Offshore Faults along Central and Northern California

Science Center Objects

From Point Conception to Cape Mendocino, seafloor faults have been, in the past, mapped in varying ways and without enough detail to assess their earthquake potential. To provide this important information, USGS uses advanced technology to image offshore faults that could trigger devastating earthquakes near densely populated areas and a nuclear power plant.

A man guides a data collection instrument, that is strapped to two large pontoons, out into water from the stern of a boat.

USGS scientist Jackson Currie deploys a chirp sub-bottom profiler (in the center) from research vessel Parke Snavely. The chirp is attached to pontoons to keep the equipment from running aground in the shallow waters of San Pablo Bay, California.

Researching earthquake hazards on land has its challenges; fault lines can run right through cities, where it’s difficult to use sound-generating equipment to image the underground fractures. In marine environments, a ship can send sound through water and sediment and back to produce a clear image, but other problems may arise. In San Pablo Bay, for example, USGS scientists wanted to pin down the location of an important segment of the Hayward fault—a fault considered most likely to produce the next large earthquake in the San Francisco Bay area. However, San Pablo Bay averages only 2 meters deep—too shallow for a large ship—and gas just beneath the seafloor interferes with the imaging. So, USGS geophysicist Janet Watt used a small boat that could operate in shallow waters to launch seismic equipment called a “chirp” floated on pontoons. What she found from the very detailed images beneath the seafloor confirmed suspicions that the Hayward fault actually heads toward the Rodgers Creek fault, a relationship that could generate a larger magnitude earthquake—even stronger than the 1989 magnitude 6.9 Loma Prieta earthquake—if both rupture together. 


The Central California coast is known for its natural beauty. Much of this stunning landscape is shaped by movements along active faults between the North American and Pacific tectonic plates on the U.S. west coast. The same forces that create the high coastal mountains and control the paths of coastal rivers also lead to devastating earthquakes that endanger coastal populations and infrastructure. One important example—the Great 1906 San Francisco earthquake (magnitude 7.8)—occurred along the San Andreas fault a few kilometers off the coast of San Francisco. This event resulted in about 3,000 deaths and destroyed more than three quarters of San Francisco.

The USGS plays a prominent role in assessing earthquake hazards, providing information that informs building codes, insurance rate structures, relicensing of nuclear power plants, risk evaluations, and public policy. Such assessment depends on accurate descriptions of faults, including their location, length, geometry, slip rate, and rupture history, as well as the connections between faults. Documenting the offshore portions of significant active faults in central and northern California contributes crucial information to the national earthquake-hazard assessment effort. Mapping efforts in the past have been patchy and used a variety of instruments that produced lower-resolution data. By applying consistent mapping techniques with state-of-the art tools, scientists can vastly improve knowledge about faults, and compare data collected at different times and in different regions.

The most significant faults within the plate boundary in central and northern California include the San Andreas, San Gregorio-Hosgri, and Hayward-Rodgers Creek fault zones. Each of these fault zones has important offshore sections that, until recently, were not mapped in great detail. For 300 kilometers between Pacifica and Cape Mendocino, about 60 percent of the trace of the San Andreas fault lies beneath the ocean floor. West of the San Andreas fault, the 400-kilometer-long San Gregorio-Hosgri fault extends primarily offshore between Point Conception and Bolinas, and sits within 3 nautical miles (in state waters) of the Diablo Canyon Power Plant. East of the San Andreas fault, the Hayward and Rodgers Creek faults are considered the most likely faults in the San Francisco Bay area to have a damaging (magnitude greater than 6.7) earthquake in the next 30 years. New geophysical evidence suggests the Hayward and Rodgers Creek faults may be directly connected north of San Pablo Bay—resolving a long-standing debate among scientists.

Oblique-angle illustration showing the seafloor offshore of Half Moon Bay, California, with bumps of bedrock protruding.

Perspective view of seafloor offshore of Half Moon Bay, showing scarp (arrows) along the eastern strand of the San Gregorio fault zone. Rocks are notably upwarped and folded adjacent to the fault.

