What the Coast Is Telling Us

Hatteras Is Losing Houses to the Ocean. Louisiana Is Losing a Football Field of Wetlands Every 100 Minutes. Neither Problem Will Be Solved by Any Single Profession. That Is the Design Brief for the Next Decade.

Thirty-One Houses

As of February 2026, thirty-one unoccupied oceanfront houses on Hatteras Island have collapsed into the Atlantic. Nine of them fell in a single week in October 2025. The ones that are left are being watched. Some will go this year. Some will go next year. All of them, eventually, will go.

The shoreline at Buxton has been retreating at roughly seventeen feet per year for two centuries. The shoreline at Rodanthe has been retreating at ten to fifteen feet per year for decades. Both numbers are being revised upward. Drone and lidar mapping after the October storms found that one week of back-to-back weather had pulled the shoreline back by an average of sixty-eight feet, with some stretches losing up to a hundred and forty-one feet of beach. Sand was pushed inland as far as two hundred and ninety-eight feet.

North Carolina has one of the strictest bans on hardened shoreline structures in the country, adopted in the 1980s. Seawalls and revetments are, with narrow exceptions, not allowed. The state is currently debating whether to repeal that ban because sixty years of beach nourishment, groins, sandbags, and terminal groins have not stopped the retreat. Nourishment projects are not lasting their design cycles. The state’s Division of Coastal Management recently told the Coastal Resources Commission that the framework is no longer keeping pace with conditions. The NCDEQ representative put it plainly at the November 2025 meeting. Inaction is the enemy.

Stanley Riggs, one of the coast’s most credible voices for decades, has proposed a different approach. Stop defending the shoreline. Establish a buffer zone, the Buxton Beach Historic Park, that belongs to the ocean and its dynamics. Relocate remaining houses. Clean up the sandbag graveyards, the septic tanks, the concrete swimming pools left behind by structures already lost. Accept that the island is doing what barrier islands have always done. Work with that motion, not against it.

This is what coastal landscape architecture looks like now. Not placemaking. Not amenity design. Triage.

Two Thousand Square Miles

Nine hundred miles south, Louisiana has a different version of the same problem. Not resistance failing. Control succeeding too well.

Since the 1930s, Louisiana has lost roughly two thousand square miles of land. That is an area the size of Delaware, gone. The loss rate right now is approximately one football field of wetlands every one hundred minutes. Terrebonne and Lafourche Parishes, with a combined population of over two hundred thousand, are projected to be underwater within fifty to eighty years at current rates. Two parishes. Quarter of a million people. Inside a human lifetime.

The cause is legible. The Mississippi River built southeast Louisiana over seven thousand years by changing its course every thousand to fifteen hundred years, laying down delta lobes of sediment. That was how the coast came to exist. In 1928, after the Great Flood of 1927, Congress authorized the Army Corps of Engineers to shore up the levee system along the entire lower Mississippi. By the 1950s, LSU geology professor James Morgan was already documenting the land loss that the levees had set in motion. The sediment that had built Louisiana for seven thousand years was now being delivered straight into the Gulf of Mexico, past the delta, into deep water where it does no one any good. The coast was being starved.

The levees did what they were asked to do. They protected New Orleans. They protected the navigation channel that carries eighteen percent of United States waterborne commerce. They kept the Port of South Louisiana open. By any engineering standard applied to any single project, the levees were a success. By any landscape or ecological standard, they are one of the largest and slowest-motion disasters in American environmental history.

The state knows this. The 2017 Louisiana Coastal Master Plan identified eight proposed sediment diversions along the Mississippi as a central strategy for rebuilding land. The Mid-Barataria Sediment Diversion, a two-point-two-billion-dollar project authorized by the Army Corps in late 2022, would have been the first of its kind. It was designed to mimic what the river did for seven thousand years, diverting muddy water into Barataria Bay during peak flood season to rebuild marsh. Projected to create or save as much as forty-seven square miles of land over fifty years.

In 2024 the permit was suspended. As of mid-2025, the project’s future is uncertain. Opposition came from the fisheries communities downstream, who would bear the costs of a sudden change in salinity. The oysters would take the hit. The shrimp would take the hit. Real people whose livelihoods depend on the salinity levels the coastal engineering of the last century accidentally produced. The diversion’s intended beneficiaries and its documented costs exist in the same waters.

This is the pattern. The coast is a complex system. Every intervention produces second-order consequences that were invisible to the discipline authoring the first-order solution. The levees saved New Orleans and starved the delta. The sediment diversion would rebuild the delta and injure the fisheries. Someone is always holding the bag.

No landscape architect, no engineer, no architect, no coastal geomorphologist, no fisheries biologist, no hydrologist solves this alone. The problem is larger than any single discipline’s frame of reference.

