VICKSBURG, Miss., — Researchers from the U.S. Army Engineer Research and Development Center (ERDC), along with the U.S. Geological Service, Marda Science and James Madison University, are engaging citizen scientists in a national SandSnap initiative to amass a spatial and temporally varying nationwide beach grain-size database.
Across the U.S., coastal shorelines are data-rich environments; however, there is currently no nationwide beach grain-size database.
Many of our coastal areas require active management to balance environmental, economic, health and human activities. Sand grain size impacts not only beach shape, but also how easily sediment erodes. Resource managers need this data to make sound management decisions.
“You can get high-resolution airborne light detection and ranging data and shoreline position from satellite and arial imagery on the position and shape of beaches for long stretches of shoreline, as well as wave information from buoys, wave information studies and coastal hazard systems,” said Dr. Brian McFall, a research civil engineer with the ERDC’s Coastal and Hydraulics Laboratory. “With these different techniques, you can get a decent understanding of the hydrodynamic forces acting on our beaches, but really the big key data piece that we’re missing is what the beach is actually made out of — that’s the void we are trying to fill.”
The high costs of collecting beach grain size data over a large geographical area with traditional methods, such as sample collection and sieve analysis, make it impracticable. Engaging citizen scientists in scientific information collection is a more cost-effective way to collect large amounts of data while increasing interest in research through public engagement. Previous work by McFall and his partners demonstrated the feasibility of creating a nationwide database of high-fidelity beach grain size data generated by analyzing sediment imagery obtained from smartphone-equipped citizen scientists.
The images will be analyzed using machine learning, or a neural framework, to determine the sediment gradation, and the results will be stored for easy online access. The ongoing effort involves working with universities and U.S. Army Corps of Engineers districts to collect sediment samples to program the framework, create and maintain an interactive web interface and continue public outreach.
“It’s a program called SediNet that was developed by one of our team members, Dan Buscombe,” said McFall. “The whole SandSnap process for the user is as easy as one, two, three. First, they need to find a spot on the beach and wipe a small area with their hand. Next, set down any size U.S. coin and take a photo. Finally, upload the image with the location, cross-shore position — or shoreline position — and coin type. The image then passes through a few filters before it is analyzed for grain size. The user is then given the results with a fun beach fact while we concurrently store the results in a database.”
The project began in late 2019 but had a slow start due to the COVID-19 pandemic. While the team was developing the software and web application, they also had to figure out how to reach the public in this new environment.
“We wanted to host outreach events, but with the pandemic restrictions, we had to figure out how to do that while maintaining social distancing,” said McFall. “We came up with library science, technology, engineering and mathematics (STEM) activity kits, or beach kits.”
The team built four kits, with each kit containing four activities, including SandSnap.
“We’re testing these out in two locations in New York,” said McFall. “We finished creating them and got them into the libraries’ hands around October 2020 — which is not the perfect time, because it was the fall season, and few people are going to the beach. We expected to really test these out this summer to see what works, what the kids like, and most importantly, what’s effective and what’s not. Our long-term goal is creating about 10 of these bags and distributing them around the country at targeted locations where we want to collect data.”
As the summer months wind down, the team has set their sights on the next phase of the project.
“It’s been a unique year,” said McFall. “We’re working with a lot of contributors and colleagues for beach sample collection to train the neural network, and we’re developing the interactive web application, which is scheduled for completion this September.”
“There’s really two parts to this,” McFall continued. “There’s the technical component, and no matter how perfect that technical component is or how accurate we make it, if we don’t have a good public outreach component that engages citizen scientists, then the project fails. Likewise, if we have a great citizen scientists’ component, but the technical side is weak, then the project fails as well. Both have to be balanced and successfully executed.”