USGS Monitors Huge Landslides on California's Big Sur Coast, Shares Information with California Department of Transportation

By Coastal and Marine Hazards and Resources Program November 1, 2017

On May 20, 2017, more than 2 million cubic meters of rock and dirt—enough to fill a line of dump trucks nearly a thousand miles long—collapsed down the steep slopes at Mud Creek on California’s Big Sur coast, about 140 miles south of San Francisco. A pile of rubble almost a third of a mile wide buried California State Highway 1 over 65 feet deep and added about 13 acres of new la

This article is part of the August-October 2017 issue of the Sound Waves newsletter.

The catastrophic collapse followed a prolonged period of slower movement, which was being monitored by USGS scientists in cooperation with the California Department of Transportation (Caltrans). Several times since March 2017, for example, USGS scientists had collected air photos along this stretch of coast. That team, led by research geologist Jon Warrick, uses “structure-from-motion” software to transform air photos into 3D maps, or digital elevation models, that they use to precisely measure changes in ground elevation.

USGS air photo of the Mud Creek landslide, taken on May 27, 2017. Photo credit: USGS

Warrick thought the team had an exciting story to tell when photos shot on May 19 documented a moderate-sized landslide blocking the highway at Mud Creek. As the scientists were analyzing the May 19 photos, however, they learned of the much larger slide that occurred the next day, May 20.

“We were utterly impressed by the landslide we saw on May 19, so when the entire mountainside failed on May 20, we were stunned,” said Warrick.

On May 27, Warrick’s team collected and processed a fresh set of air photos. Comparing 3D maps derived from the May 27 photos with those from May 19 enabled them to calculate changes in ground elevation and measure the area, thickness, and volume of the slide.

3D maps created by computer processing of air photos show what the Mud Creek area looked like before and after the landslide — Topographic “point clouds” (or 3D maps) created by computer processing of air photos show what the Mud Creek area looked like on March 8, 2017 (top), May 19 (center), and May 27 (bottom). Image credit: USGS

3 close-up views of the toe of the Mud Creek landslide created from air photos after the landslide — Close-up views of the toe of the Mud Creek landslide created from air photos taken May 27 (top), June 13 (middle), and June 26 (bottom). Caltrans built roads on top of the slide for assessment and monitoring activities. Image credit: USGS

“These photos not only show us what the ground looks like, but they also allow us to make precise measurements of the landscape that can inform resource managers and field crews,” said Andy Ritchie, a USGS scientist who has been instrumental in developing analysis techniques for the project.

The team collected more photos on May 31, taking advantage of another clear day on a coast that is often foggy during late spring and early summer. Processing these photos revealed that parts of the upper slopes were still moving, and waves had begun to erode the edge of the lobe of sediment extending into the sea. Photos shot on June 13 and 26 showed still more movement and erosion: between May 27 and June 26, the seaward edge of the slide had retreated about 30 feet.

Topographic-change image produced by comparing 3D map derived from May 27 air photos with that derived from June 13 air photos. Note that material has eroded from the seaward edge of the toe (yellow and orange tones), and material has built up on beaches to either side of the slide (blue tones). Image credit: USGS

Gerry Hatcher with the system he designed to record the precise time and geographic location of each air photo Van Wagenen shoots. Photo credit: USGS

The USGS scientists continue to monitor the slide by collecting imagery every couple of weeks, weather permitting. Pilot Bob Van Wagenen, contracted through the Department of the Interior’s Office of Aviation Services, takes air photos for Warrick’s team, flying out of the Watsonville Municipal Airport in a Cessna 182R. He uses a camera-plus-GPS system designed by USGS ocean engineer Gerry Hatcher to record the precise time and geographic location of each photo—information that speeds processing and increases accuracy of the maps derived from the photos.

Two photographs taken from the sky looking down on a landslide into the ocean water with a plume of murky water along the coast. — Drone’s-eye views of the toe of the Mud Creek landslide, from videos shot by Shawn Harrison on July 12, 2017. Credit: Shawn Harrison, USGS

Preliminary seafloor bathymetry (in rainbow colors) collected by the USGS research vessel Parke Snavely on July 11, 2017. Relative depths shown in color, superimposed on a shaded-relief map from the June 26 USGS air-photo survey. Note white data gap next to the shore where water was too shallow for the Snavely to map. Shawn Harrison hopes to fill this gap with data derived from drone videos. Credit: Andy Ritchie, USGS

Several of the project scientists have been certified to fly government drones (see “Coastal and Marine Geology is Airborne!”). On July 12, USGS Mendenhall Research Fellow Shawn Harrison used a drone with attached video camera to image the slide area in even greater detail.

