Footage Shows How Dangerous California's Coastline Cliffs Are

Apparently, the cliffside right here, this close to PCH, is falling. Right now, on a stretch of coastline in San Clemente, California, waves are washing directly over an active railroad. Not near it, over it. This is part of the LO-SSAN corridor, the second busiest passenger rail line in the country, carrying Amtrak, Metrolink, and freight up and down the coast. This September, the ocean shut part of it down. High surf climbed the slope beside the tracks, damaged rock that had already been placed to protect it, and forced officials to close the line indefinitely. A railroad running out of ground. The San Clemente closure didn't come out of nowhere. Orange County Transportation Authority officials say high surf from Hurricane Marie hit the slope over Labor Day weekend, and when crews went back days later, they found the damage was worse than first thought. Waves had overtopped the tracks and torn into protective rock that had been placed there specifically to hold the slope together. That's a strong signal, because rock armor is usually the last line of defense before a rail line becomes unusable. San Clemente itself is a small city, around 62,000 people, built along a narrow shelf between hillside and shoreline. For decades, the beaches here were fed by sediment carried down from inland rivers and creeks. Urban development upstream slowed that supply to a trickle. At the same time, warmer ocean water is pushing waves higher and further inland than they used to reach. Less sand arriving, bigger waves arriving anyway. That combination is why this particular town is losing ground faster than most. The scale of the shortfall is striking. City officials estimate San Clemente needs about 3 million cubic yards of sand just to restore what's been lost. Trucking that much sand in would take roughly 300,000 individual truck trips. That's not a maintenance project, that's closer to rebuilding a beach from scratch, one load at a time. With the sand fix that far out of reach, officials are now weighing harder options. One is relocating the tracks further inland, a move that's been called too expensive for years but is being reconsidered now that the closures keep repeating. Another is simply accepting more closures as a recurring cost of doing business on this stretch of coast. A third, more drastic option under discussion is halting rail service through this segment altogether, Rerouting freight and passengers around the problem in Why the coast was always going to give way Roughly 70% of the California coastline is made up of eroding cliffs, not gently sloping beach. San Clemente sits on one edge of a much larger, older problem that stretches from the Mexican border up past San Francisco. These aren't hard granite walls built to withstand centuries of pounding surf. Many are sedimentary bluffs, layers of compacted sand, silt, and clay, stacked up over thousands of years. And that composition matters enormously for how they fail. Waves attack the base of these cliffs first, carving out what geologists call a wave-cut notch, a hollow groove at the point where ocean meets rock. Rainfall does damage from the opposite direction, soaking into the bluff from above and softening the material that's supposed to be holding the notch's roof up. Push from below, wait from above. Eventually, the unsupported section simply can't hold itself together, and it drops, sometimes as a slow slump, sometimes all at once. Taking whatever sits on top of it along for the fall. That's the key detail people miss when they look at official erosion statistics. Statewide, California's cliffs retreat by an average of about six centimeters, roughly two inches, per year. That sounds almost harmless, something you'd barely notice in a lifetime, but averages hide the real hazard. In places like Rancho Palos Verdes and Big Sur, retreat has been measured at more than 16 feet in a single year. Over a hundred times the statewide average. Even farther north, stretches of Humboldt and Mendocino County cliffs show similarly extreme rates, cut into softer, more exposed rock with almost no shelter from open ocean swell. The danger was never the slow, steady inches. It's the sudden, violent collapse that happens without warning, when a section of cliff that looked stable for years gives way in seconds, dropping tons of rock and soil onto whatever beach, Trail or track happens to sit below. That unpredictability is exactly what turned a scientific curiosity into an urgent public safety question. Because in one California community, it turned fatal. The collapse that changed everything. On August 2nd, 2019, at Grand View Beach in Encinitas, a slab of bluff, roughly 30 feet wide and 25 feet tall, broke away without warning and came down onto the sand below. Three women died in that collapse. There was no storm that day, no visible crack anyone had flagged. The bluff simply gave way on a stretch of coast people walked past every day without a second thought. That tragedy is why California started treating cliff monitoring as a public safety issue. Not just infrastructure planning. Lawmakers responded with AB66, funding a three-year, $2.5 million research effort built around one question. Could