This summer I worked with UMass Amherst´s Salt Marsh UAS team, which is funded by the EPA and started in 2018 headed by Scott Jackson and Charlie Schweik. We fly drones over the salt marshes along the Massachusetts coast and collect images of the marsh area. We then take those images and stitch them together in Agisoft Metashape to create DEMs and orthomosaics. The drones we use have sensors attached which take high resolution images in near infrared, shortwave infrared and longwave infrared which shows absorption of light and inundation of water retention in the salt marsh. We use this spectral data to analyze and compare how the marsh is changing and adapting. We create the DEMs and orthomosaics for the classification model we train. The classification model takes these orthomosaics and classifies the landscape into either vegetation, water features or bare ground.
For my own individual project, I delineated bare ground areas at one of our sites at Peggotty Beach in Scituate, MA. I then took the data points I collected and input them into QGIS. I used my data on the classification model to see if it was accurately classifying the landscape. I also used the data I collected to compare the bare ground areas from 2019 to present day to see how those areas have changed over time.