Harrison timed his drone experiment to closely follow offshore mapping on July 11 by Ritchie and other scientists from the USGS Pacific Coastal and Marine Science Center (PCMSC). They used sonar mounted on the USGS research vessel Parke Snavely to map the underwater slide debris and surrounding seafloor. The research vessel could not get right up to the shoreline, which is where Harrison’s drone work comes in: if successful, his analysis of images from the drone videos will enable Harrison to produce a bathymetric map of the bottom between the shore and the water depths that the Snavely can map.

Eventually, Warrick plans to deploy instruments on the seafloor to measure currents and waves. All these offshore studies will help the researchers determine what happens to landslide debris that ends up beneath the water. How does it affect the seabed? Will it be eroded away and dispersed? Where will it go?

“We have a special opportunity to track how the land and ocean respond to these rare but large landslide events,” stated Warrick. “We know that there has been a long history of landsliding along the U.S. west coast, and this recent slide will allow us to better understand how the coast changes over time.”

Working with landslide researchers

Big Sur coast. Red squares mark some of the sites damaged by 2016–17 winter storms, including Mud Creek and Paul’s Slide areas. Basemap is the intellectual property of Esri and is used herein under license. Copyright © 2017 Esri and its licensors. All rights reserved. Image credit: USGS

The PCMSC scientists are collaborating with research geologist and landslide expert Kevin Schmidt of the USGS Geology, Minerals, Energy, and Geophysics Science Center (GMEG). Schmidt has been studying several landslides on the Big Sur coast, including Paul’s Slide, which is about 13 miles up the coast from Mud Creek and was still limiting access along Highway 1 in mid-July.

“Kevin really is the reason we started photographing Big Sur,” said Warrick. About a year and a half ago, Warrick told Schmidt about regular flights he was conducting to identify and measure coastal change from San Francisco to Carmel as part of the USGS Remote Sensing Coastal-Change project.

“After we talked,” recounts Warrick, “Kevin said, ‘Gosh, you should fly Big Sur so we can compare your structure-from-motion with our GPS [Global Positioning System] surveys.’ He convinced me.” Soon Warrick’s team was photographing most of the Big Sur coast. They extended their flights farther south in March, enabling them to capture the May 20 landslide at Mud Creek.

As part of the collaboration between Warrick and Schmidt, GMEG geologist Beth Haddon, who recently began working with Schmidt, will apply structure-from-motion processing to Warrick’s air photos of Paul’s Slide. Information gained from the photos will be added to Schmidt’s on-the-ground studies, including GPS surveys over wet and dry winters that Schmidt has been conducting for years.

Schmidt and his team “are the landslide experts,” said Warrick. “They know how the earth moves. We are providing remote-sensing data and expertise.”

Air photo of Paul's slide (left); photo of repair equipment/construction (right) — Paul’s Slide. Left: Air photo taken May 27, 2017, one of many that Beth Haddon will analyze with structure-from-motion software to measure changes in ground elevation. Right: Caltrans employees and contractors use an excavator and bulldozer to remove Paul’s Slide debris from California State Highway 1. Note recent construction of concrete and a line of grouted reinforcing bars (“nails”) to slow undercutting of southbound lane. Photo taken June 6, 2017, by Kevin Schmidt.

Collaborating with Caltrans

Schmidt and Warrick are sharing their findings with Caltrans to help the agency’s geologists and engineers assess and monitor the slides.

“We have been utilizing data and information from the USGS for landslide modeling and risk management at the Mud Creek landslide complex,” said Tom Whitman, senior engineering geologist with Caltrans. “Given the size and magnitude of the landslide complex and its remote location, the information from the USGS has given us the opportunity to evaluate the landslide complex through its development and post failure in a way that no other data could. I look forward to additional reports over the next year.”

