Implemented fully lazy climate_statistics

[schlunma]

While working on #1193 I found that the preprocessor climate_statistics is not lazy if period='full' and a weighted operator is used. The reason is the use of iris.util.broadcast_to_shape here, which broadcasts the time weights to the original shape of the cube.

However, since the release of iris v3, this broadcasting is no longer necessary. PR SciTools/iris#3943 introduced the support of 1D weights for collapsing cubes. This PR implements this for our preprocessor.

Note that while our implementation might have worked previously for multidimensional time arrays, the calculation of the weights didn't make sense at all. Therefore, in this new implementation an error is raised if the time coordinate is not 1D. However, I don't even know if there is any use case for this.

I tested some recipes and all worked fine 👍 I also didn't modify the climate_statistics tests at all, everything worked fine with the new implementation.

Closes #336.

Link to documentation:

---

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

Merging #1194 (09f2809) into main (3b653c4) will decrease coverage by 0.00%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #1194      +/-   ##
==========================================
- Coverage   85.51%   85.51%   -0.01%     
==========================================
  Files         188      188              
  Lines        9147     9146       -1     
==========================================
- Hits         7822     7821       -1     
  Misses       1325     1325              

Impacted Files	Coverage Δ
esmvalcore/preprocessor/_time.py	94.13% <100.00%> (-0.03%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 3b653c4...09f2809. Read the comment docs.

[schlunma]

I think codecov can be ignored here: The coverage of the changed function is 100%, but the relative coverage of the entire file decreased since the total number of lines decreased (and thus the number of hits)...Would be nice if codecov was smart enough to catch this!

[bouweandela]

Looks good! I'll have a look at improving the codecov configuration so the checks are useful, see #1195