Central California map with fault lines, larger faults and location of Diablo Canyon Power Plant are are labeled.

Main faults along the northern and central California Coast. (DCPP= Diablo Canyon Power Plant)

What the USGS is doing

The USGS is collecting higher-resolution offshore geophysical data to better characterize these faults. Scientists bounce sound off the seafloor to image the bottom (bathymetric mapping), or image the layers of sediment and rock beneath the seafloor (seismic mapping). For example, multibeam sonar and chirp systems both use high frequency sound to create detailed views, respectively, of the seafloor and features beneath the seafloor. Lower frequency sound sources, such as the mini-sparker, can penetrate deeper and image as much as 300 meters below the seafloor. In addition, detailed measurements of the Earth’s gravitational and magnetic fields near the seafloor can tell us about the physical properties of rocks on and below the seafloor, and help scientists locate and estimate the shape (dip) of the faults that cut those rocks.

Since 2008, 18 research cruises spanning nearly 200 days at sea have given USGS scientists the opportunity to collect more than 7,000 kilometers of high-resolution seismic data and magnetic profiles, and to map over 400 square kilometers of seafloor in very high detail. Running the seismic equipment across a fault multiple times—in straight lines from one-half to one kilometer apart—can pick up valuable fault details. For example, establishing how features on the seafloor, or features below the seafloor, are offset can reveal how fast a fault is moving (slip rate), and when the last earthquake occurred along a fault. Without this level of detailed imaging, it’s difficult to accurately describe faults and their interactions.

This work involves many outside collaborators, including students and faculty at Oregon State University. The California Seafloor Mapping Program(CSMP), funded in large part by the state of California, has supported the collection of nearly 5,000 square kilometers of high-resolution bathymetric data in state waters (from shore out to 3 nautical miles), including virtually all of the central and northern California coast as well as San Pablo Bay just north of San Francisco.

Man points to a computer screen with his pen, there are other computer screens and equipment with keyboards, on a ship.

During a geophysical cruise off northern California, Sam Johnson from USGS points out the location of the San Andreas fault on a map generated by multibeam data to correlate with the new, incoming seismic-reflection data. Taken during the San Andreas Fault 2010 Expedition, NOAA Ocean Exploration and Research.

Three men and one woman stand, smiling, on the fantail of a docked boat in a marina, holding an instrument used to collect data.

USGS Pacific Coastal and Marine Science Center field crew showing off the new magnetometer, named Magnetron, on fantail of Research Vessel (R/V) Parke Snavely.

What the USGS has learned

A man, standing at a table and talking, gestures to maps laying on the table.

Geophysicist Sam Johnson explaining details of the San Andreas fault zone at USGS offices in Santa Cruz.

San Andreas

Mapping along the San Andreas fault between San Francisco and Cape Mendocino has revealed the complexity of strike-slip faults, including many strands of the fault that are active, and fault-bounded basins and uplifts. Uplifts within the fault interrupt sediment movement in several locations and help control the shape of the coast. Sediment-filled basins can amplify ground motion and shaking in an earthquake, and fault strands can indicate possible fault movement along other branches, making it challenging to calculate slip rates. Additionally, near Bodega Bay, the main San Andreas fault was found to be located about 800 meters west of its previously mapped location. Adjacent to the fault, strong ground motions have generated significant seafloor failures (debris flows) on the gently sloping (1°) shelf. Such areas are important to avoid when placing offshore structures. Farther north, USGS mapped the offshore section of the San Andreas for the first time in detail from where it goes offshore at Cape Arena to its termination at the junction of three tectonic plates off Cape Mendocino, California.

Hosgri-San Gregorio

USGS mapped the Hosgri fault zone in high-resolution for about 100 kilometers between Piedras Blancas and Point Sal in central California, where the fault runs within 3 nautical miles of the Diablo Canyon Power Plant. This comprehensive imaging helped highlight fault connections, which are important because longer faults can produce larger earthquakes. This level of detail also revealed the diversity of deformation along the fault, showing uplifts and depressions from small bends—a complexity not captured by previous mapping techniques. By using cutting-edge analysis, USGS scientists also examined seismic data in three dimensions along a fault bend in the Hosgri fault zone to help visualize the fault and spot pathways that fluids might follow.

