Following our last successful trip to Lulworth Cove, Dorset, we decided it was worth taking another trip to the Jurassic coast for a second data gathering mission. Last Wednesday (23rd Aug) our team met at ZSL (London Zoo) in the early hours of the morning to collect our transport for the day (Fig 1). We headed south from London in the wee hours of the morning in an attempt to beat morning rush hour. After a slow start, we were on our way to the Jurassic coast once again!
Fig 1. Our transport with kayak in-tow on the shore of Kimmeridge Bay
We arrived in Weymouth just before noon, to the Jurassic Coast Activities HQ, where we picked up our kayak for the day. We turned around and headed back east, kayak in-tow (Fig 1), along the coast to Kimmeridge bay, hoping that the weather forecast had stayed true. As any mariner will attest; the sea forecast rarely stays true and unsurprisingly it was not the wind direction or speed we would have liked, nevertheless we began readying ourselves on the shore line (Fig 2). Unlike our last ground-truthing trip to Lulworth, we had decided to not gather singlebeam sonar data and therefore, we were able to focus on gathering good quality video footage for ground-truthing.
Fig 2. Preparing our camera set-up on the shoreline
Following our last trip to Lulworth Cove, we were able to use the ground-truthing data gathered, along with processed multibeam sonar data to predict kelp habitat on the Dorset coast (using species distribution models or SDMs). However, to improve classification success, we decided it was worth gathering more ground-truth information to further train our model. We settled on Kimmeridge Bay due to a number of factors i.e. ease of access and coverage extent of multibeam sonar data.
After a quick bite to eat and after we had ‘gotten our bearings’, we were ready to go. Before heading out, we made a quick stop into the Purbeck Marine Wildlife Reserve & Fine Foundation Marine Centre to speak to some of the local experts about kelp distribution in the bay. Our trip was in two parts; we needed to collect video footage of kelp coverage (presence and absence), but following that, we wanted to collect Asparagopsis armata samples for an on-going project called ASPAZOR. Following our stint in the kayak, we planned to snorkel in the bay and collect what samples we could. However, after speaking to the staff in the reserve centre, we confirmed that our chances of finding the red algae in the Bay were relatively slim.
Fig 3. The GPS track line of the route we took in Kimmeridge Bay
Prior to leaving on our trip we had examined the extent of multibeam sonar coverage in Kimmeridge Bay, and so, we we’re able to plan our survey track accordingly. Although we deviated somewhat (Fig 3), using the planned route allowed us to stay within areas of multibeam sonar coverage. With the GPS in hand guiding us over areas where we believed there to be kelp, and the camera in tow, we gathered as much good footage as possible. We set out on our two man kayak to collect what footage we could.
Fig 4. A ‘fucoid waterfall’ spotted on the shore in Kimmeridge Bay at low tide
At first it was unclear if our set up was working as well as we hoped. After a few trial runs we were able to confirm the camera was capturing footage of the seafloor, whether we were capturing footage of kelp or not was another question. We collected as much footage as we could, in areas where we believed kelp was present and absent. When we were finished we headed back to shore to kit up again, in our snorkel gear and search for Asparagopsis.
The visibility was less than ideal, which was a worry for our video footage but upon leaving the Bay the video footage was clearer. However, we were limited in our ability to strive far while snorkelling and time had gotten the better of us. After a quick search, without any luck, we headed back to shore to make our trip back to London.
Fig 5. A series of short clips of the seabed in Kimmeridge Bay (spot the Wrasse at 0:45)
We packed up and headed back towards Weymouth to drop our kayak back to Jurassic Coast Activities, after which we began the long drive back to London. On the way back we were able to confirm the trip had been a success following a more extensive review of the footage (Fig 5). Fingers-crossed (with the good footage we’ve collected) we will be able to better train our model/ verify the results of our previous outputs.
Watch this space!
Fig 6. Kimmeridge Bay, Dorset