What the Work Actually Is

Coastal mitigation in 2031 is not a design project. It is a prediction problem that ends in design decisions.

Consider what the Hatteras question actually requires. Where will the shoreline be in five years. In ten. In thirty. Not a single line on a map. A probability distribution of positions, weighted by storm sequence, by the actual rate of relative sea level rise, by the specific geology of each barrier segment, by the state of the inlet dynamics at Oregon Inlet and Hatteras Inlet, by the sediment supply from offshore shoals, by the frequency of overwash events that widen the island westward as they narrow it from the east. No one answers that question with a drawing. It is answered with a model that integrates hydrology, meteorology, geology, and probability theory, run under a range of climate scenarios, tested against the actual record of the last twenty years of storms.

Once the model answers, the design question begins. Which stretches of the island are defensible and on what time horizon. Which are not. Which structures can be retained. Which should be relocated now. Which should be relocated in the next cycle. Where does NC Highway 12 go when the current alignment is no longer maintainable. What happens to the communities whose entire economy is built on beach access. What does a barrier island village look like when it has been designed to move with the island rather than anchor against it.

The Louisiana question is parallel. The sediment diversion is not one project. It is a choice between several versions of Louisiana in 2075. The version where the diversion runs at full authorized capacity. The version where it runs at reduced capacity to minimize fisheries disruption. The version where it does not run at all and the coast continues to retreat at current rates. The version where it runs alongside a fisheries transition program that pays oyster communities to relocate or to shift to species that tolerate the new salinity. Every version has a landscape and an economy and a set of communities. Every version has winners and losers. The question is which version gets built.

No single profession can hold this question. Landscape architecture cannot. Engineering cannot. Architecture cannot. Coastal geomorphology cannot. The problem sits in the intersection of all of them, and the intersection is where the design work has to happen.

The Shared Model

Here is where the tools come in. The reason this kind of integrated, cross-disciplinary design work has not been done at scale before is not that nobody thought of it. It is that the translation cost between disciplines ate every project that tried.

A landscape architect and a coastal engineer have, historically, not been able to work from the same data. The engineer has a hydrological model in one software package. The landscape architect has a site plan in another software package. The architect has a building model in a third. The geomorphologist has a shoreline migration projection in a fourth. To get them talking to each other requires a human being to translate between the models, which takes weeks, which costs money, which means the translation only happens at project milestones, which means the disciplines are never actually thinking together. They are passing finished artifacts across a wall.

AI did not change what each discipline knows. It changed what it costs to combine what they know.

In 2031, the prediction center for a project at Hatteras or in Barataria Bay is a room where the landscape architect, the coastal engineer, the architect, the ecologist, and the municipal planner are all looking at the same integrated model. They are not looking at different dashboards of the same data. They are asking the model questions together.

What happens to the dune if we retreat the road a hundred feet west. What does that do to the salt spray exposure on the structures behind it. What does that do to the habitat corridor. What does that do to the estimated cost of maintenance over thirty years. What happens to the fisheries downstream if the diversion runs at seventy-five thousand cubic feet per second in peak flood. What if it runs at fifty thousand. What if it runs at fifty thousand only during March and April and the rest of the year the gates are closed.

The model runs in minutes. The disciplines argue in real time. The design decision that comes out of that conversation is informed by more integrated knowledge than any design decision in the history of the profession has been informed by.

This is not a description of what might be possible. Tools like this already exist in specialized forms. The MissDelta research consortium funded in 2023 by the National Academies, led by Tulane and LSU, has been building exactly this kind of integrated coastal model. The RCOAST startup that mapped the October 2025 Buxton collapses in three dimensions is doing a version of this work. The tools are becoming available to firms, not just research universities, on a timeline that makes 2031 a reasonable horizon for widespread adoption.

The question is not whether the tools will exist. The question is whether the disciplines will have learned how to work with them, and with each other, by the time the tools arrive.

The old hierarchy, which was architect on top, engineer in the middle, landscape architect managing the ground plane underneath, is not being overthrown by AI. It is being dissolved by the nature of the problems the professions are now being asked to solve. The coast does not care about the hierarchy. The coast cares about whether the people working on it can actually think together.

What This Means for Design Education

Here is where the post turns.

NC State’s College of Design is roughly three hours inland from Hatteras. It trains landscape architects, architects, and graphic designers. It is a good program. It has produced generations of capable professionals. I went there. I got my Master of Landscape Architecture and Environmental Planning there in 2022. I am grateful to the faculty who taught me and the peers I learned alongside.

It is also, like most American design programs, not yet built for what the coast is about to demand.

The studio model of design education is largely unchanged from its mid-twentieth-century form. A student sits at a desk, receives a site, develops a design, is critiqued on the design. The tools evolve. The underlying pedagogy does not. The separation between landscape architecture, architecture, and engineering is structural, baked into the accreditation system, reinforced by the professional licensure regimes that follow graduation. Students who cross the boundary do so on their own initiative, against the grain of their program.

