Add min error of 0.1 for airnow

[huiying-noaa]

Add min error of 0.1 for AirNOW. Please let me know if there's any comments.

Issues: current version has 0 error in our development test for PM2.5 DA for RRFS-SD https://github.com/NOAA-EMC/rrfs-workflow/tree/rrfs-mpas-jedi. The RDASApp used for RRFS DA directly points to this repo.

Simple PR, no dependency.

I have performed a self-review of my own code and tested results.
Original PM2.5 error: _, 0.05, 0.47, 0, 0.12
New PM2.5 error: _, 0.1, 0.47, 0.1, 0.12.

[EricJames-NOAA]

Looks good to me! Thank you for opening this PR.

[guoqing-noaa]

@huiying-noaa We will need to make sure the CI tests pass first.

[huiying-noaa]

Hi @weiwilliam, Do you have any leads on the ci test error? Many thanks!

Node 20 is being deprecated. This workflow is running with Node 24 by default. If you need to temporarily use Node 20, you can set the ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION=true environment variable. For more information see: https://github.blog/changelog/2025-09-19-deprecation-of-node-20-on-github-actions-runners/
Run actions/create-github-app-token@v1
/home/runner/work/_actions/actions/create-github-app-token/v1/dist/main.cjs:42559
throw new Error("Input required and not supplied: private-key");
^

Error: Input required and not supplied: private-key
at Object. (/home/runner/work/_actions/actions/create-github-app-token/v1/dist/main.cjs:42559:9)
at Module._compile (node:internal/modules/cjs/loader:1854:14)
at Object..js (node:internal/modules/cjs/loader:1985:10)
at Module.load (node:internal/modules/cjs/loader:1577:32)
at Module._load (node:internal/modules/cjs/loader:1379:12)
at wrapModuleLoad (node:internal/modules/cjs/loader:255:19)
at Module.executeUserEntryPoint [as runMain] (node:internal/modules/run_main:154:5)
at node:internal/main/run_main_module:33:47

Node.js v24.16.0

[eap]

The problem with CI in this PR is because the PR is coming from a fork which does not have access to our org secrets. I just talked to Steve H who noted that this repo is expected to have future pull requests from collaborators without JCSDA write access. Therefore I will adopt the pull_request_target approach currently used by CRTMv3 which allows for testing of forks by running a checked-in version of the workflow.

I've created the PR #1794 to address the issue. Once that PR is merged, you can sync this and the tests should run as expected.

[weiwilliam]

@huiying-noaa You need to create new testoutput. You can check

ioda-converters/test/CMakeLists.txt

Lines 1528 to 1551 in cea25d6

	ecbuild_add_test( TARGET test_${PROJECT_NAME}_airnow
	TYPE SCRIPT
	ENVIRONMENT "PYTHONPATH=${IODACONV_PYTHONPATH}"
	COMMAND bash
	ARGS ${CMAKE_BINARY_DIR}/bin/iodaconv_comp.sh
	netcdf
	"${Python3_EXECUTABLE} ${CMAKE_BINARY_DIR}/bin/airnow2ioda_nc.py
	-i testinput/airnow_2020081306.dat
	-s testinput/airnow_sites.dat
	-o testrun/airnow_2020081306.nc"
	airnow_2020081306.nc ${IODA_CONV_COMP_TOL})
	ecbuild_add_test( TARGET test_${PROJECT_NAME}_airnow_epalist
	TYPE SCRIPT
	ENVIRONMENT "PYTHONPATH=${IODACONV_PYTHONPATH}"
	COMMAND bash
	ARGS ${CMAKE_BINARY_DIR}/bin/iodaconv_comp.sh
	netcdf
	"${Python3_EXECUTABLE} ${CMAKE_BINARY_DIR}/bin/airnow2ioda_nc.py
	-i testinput/airnow_2020081306.dat
	-s testinput/airnow_sites_epalist_20240716.csv
	--epa_list
	-o testrun/airnow_2020081306_epalist.nc"
	airnow_2020081306_epalist.nc ${IODA_CONV_COMP_TOL})

for reference.
After you created the new testoutputs, replace the test/testoutput/airnow_2020081306.nc and test/testoutput/airnow_2020081306_epalist.nc with them.
Then push another commit to your branch.
Thanks!

[huiying-noaa]

Thanks for checking and providing the instructions, @weiwilliam!

[BenjaminRuston]

@huiying-noaa and note you can specify error dynamically, assume you are doing this, but adding this to a yaml inserts a value which doesn't necessarily even need to be in the input file:

  # Assign the initial observation error (constant value, 3.5 m/s right now).
  - filter: Perform Action
    filter variables:
    - name: windEastward
    - name: windNorthward
    action:
      name: assign error
      error parameter: 3.5

[mer-a-o]

This assigns 0.1 for all species. Is it possible to make it for only PM2.5? This converter is used for other species such as O3 and NO2 and 0.1 may not be a suitable for these species. This requires some more changes to the converter but can address the issue properly.

[huiying-noaa]

Thanks for your comments, @BenjaminRuston and @mer-a-o! I'll first try adding minimum error in yaml to achieve the same.

[BenjaminRuston]

@huiying-noaa do not believe you've got the species key capability added as yaml parameter correct?

[huiying-noaa]

Hi @BenjaminRuston , I'm mostly on vacation and sick leave after the last discussion. For the few tests I was able to conduct, the conditional error assignment in yaml by filtering with ObsValue or ObsError of pm2.5 hasn't been achieved yet. I do see the error assign process in log, but values all remain the same. I'll do more tests early next week and report back.

[huiying-noaa]

Hello all, Many thanks for all your input. I was able to achieve the same results using yaml settings. All communication is helpful for my future commits. Thanks again!
Closing this PR now.