Collaboration between Caltrans and the USGS along the Big Sur coast goes back years. In the early 2000s, for example, then PCMSC research geologist Cheryl Hapke (now at the USGS St. Petersburg Coastal and Marine Science Center in Florida) conducted studies to determine the rates at which landslides were moving material from the Big Sur mountains into the waters of the adjacent Monterey Bay National Marine Sanctuary (see “New USGS Fact Sheet About Landslides Delivering Slope Material to Nearshore Waters on California’s Big Sur Coast”). She used digital photogrammetric techniques to create 3D maps and derive historical landslide volumes from air photos. The technique she used is like the structure-from-motion approach, but what takes days with the new technology took many months back in the 2000s. Hapke also analyzed air photos and hyperspectral data (measurements of reflected sunlight) to map locations of historical landslides along the length of the Big Sur coast and correlate their occurrence with rock types and deformation, such as faults and shear zones (see “Hyperspectral Data Analysis for Mapping Coastal Landslide Hazards Along the Big Sur Coast”). One of her primary goals was to help Caltrans manage scenic Highway 1 while protecting the valuable natural resources of the marine sanctuary.

Hapke is now sharing her knowledge of the Big Sur region with Warrick’s team, and she is eager to leap back into studies of Big Sur coastal change. Hapke can provide the team with historical data from the central California coast to help identify trends in coastal retreat caused by landslides.

“I’ve got drawers filled with CDs of aerial photos,” she said, as well as detailed notes on historical holdings that she researched at Caltrans archives in Sacramento, California, and the National Archives in Maryland. Many of the photos can now be processed with structure-from-motion software to provide a more robust history of coastal change in the region. Warrick’s team had great success in a similar effort, when they measured the volume of material eroded from coastal cliffs at Fort Funston in San Francisco by analyzing photos collected every few years by the California Coastal Records Project (see “New Maps from Old Photos: Measuring Coastal Erosion in California”). Historical analysis can help the scientists better understand the processes by which the Mud Creek Slide is evolving as well as identify broader, regional trends.

A woman sits at a computer workstation, smiling and pointing to the computer screen to show her work. — Cheryl Hapke is now sharing her knowledge of the Big Sur region with Warrick’s team, and she is eager to leap back into studies of Big Sur coastal change. Photo credit: USGS

Diagram of deep-seated landslide, from USGS Fact Sheet 3004–3072, “Landslide Types and Processes.” Image credit: USGS

Schmidt began studying Big Sur landslides in 2001 doing research that complemented Hapke’s. Schmidt and his colleague Mark Reid (USGS California Volcano Observatory) measured the strength of different types of rock along the Big Sur coast and studied the relation between rock strength and the occurrence of deep-seated landslides. Slides are considered deep-seated if the slip occurs on a surface more than 10 to 15 feet below the ground.

“Deep-seated rock slides along the Big Sur coast sculpt the landscape, transport sediment to the Pacific Ocean, and frequently damage Highway 1,” said Schmidt.

Two photos showing people standing near some kind of pole, collecting data up on top of a cliff overlooking the ocean. — Left: Mark Reid (left) and Kevin Schmidt collecting GPS data along the Big Sur coast on September 19, 2007. Photo credit: Maiana Hanshaw, then with the USGS, now with the in Switzerland (swisstopo). Right: Kevin Schmidt collecting GPS data on the Big Sur coast on March 19, 2015. Photo credit: Mark Reid, USGS.

Reid and Schmidt also began conducting regular GPS surveys to study rates of landslide movement relative to rainfall amounts and wave energy over a 2.5-mile stretch of coast near Lucia, close to Paul’s Slide. They continue to conduct those surveys twice a year.

“We carry out our surveys in March and September of each year,” said Schmidt, “and we see the highest displacement rates in the March-to-September timeframe, not over the winter. In contrast to shallow landslides that can mobilize into debris flows during rain storms, these mountain-scale, deep-seated slides respond over much longer time scales.”

Large landslides with deep failure surfaces, like Paul’s Slide and the Mud Creek Slide, tend to slip weeks to months after heavy winter rains, when water has had time to percolate deep into the ground, raise the water table, and increase the pressures that exist in the pores—or spaces—between particles of earth. Although these “pore pressures” act on a very small scale to push rock and soil particles apart, collectively they can weaken the strength of entire hillsides.

“The Big Sur coast has a wide variety of active landslides, with speeds varying from millimeters per year to meters per minute,” said Reid. “Having an assortment of tools, such as repeat photogrammetry and high-precision GPS surveying, allows us to detect where, when, and how fast sliding happens.”

The multidisciplinary group of USGS scientists—Warrick, Ritchie, Schmidt, Reid, Hapke, Haddon, Hatcher, Harrison, and others—will continue studying the changing California coast and sharing their findings and insights with colleagues, Caltrans, and beyond.

Stay tuned!