This is not going to be sufficient. The coast is going to need graduates who can read a hydrological model, argue with a coastal engineer, understand the statistical mechanics of a storm probability distribution, collaborate with an ecologist on a salinity gradient, and then translate the result into a design intervention that communities can evaluate. That is not a landscape architect. It is not an engineer. It is not an architect. It is something the current accreditation system does not have a category for, because the current accreditation system was built for a coast that was not falling into the ocean and a river that was not starving a delta.

The curriculum will have to change. Not the tools taught. The curriculum. Studios will need to be cross-disciplinary by default. Students will need to be fluent in data analysis at a level that current programs treat as optional. Climate science will need to be a core sequence, not an elective. The statistical reasoning required to work with probability-weighted shoreline projections will need to be taught, not assumed. The difference between a design that is generated and a design that is defended, the difference between a rendering that is persuasive and a rendering that is honest, will need to be a pedagogical focus, not a stray remark in a jury.

Design education has been producing graduates for a profession that is about to cease to exist in the form the programs were built to serve. The programs that see this coming and restructure will be producing the professionals the coast will hire in 2031. The programs that do not will be producing graduates whose credentials were accurate for a problem that is not the problem anymore.

NC State has the opportunity to lead on this. The geography is right. The state’s coast is the case study. The cross-disciplinary infrastructure of the university is real. The demand signal from Dare County and from Louisiana and from every coastal region facing the same questions is as clear as a demand signal can be. The question is whether the institutional incentives will allow the curriculum to move at the speed the problem is moving. If they do not, a different program will lead, and NC State’s graduates will be good at the things the profession used to need.

This Comes Back to You

If you are a landscape architect, the five-million-dollar stormwater project on your desk right now is practice for the twenty-million-dollar coastal mitigation project that is about to arrive. The way you set up that project, the data you demand, the disciplines you pull into the conversation, the software you invest in learning, is practice. The firms that treat present work as practice for future work will be ready. The firms that treat present work as the ceiling will find themselves doing the same projects five years from now while the next generation of work goes to firms they have never heard of.

If you are an engineer, the same is true with the polarity reversed. The hydrological and coastal engineering firms that see themselves as the senior partner in a coastal project, the ones that understand landscape architecture as the profession that handles the surface treatment after the real work is done, are about to lose their entire business model. The coast is a design problem. The engineering is necessary. It is not sufficient.

If you are an architect, your problem is the hardest. Coastal architecture in the sense that most architects understand coastal architecture is about to not be a category anymore.

What is coming is regional planning disguised as building design. You will not be designing houses that sit on a shoreline. You will be negotiating which houses stay, which move, and what the moved ones look like when they arrive in their new location.

That is a different profession. Some architects will become it. Most will be replaced by the ones who already did.

If you are a student in any of these disciplines, or in climate science, or in geography, or in environmental engineering, you are sitting in the most interesting position in the room. The credentials you are accumulating right now are going to be evaluated in 2031 not by how well they match the old profession but by how well they prepare you for the actual problem. Build the foundation underneath the tools. Learn to read a model. Learn to argue across disciplinary boundaries without losing your standing in your own. The professionals who come out of this decade with those capacities will run the next fifty years of coastal work.

If you are a client, a municipality, a state agency, a federal agency, you are about to have to make procurement decisions that the existing firm categories do not neatly map to. Ask who is actually at the table. Ask whether the team is a team or a set of subcontractors. Ask whether the disciplines are arguing with each other or taking turns presenting. The projects that get built well in 2031 will be bought by clients who learned how to evaluate integration, not by clients who bought the firms with the best brochures.

The Terrain Lens — Applied

The coast is not waiting for the professions to get their act together. The shoreline moves at the speed the shoreline moves. The wetlands subside at the speed they subside. The storms come when the storms come.

Pick one coastal place you know. A beach you grew up going to. A city you have worked in. A marsh you have walked. Go find the most recent erosion data, subsidence data, or shoreline retreat data for that place. It will be available. Look at the number.

Now ask what you would design for that place if you were given the brief today. Not a rendering. A strategy. What do you defend. What do you relocate. What do you accept. Which disciplines do you need in the room. What data do you need to make the call.

What is your time horizon.

If you can answer those questions, the rest of this post is a framework for work you are already prepared to do. If you cannot, you are in the same position as most of the profession, including the one that trained you, which is starting from behind on a clock that is not asking anyone’s permission to keep moving.

The coast is telling us something. It has been telling us for a hundred years. The difference now is that the tools to hear it clearly, to integrate what it is saying across every discipline that needs to respond, are finally arriving. The question is whether the disciplines arrive with them.