access_JEDIAtNCI_MiniAcademyJune2020_Activity2A - ACCESS-NRI/accessdev-Trac-archive GitHub Wiki

Interactive activity 2A

This covers "Tuesday A: Adding new observations to be assimilated"

Steps 1

Follow the instructions in Monterey tutorial taking into account the differences in the file locations and names. For example, to produce a plot of brightness temperature observations from channel 15 of AMSU-A onboard NOAA-19 (instead of AMSU-A/MetOp-A as is the case in the Monterey tutorial) you will use,

hofx.py -f Data/hofx/amsua_n19_hyb-3dvar-gfs_2018041500_m_NPROC.nc4 -v brightness_temperature_15@ObsValue

Make similar changes and produce plots of model background in observation space, cost function, innovation and analysis increments.

Move the plots to the Day2_Run1 directory as suggested in the Monterey tutorial.

Next, do the suggested exercise in the Monterey tutorial, which looks at observation cost (Jo). Bear in mind that the log file that is written out from our hybrid 3DVar run is hyb-3dvar_gfs.run

Steps 2

We will need to change this step slightly since adding AMSU-A/Aqua to the basic yaml configuration as suggested in the Monterey tutorial causes a problem with CRTM. Instead we will add AMSU-A from Metop-A to our yaml configuration file so that the radiance observations from this instrument are assimilated in our hybrid 3DVar.

Follow the instruction in the Monterey tutorial but note some changes,

• you are adding AMSU-A from Metop-A instead of AMSU-A from Aqua
• the file that contains the observations of AMSU-A/Metop-A in our case is,

  Data/obs/testinput_tier_1/amsua_metop-a_obs_2018041500_m.nc4
  

• assimilate channels, 4 through to 6 and 9 through to 14

Steps 3

Follow the Monterey tutorial instruction paying attention to the difference in the run script and file names, etc.

Repeat Step 1 to generate various plots. Follow the rest of Monterey tutorial.

Steps 4

Noting that we are adding observation filtering for AMSU-A/MetOp-A follow the instructions in the Monterey tutorial.