anyone have seen this coming? The plan combined LiDAR mapping, using laser pulses to build precise 3D models of the cliff face, with a longer-term push to install physical sensors directly into the rock at known danger points. Nobody thought the goal was stopping erosion. The goal was warning time. Even a little. Enough to close a trail before the next Grand View Beach happened. That funding is what eventually put real instruments into the ground. And this year, those instruments proved they work. Listening to the Cliffs, inside the 2026 Early Warning Study. In July, Scripps Institution of Oceanography released results from a four-year monitoring effort at three San Diego County sites, Del Mar, Beacons Beach, and San Elijo State Beach. The finding was direct. A network of ground sensors can give genuine warning before a dangerous cliff failure, sometimes hours ahead, sometimes days. The equipment explains how something as chaotic as a rock collapse becomes predictable at all. Researchers installed tiltmeters, which track the cliff tilting almost imperceptibly before it breaks, alongside mechanical and optical extensometers, measuring tiny amounts of stretching in the rock, a seismometer to pick up vibration, and optical displacement sensors tracking movement down to fractions of a millimetre. Drones and truck-mounted LiDAR filled in the picture from above, remapping the cliff face repeatedly to catch changes invisible to the naked eye. Some of that equipment is almost absurdly sensitive. The tiltmeters can register ground tilting equal to one-eighth the width of a human hair, and fiber-optic sensors in the array can detect deformation down to one-billionth of a meter. That's not measuring cracks, that's measuring rock beginning to move before there's a crack at all. Over the four years, the team documented five separate collapses, two at Del Mar and three at San Elijo, ranging from a few cubic yards of material to several hundred. In every case, sensors picked up ground movement before the rock came down. The clearest example happened at Del Mar in April 2024. After heavy rain soaked the bluff, tiltmeter readings started accelerating day by day. And by the time the trend was unmistakable, researchers knew a failure was coming, before it happened, not after. That's the difference between a monitoring system It's worth being honest about where this technology still falls short. At Beacons Beach, a site with a long history of slow-moving landslide activity, permit delays meant the most sensitive sensors weren't collecting usable data until 2025. And researchers there concluded they need more time watching that slope before drawing firm conclusions. A planned second phase, adding fiber-optic strain meters at more locations, is still Still waiting on permits of its own. And a scientist involved in the broader USGS work on this coastline put the scaling problem plainly. Crumbling cliffs stretch across more than 530 miles of California, and building this kind of sensor network out to cover all of it would cost a serious amount of money. This is the kind of story we keep building on at DiscoverTube. Because the science doesn't stop at one press release, it keeps developing site by site. So the question shifts from whether early warning works, which it clearly does at these three beaches, to whether anyone can afford to listen to the whole coastline at once. Mapping a moving coastline. That scaling question has a partial answer already, and it arrived within weeks of the sensor results. Scripps released a new interactive tool called the California Coastal Cliff Erosion Viewer, letting anyone pull up almost any stretch of coast and see documented erosion rates for that specific spot. It covers 866 kilometers, about 538 miles, of mapped cliff built from repeated airborne LiDAR passes stitched into a single searchable record. What the map shows confirms something researchers had long suspected, but never had laid out this clearly. Erosion isn't evenly distributed, and the North is losing ground faster than almost anywhere else in the state. Humboldt, Mendocino, and Del Norte counties post some of the highest retreat rates recorded, driven by softer rock and direct exposure to open ocean swell with none of the sheltering that Southern California bays and headlands provide. Point Arguello and the bluffs near Big Sur and Martin Beach show similarly extreme numbers further south. Proof this isn't purely a north-south divide, but a matter of which rock faces the open Pacific unprotected. Orange County, by contrast, shows some of the lowest rates on the map, in some stretches barely a fraction of an inch a year. Part of that gap comes down to armoring. About 20% of Southern California's cliffs have been reinforced with riprap or seawalls, Compared to just 1.5% up north, And armored sections erode at roughly half the rate of exposed rock nearby. That protection comes with its own cost, though. Where a seawall, or rock revetment, holds a cliff in place, the beach in front of it often can't retreat landward the way it naturally would, and it simply gets squeezed narrower with every high tide. Scripps isn't stopping at this single snapshot, either. Their California mapping program plans to resurvey the