Water Observations tables

[benji-glitsos-ga]

• Improving the bit set table
  • Renamed 'Flagging' column to 'Classification'
  • Used title case for these capitalisations to match DEA Maps.
  • Added plain language descriptions based on Bex's notes.
  • Rearranged columns of table so that Classification is the first column and Bit and Decimal value are next to each other.
  • Converted to list-table code format for ease of editing.
• Added cross-over table based on Bex's spreadsheet.
  • Added documentation based on Bex's notes.

Preview: https://pr-381-preview.khpreview.dea.ga.gov.au/data/product/dea-water-observations-landsat/?tab=description#technical-information

[github-actions]

Preview is ready!

View the preview:

🌐 PR Preview 381

You can also view the deploy log or read the documentation.

[benji-glitsos-ga]

Caitlin has confirmed that all the values in the 'Combination decimal values' table are correct.

[BexDunn]

Looks good to me Benji

[BexDunn]

Looks good team

[robbibt]

@BexDunn Thoughts on whether we might want to simplify the table to remove these bespoke double-flag options? My concern is that they make things more complicated than need be - they kind of infer the presence of a "Cloud + water" flag where none exists. It might also add confusion as to why certain other useful combinations (e.g. cloud + shadow) aren't included in the table.

Alternatively, we could just include the main 8 bits (cloud, shadow etc) on the table, and then discuss even more complicated options in the paragraph above it.

[BexDunn]

Hey @robbibt

• The absence of Cloud+Shadow is either due to me missing that combination or I may have thought that the two were mutually exclusive bits for some reason. Are any other combinations missing?
• The original inspiration was the colour map that is used on DEA Maps, which maybe we should also add into this information somehow. The colour map has more options than the original 8 bits in the data where decimal values correspond to multiple bitflags active:

• The original purpose of the table was to work out all the mathematically possible values (not necessarily all the values actually physically possible) to help diagnose an issue that a stakeholder had where they had values of 84 present in the data. I started with the colour map, but as the combination bitflag values weren't explained or quantified anywhere, I needed to write the table to understand what was happening. I think the table could be useful, but we could split it into two if it isn't helpful, or line it up against the colour map somehow to make the explanation clearer.

[benji-glitsos-ga]

Now that I look at the colour map, I realise I have more work to do in making this easy to read:

• Terms like 'Cloudy Water' really mean 'Cloud + Water', so I should represent that somehow.
• I should use the colour coding on the table to make it easier to use

I'll open a new PR soon to improve it.