Those following this story in previous updates (https://rantransproject.com/a-shot-from-above-satellites-and-drones/) can see our leaps of progress.

Philippe and Mael are refining the algal mat spatial detection toolkit. From their base in ARGANS France, they are accessing better data from the Planet company in addition to data from sentinel 2. This new data is most costly but includes; very high spatial resolution satellite images (about 3m), daily repetitiveness, atmospheric correction, and 8 bands camera sensors, compared to 4 bands previously, that are producing sharper and more accurate images.

Also included in this piece of work are aerial overflights data and images from partners such as CEVA drone in Baie des Veys and Ledano cove, as well as NE drone in Langstone and Poole harbours; that are carried out each year on the French (CEVA) and English (NE) coast. This is important, allowing the team to order in priority the “Planet” images that best corresponds with overflight dates and locations.

The smart efforts from the team continued into developing a new methodology that includes complex calculations and applying what they call “the Fuzzy classification”. So, using one or more overflight reference images, the new method selects the set of pixels containing only green algae and calculate an average value that is then represented on a vector (blue line). Lots of training data is being used to tune this system to be most accurate, and to obtain a fuzzy index data between 0-1.

Results of very-high resolution satellite images of algal mat are created in the various following stages;

In the first stages the reference image is used to create reference dataset by selecting the green algae pixels

RGB Planet image of 13/07/2022 (Lédano Cove- The Trieux river)

Next, the reference dataset showing only the selected algae pixels is created

(Lédano Cove- The Trieux river)

Fuzzy index calculation on the reference image.

(Lédano Cove- The Trieux river)

Using the fuzzy index that is higher than 0.4 to classify areas with green algae

(Lédano Cove- The Trieux river)

The approach is also applied in Holes Bay, Poole harbour, England for green algal mat identification

In Holes Bay Poole Harbour, using the fuzzy index that is higher than 0.4 to classify areas with green algae

In Holes Bay Poole Harbour, Saltmarshes are being classified using the fuzzy index that is higher levels than that use to classify areas with green algae

(Saltmarsh identification -Holes Bay, Poole Harbour)

This work continues in 2023 to better identify saltmarshes, as well as seagrass, and intertidal substrates (sand or mud).

Find out more about RaNTrans by signing up to our Newsletter (rantransproject.com/#signup) visiting the website (rantransproject.com) and following us on twitter (@ProjectRaNTrans)

Have a lovely Christmas and looking forward to working with you in the New Year 2023.