entire coastline this fall, and again next spring. Specifically to measure what this year's ocean conditions do to the cliffs in the meantime, giving researchers a before and after record of exactly how much damage one severe season can cause. The season that pushed the sea higher. That before and after record matters more this year than any other, because 2026 is shaping up to be one of the most punishing coastal seasons on record. Climate scientists tracking the Pacific describe conditions right now as a Super El Nino, layered on top of an unusually active hurricane season. Those two forces together are doing something specific to the ocean's surface. Warm water expands. Storms pile that warm water higher against the coast, And the result is sea level running about 6 inches above normal since March, a level expected to peak this coming November and December. Six inches doesn't sound dramatic on its own, but remember what's happening underneath it. Those cliffs are already being undercut at the base by ordinary wave action. Add six extra inches of water reaching that same base, and waves that used to fall short now slam directly into the weak point every single tide cycle. That's the same physical process eating into San Clemente's rail line, just playing out up and down hundreds of miles of coast at once. Local governments are already responding as if this is an emergency, because in several places, it is one. Carlsbad, just north of San Diego, declared a local emergency in September and rushed in a 192-foot rock revetment to protect an exposed bluff before winter surf arrives in force. Officials covering the stretch from Oceanside down to Del Mar have warned residents that unprotected properties could see the most damaging erosion season in decades. And they're blunt that some homes have rip-rap defences, while plenty of others have nothing at all. It helps to hold two numbers side by side here. This single season is pushing sea level 6 inches above normal for a few months. But scientists studying the longer curve say bluff tops across Southern California Could lose 62 to 135 feet of ground by the end of the century in the worst affected stretches. 135 feet is close to a third of a football field. One number is this winter's spike. The other is where that spike is dragging the baseline. Short-term armor like Carlsbad's revetment can blunt a single bad season. It was never designed to answer the second number. Holding the line or learning to let go. That's the choice quietly sitting underneath every emergency declaration and every new sensor reading. Do you keep defending a fixed line against a rising sea, or do you accept the coast is going to move and plan around that instead? Sea walls and rock revetments work in the narrow sense that they stop a specific bluff from retreating for a while, but they create a side effect engineers call Coastal Squeeze. The rock holds the cliff in place, the sea keeps rising anyway, And the beach in front of that wall has nowhere left to retreat to. It simply narrows, tide after tide, until in some spots it disappears at high water entirely. Del Mar has been negotiating this exact trade-off with regional transportation planners for years, discussing whether to physically move rail tracks off the most unstable stretch of bluff rather than keep armoring it in place. Nothing has been decided. A USGS oceanographer working on this coastline Put the core problem plainly, sea level rise is the constant pressure, it isn't going to pause, and it's going to keep pushing beaches landward whether or not any given town has the money to fight it. Worth noting, the Pacific Northwest gets a partial break here that California doesn't. North of Cape Mendocino, tectonic plate movement is actually lifting the land itself, roughly Roughly half an inch to just over an inch per decade. Which offsets some of the rise other stretches of the coast feel at full force. Southern California has no such cushion. So here's a genuine question worth sitting with, and it's one we'd like to hear your take on in the comments over on DiscoverTube. If armoring one town's bluff just pushes the erosion problem down the coast to the next unprotected stretch, is engineered defense even the right long-term answer? Or is retreat the only honest option? San Clemente is where all of this collapses back into one concrete decision. Move the tracks inland, accept repeated closures as routine, or shut the corridor down. Whichever way officials lean, the ground itself has already made its position clear. What's changed in 2026 isn't that California's coast is suddenly falling apart faster than before. What's changed is that we can now watch it happen, sometimes days ahead of the collapse Thanks to sensors sensitive enough to catch a cliff tilting a fraction of a hair's width before it breaks. Measurement finally caught up to the mechanism. The fix hasn't. If you want to keep following where this goes, next season's re-survey, the next closure, the next emergency declaration, that's exactly the kind of unfolding story DiscoverTube was built to track. So subscribe if you want to stay with it, and tell us in the comments. Would you rather see California keep armoring its coastline or start pulling back from it?