Fighting North Carolina’s coastal erosion starts with knowing where to begin.

What goes when the marsh goes

Jacob Boyd is 42 and has lived in eastern North Carolina his whole life. He runs the salt marsh program at the North Carolina Coastal Federation, which means his job is finding the places where the state’s money will do the most good holding a shoreline in place.

Like most people who live on this coast, he does not need a dataset to tell him it’s moving and changing. He has watched it with his own eyes, from higher tides and more frequent flooding on clear days to marshes and shorelines that are simply gone. What he points to first is the ghost forests, stands of dead trees killed from below by salt water pushing inland.

Behind the ghost forests, marsh moves in. In front of them, marsh turns into open water. The whole band of marsh is moving inland, and its seaward edge, which is what this project measured, is pulling back across most of the coast. Where there is room, and where the land rises gently enough, the marsh can migrate inland. Along much of this coast there is no room, because roads, houses and other infrastructure are in the way.

What is lost in that slide is not only ground. Marsh is where young fish and crabs mature. It filters the runoff coming off the land before that water reaches the bays. It takes the energy out of a storm surge before the surge reaches a road or a house. When marsh becomes open water, what is left is bottom habitat, which is still habitat, but it does almost none of that work.

Seeing all of that was never the difficult part. Measuring it was. Grant money and permits go to projects that can show evidence, and evidence means data. Everywhere along this coast needs work, which is another way of saying that, to the eye, there is no obvious place to start.

There are ways to secure a shoreline. None of them are free.

Erosion is not untreatable. A shoreline can be armored against the water using traditional methods such as bulkheads and riprap, though those have been shown to increase erosion nearby, require costly repairs, and provide no ecological benefit. Nature-based options rebuild a shoreline to absorb what is hitting it while creating habitat. The options run from breakwaters set out in the shallows to living shorelines, which combine engineered structure with natural habitat so that oysters, marsh grass and trapped sediment do part of the work and the protection grows stronger as it establishes. These methods have improved a great deal in the past decade as the materials and the modeling behind them have gotten better.

But they all come with costs. Each project is real, physical construction work in open water, paid for out of grants that can only be awarded once and spent once. The Coastal Federation cannot build along 2,500 miles of shoreline. Nobody can. So the job comes down to choosing which stretches to protect, in what order, and on what evidence.

More coast than anyone could look at

Measuring 2,500 miles of shoreline across ten years of photographs, finely enough to catch a marsh edge moving two feet in a year, is not work a person can do with a magnifier and a light table. There is simply too much of it. And the field check that would normally back it up is not available either, because much of this coast cannot be reached on foot.

Natrx built its analysis system, Natrx Assess, for exactly this. A model reads every image and decides, pixel by pixel, what is land and what is water. The rates of retreat come from the same shoreline change methods coastal scientists have used for decades, and a person overrules the model wherever the imagery is ambiguous.

The project ran in two passes, and the reason is that the coast is not a simple line. It’s a maze of creeks, bays and marsh islands. Measuring all of that precisely is slow and expensive, so the first pass was wide and quick, and the second was run only where more precision warranted the additional cost. The first pass looked at every pixel in an image and decided one thing about it: marsh, or not marsh. Do that for one year, do it again for a later one, subtract one from the other, and what is left is everywhere the marsh went away. It covered everything fast. It did not tell you how fast any one place is moving.

The second was narrow and exact. Natrx set about 93,000 fixed points along those 2,500 miles, each one on land with a line running from it out to the water’s edge. In five sets of aerial photographs spanning 2012 to 2022, the models traced that line to the edge of the marsh grass. Measure the same line five times across a decade and the shoreline’s movement at that spot becomes a rate, in feet per year.

The wide pass narrowed the coast to the stretches worth a closer look. Which of those got measured in detail was settled with the Federation, working closely with its partners.

What the first pass gives you is a ranking. It says which stretches are worth the expensive look, which is the difference between working from whatever somebody happened to observe recently and seeing the whole picture, including the parts nobody has ever visited. It is not a substitute for the measurement: lined up against the precise work, it accounts for less than four percent of why one shoreline retreats faster than another. Narrowing the search is the job, and it does that.

