| 158 | + |
| 159 | +* No change from GFS v16.1.3 |
| 160 | + |
0 commit comments