����ֱ��

The CRCA would like to congratulate Jillian Carr ’21 on the completion of her thesis. Jill was a Geo-Information Science Master of Science student at ����ֱ�� who works in the field of marine science for the Massachusetts Bays National Estuary Partnership. After taking Professor Stephen Young’s “Drones” class, she knew she wanted to investigate a gap she saw in the mapping of marine ecosystems in existing Massachusetts state programs. She developed a study using drones to map several coastal and marine habitats and evaluated how the technology could fit into programs that already exist for state mapping, and to supplement the important work that was already being implemented.

People have been using drones to map marine habitats in other parts of the world, Jill notes, with “great success in places where water clarity is good” such as in the Caribbean. However, drone mapping is not as well-tested in Massachusetts due to the murkier, more turbid quality of the waters. Jill recalls her time researching and testing for her thesis was difficult and took the majority of two summers and two winters to complete. She used footage collected in the summer versus the winter to concretely understand how the habitats would change between seasons and over time.

Using state-produced maps that already existed as her reference datasets, Jill learned to train the image analysis software to identify and differentiate between habitats present in her drone imagery. She explains that this aspect of her project was fascinating. She notes that eelgrass, a critical habitat for local marine life that’s in decline, was particularly difficult to map, and the high-resolution maps she produced from her footage created, much to her delight, extremely accurate data. One of the strongest suggestions she makes in the discussion component of her thesis was that a drone program could supplement existing state programs to document the changing nature of eelgrass over time at the meadow-scale and could be used to collect more frequent imagery than the state programs currently collect. Jill sees this as a readily available and relatively inexpensive solution that would help understand and protect this crucial aspect of our local marine ecosystems.

In terms of the machine-learning component, Jill notes that she was navigating a fairly steep learning curve since the software she had at her disposal wasn’t built for use with drones. However, one of the biggest benefits of her thesis, in her view, is the workflow system she developed that will hopefully make it a lot easier for other scholars and conservationists to replicate her process for similar projects in the future.

We thank Jill for her important work, and we can’t wait to see the continuation of her astounding accomplishments into the future of her career.