Fixed potential memory leak in local_solar_time

[schlunma]

Description

See SciTools/iris#5767: wrapping large arguments, such as numpy arrays in a functools.partial is an example of how functions can be created that will make this into a sizeable memory leak. This fixes this by using all numpy arrays as positional arguments in dask.array.apply_gufunc.

Closes #2355

---

Before you get started

• ☝ Create an issue to discuss what you are going to do

Checklist

It is the responsibility of the author to make sure the pull request is ready to review. The icons indicate whether the item will be subject to the 🛠 Technical or 🧪 Scientific review.

• 🧪 The new functionality is relevant and scientifically sound
• 🛠 This pull request has a descriptive title and labels
• 🛠 Code is written according to the code quality guidelines
• 🧪 and 🛠 Documentation is available
• 🛠 Unit tests have been added
• 🛠 Changes are backward compatible
• 🛠 Any changed dependencies have been added or removed correctly
• 🛠 The list of authors is up to date
• 🛠 All checks below this pull request were successful

---

To help with the number pull requests:

• 🙏 We kindly ask you to review two other open pull requests in this repository

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 94.11%. Comparing base (2be264d) to head (72859ef).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2356      +/-   ##
==========================================
- Coverage   94.11%   94.11%   -0.01%     
==========================================
  Files         241      241              
  Lines       13381    13379       -2     
==========================================
- Hits        12594    12592       -2     
  Misses        787      787              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[bouweandela]

Thanks @schlunma Code looks good to me. I also did a test run. I did notice some strange vertical lines with very low values in them though. Here is the recipe:

datasets:
  - {dataset: ERA5, project: native6, tier: 3, type: reanaly}

preprocessors:
  local_solar_time:
    local_solar_time:

diagnostics:

  map:
    description: Global map of temperature in January 2000.
    themes:
      - phys
    realms:
      - atmos
    variables:
      pr:
        mip: E1hr
        preprocessor: local_solar_time
        timerange: 19900102/P2D
        caption: |
          Global map of {long_name} in January 2000 according to {dataset}.
    scripts:
      script1:
        script: examples/diagnostic.py
        quickplot:
          plot_type: pcolormesh
          cmap: Reds

and the plot (plotted time step 23), the white lines have values of order 1e-18:

This does not appear to be a new issue though, it also happens with the main branch.

[schlunma]

Hmm, when I was testing this I've never seen these weird lines. I think these are missing values, not very small values. My guess is that it's somehow related to rounding, which leads to some longitude/time combinations having missing values. When this dataset is regridded to a course grid, these lines are gone.

Unfortunately I don't have the time to investigate this further, but this also shouldn't be a big issue to be honest, except if someone is interested in only these longitudes bands individually.