Refinement of thresholding

[rosepearson]

• Leigh looking into indices to better distinguish between thin cloud and ocean glint
• We may move to a two phase
  1. Detect bed centres with on set of indices that are resistant to thin cloud and ocean glint
  2. Buffer around these areas an apply a more sensitive set of indices to pick up the complete bed
• Once indices selected we may look at developing a complete kelp extent map based on all beds (could filter by area) for 2016-2024. We can then filter to only process days without cloud cover or other non-water or non-land SCL classifications over kelp beds.