Research Update: Gauging Carbon Levels in Salt Marshes

Salt marshes store roughly 25% of the world’s carbon, despite only making up about 5% of the world’s surface. Beyond that, they are an important form of natural infrastructure, mitigating the impacts from storms and reducing flooding in nearby towns and cities.

However, salt marshes face serious challenges due to development and sea level rise. Sea level rise can drown marshes, causing the release of carbon in the atmosphere and preventing future carbon uptake.

IRIS affiliate Dr. S. Sonny Kim, Associate Professor in the College of Engineering, is working to find low-cost, reliables method of measuring carbon in salt marshes to better understand how factors like depth, salinity, pH and temperature impact organic carbon levels in soil. By understanding these factors, researchers can develop ways to protect marshes as conditions change, which will ultimately buffer against climate change and help communities and livelihoods.

“We’ve already lost 53% of the wetlands in North America, 62% in Mexico and 16% in Canada due to extreme weather events and climate change,” Kim explained. “Because of this loss, it’s really important that we have continuous monitoring of organic carbon.”

However, monitoring organic carbon in saltwater marshes has proven to be a challenge: to sample in the traditional way, using cores of soil pulled up from within the marsh, researchers must face alligators, venomous snakes, and unknown water depths.

“One time, I took one of my students out to measure the soil properties of a site. Suddenly I looked up and realized that he was stuck in the muck at the bottom of wetland. I had to toss big pieces of wood to him to stand on. After getting back from the trip, he kept joking that I’d rescued him,” Kim recalled.

By comparison, remote sensing technologies are efficient and hassle free, but to be able to use them with confidence, researchers must ensure that they are properly calibrated and providing quality data.

“Machine learning and artificial intelligence are emerging fields, and sensors are becoming cheaper and more accurate,” Kim continued, “So we thought, why not combine that with remote sensing with satellite image data, so that we can minimize inconvenience in data collection, damage to the wetlands and also have access to data in near real-time?”

This effort called for an interdisciplinary team, including researchers from Computer Science, Geography, and the Warnell School of Forestry and Natural Resources.

It’s been a year since the project was first approved, and since then the researchers have begun to develop a machine learning (ML) model, affectionately dubbed AWeSOMSense. This model will help to accurately predict below-ground soil organic matter based on surface features, such as reflectance which can help us understand patterns and trends in marsh response to sea level rise and climate change.

“This model will transform salt marsh soil organic matter sensing by using high-quality data at a geographically dense spatial pattern,” Kim explained. “The spatio-temporal dynamics will expand the footprint of the existing point data to broader scales and will provide variation of the predicted belowground soil organic matter along depth based on machine learning algorithms.”

They’ve also individually tested low-cost sensors that measure pH and salinity using solar power to assess accuracy. Expensive sensors can run upwards of $160, but the team has identified much less expensive sensors (to the tune of $16) that, when properly calibrated and filtered, provide solid data.

The next step is to test the sensors in the field by combining them into a tower, which they’ll install in a wetland.

“We’ll leave it there for about two weeks, so that it can continually measure soil properties, regardless of high or low tide or other changing conditions,” Kim explained. “We are very much looking forward to seeing the data variation from the field.”

Kim and colleagues are especially proud of this research for helping to fulfill the University of Georgia’s mission as a Land and Sea Grant university.

“We feel that we’re actually satisfying UGA’s mission to protect the land and sea within Georgia by laying the groundwork that will allow us to prevent marshes from drowning, sustain storm protection, and support the livelihoods of people in surrounding communities, which is very exciting to us,” Kim said. Read the original article here.

Author: Sarah Buckleitner
Contact: Dr. Sonny Kim