Resolution is the other reason this works. The aerial photographs were measured at one meter, so a single pixel covers about a square yard of ground. The Coastal Federation needed to catch places losing as little as two feet a year. Over a decade that is about twenty feet. In the publicly available satellite imagery, twenty feet is smaller than a single pixel, so the change never appears at all.

Nick Brady, Natrx’s chief technology officer, is matter of fact about which part of this was hard: having people who can look at an ambiguous photograph of a tidal flat and say whether they are looking at marsh or water.

Brady estimates that doing this the traditional way would have cost into the millions of dollars, and that even then it would not have been able to cover this much coast. A crew measuring a shoreline on foot can get through the shoreline of about one town. This project covered 2,500 miles. The Natrx analysis came to about one percent of the grant, leaving the rest for the restoration work itself.

The loss is not spread evenly

Half of those 93,000 spots are pulling back faster than eight and a half inches per year. Four in ten are losing more than a foot.

Rank every eroding spot from fastest to slowest, and the worst tenth of them accounts for this.

At the worst single spot measured, the marsh edge pulled back about 46 feet in one year. Dare County, on the sound side of the Outer Banks, is eroding fastest. Hyde County, on the mainland to the southwest, is close behind.

Ninety percent of the eroding coast in the study area is losing ground at slower rates. The other ten percent accounts for almost half of the loss. That is the most useful thing in the measurement. Spread a fixed budget evenly along 2,500 miles and most of it goes to protecting shoreline that was barely moving. Pinpoint the worst spots and the budget goes where it’s needed most and can be most effective.

Better measurements mean smaller, more focused projects

How much structure a site needs depends on several things: how much wave energy is actually reaching it, how fast the shore there is giving way, the elevation, and the water depth. Without measurements the safe assumption is the worst case, and the safe design is the largest one the permit allows.

With them, the structure gets matched to the site. Less material, fewer trucks, lower cost, and often better protection, because something built to the real conditions outperforms something built to a permit ceiling.

Building to the ceiling is the habit, and habits move slowly. As Boyd puts it, people are used to doing things the way they are used to doing them.

The analysis does two more things that have nothing to do with engineering. Grant money is competitive, and Boyd says a proposal with more robust data behind it beats one without, which is frequently the whole difference between funded and not funded. And permit reviewers want to see evidence, so bringing it up front moves an application along faster than identifying the needs after the fact, which requires more time and effort.

Six hundred acres, and a place to start

The Coastal Federation is taking the ranked shoreline into early permitting conversations with state and federal agencies. Boyd expects permit applications within about six months and construction inside two years. The program aims to protect up to 600 acres of marsh across eight large restoration projects.

The funding behind all of this was awarded for exactly those acres. In 2024 the Environmental Protection Agency awarded $421 million to the Atlantic Conservation Coalition, a partnership of North Carolina, South Carolina, Virginia, Maryland and The Nature Conservancy, and the Coastal Federation’s share is roughly $30 million over five years. The purpose of that grant is keeping carbon in the ground. Marsh soil holds it, and marsh that erodes into open water lets it go.

Interest is not confined to North Carolina. The Federation leads this state’s part of the South Atlantic Salt Marsh Initiative (SASMI), a four-state effort formed in 2021 that now brings together more than 400 partners across North Carolina, South Carolina, Georgia and Florida. Those four states hold about a million acres of salt marsh between them, and the Federation led the development of North Carolina’s plan for protecting its share. Boyd expects the results to travel. The method needs marsh, since it reads the line of grass as the shoreline, and most of the unhardened shoreline in the Southeast is exactly that.

Boyd has watched this happen his whole life, and has seen how people have had to come around to it. After Hurricane Florence in 2018, he says, a lot of his neighbors stopped arguing about whether the coast was changing and started asking what could be done about it.

The datasets and the mapping tools go public on the Federation’s website this fall. Anyone will be able to look up their own stretch of shoreline.