One fish, two fish… but what are those fish up to? A new website invites citizen scientists to roll up their digital sleeves and get their hands into the ocean to help biologists identify seafloor life and habitats.
The website, Seafloor Explorer, has a gallery of 100,000 images, taken by a habitat mapping underwater vehicle, HabCam. The vehicle is slowly towed over the seafloor, shooting continuous images. The images are then color-corrected and stitched together to create an accurate picture of the life and health of the continental shelf. “The novelty of this is that because it’s a continuous sampling method, that’s been the key to make this a large-scale survey that can create data points on a continental shelf scale,” Woods Hole biologist Amber York told Co.Exist.
Not only is the data being used for a scientific count of scallops, it’s also going to be part of an overall project to monitor the seafloor, using citizen scientists. Since the website launched a week ago, those images have been viewed more than 450,000 times. Each picture needs to be annotated several times, explained York.
The website asks users to classify everything they see in a photo, from the type of ground cover (sand, shells, boulders, cobble, or gravel) to the size and type of species in each picture. They hope to use information to better understand the distribution of scallops and other resources relative to their predators such as starfish and the composition of their habitat.
Seafloor Explorer is part of a growing trend that relies on the public to help scientists with tedious tasks, from finding galaxies to transcribing ancient Greek texts. Chris Lintott, director of Zooniverse, which develops interactive science sites and helped create Seafloor Explorer, says that digital technology has made citizen involvement possible. “Traditional citizen science–birdwatching, for example–involved volunteers making observations which were then mostly analyzed and used by professionals. These days, obtaining data is often automated, but we need volunteers to sort through the huge piles of images and other data that result.”
Lintott says the future will be even better for citizens and scientists to work together and also to train machines to identify data sets. “I can imagine a situation where we take live data from a telescope, or from a camera observing a natural habitat, and when something unexpected happens we ask thousands of people around the world to take a look right away. I also think you’ll see projects that combine human and machine classifications, making the best of both approaches to get all the information we can out of a dataset.”
Ultimately, the situation is a win-win: Scientists get an accurate reflection of what lies on the seafloor, and citizens get to be involved in new research. Already the Citizen Science Alliance has contributed data that has spawned 40 peer-reviewed science papers.