Comprehensive mapping is important not only for capturing nuances of the fault, but also for slip rate calculations. In the northern section, the Hosgri fault diverges into two strands, running north and west around a central uplifted block—Piedras Blancas. Slip rate on the northerly strand is about 2.6 millimeters per year, but overall slip rate must be established on both strands. Slip rate likely varies along the Hosgri fault depending on whether adjacent faults are merging or diverging from the main fault. USGS is now extending this detailed mapping both north and south along the fault.

Computer model output looking at an angle, at seismic data represented by lines and swirls, with streams of color going upward.

Three-dimensional view of the Hosgri fault 45 meters below the seafloor, revealing fault strands (black), and potential paths along the fault that fluid could follow (green/blue). The other colors represent different geologic layers.

Map of the seafloor showing its characteristics which include long, linear features that are earthquake fault zones.

Seafloor offshore of Point Estero (PE) showing east (EH) and west (WH) strands of the Hosgri fault zone. Arrow points to a seafloor slope (a 12,000 year old shoreline) that has been offset by the east Hosgri strand, indicating a slip rate of about 2.6 millimeters per year.

Hayward-Rodgers Creek

USGS scientists have been using marine magnetic and chirp seismic reflection data to help create a 3D model of the Hayward and Rodgers Creek faults, whose geometry beneath San Pablo Bay (a northern arm of San Francisco Bay) is not well known. Chirp seismic profiles collected in 2014 show a previously unrecognized strand of the Hayward fault within the bay that may connect with the Rodgers Creek fault onshore. A direct connection makes it easier for an earthquake to rupture both these faults, potentially creating a larger earthquake than if the two faults were to rupture independently. This work follows a series of studies by the USGS in which scientists have discovered that the Hayward, Calaveras, and San Andreas faults are more interconnected than previously thought.

Man squats near a small box on a dirt road near a small pond of water.

USGS scientist David Ponce measuring gravity using a gravimeter along the Hayward-Rodgers Creek fault zone just north of San Pablo Bay, California.

Two men stand and a woman sits on tilled earth near marsh grass, woman works on a black box, one man holds paper.

USGS scientists Kevin Denton (left), Katherine “Kyeti” Morgan, and David Ponce set up a magnetic base station during fieldwork along the Hayward-Rodgers Creek fault zone in wheat fields north of San Pablo Bay.

News and Videos

New hunt for old earthquakes in San Francisco and San Pablo bays: Geologists join together over when Bay broke apart
San Jose Mercury News, July 2017

Underwater Secrets of the Hayward Fault Zone: Integrated 3D imaging to understand earthquake hazards
USGS Public Lecture, May 2017

Mystery Gap: Connecting Earthquake Faults near San Francisco, California, Requires Many Approaches
USGS Sound Waves, February 2017

CBS This Morning features USGS scientists studying link between earthquake faults near San Francisco, California
USGS News, November 2016
VIDEO: “California Prepares for Next Big Earthquake”
CBS This Morning, November 2016

Coring the Hayward-Rodgers Creek fault zone in San Pablo Bay, California, to unravel the history of faulting beneath the bay
USGS News, November 2016

B-roll Video Clips: “Sampling and Coring in San Pablo Bay
Videographer: Janet Watt, USGS
Editor: Rex Sanders, USGS

Link between two earthquake faults near San Francisco, California, revealed by detailed sub-seafloor mapping
USGS News, October 2016

VIDEO: “2 fault lines under San Francisco Bay connected, could produce major quake, study says
KTVU News Channel, October 2016

Two San Francisco-area earthquake faults found to be connected, October 2016

Why San Francisco’s next quake could be much bigger than feared
New Scientist, October 2016

New Study Finds Hayward Fault Connected To Second Fault, Potentially Making An Earthquake Much Worse
SFist, October 2016 

VIDEO: “Two Bay Area earthquake faults found to be connected
KRON 4 News Channel, October 2016

Two of San Francisco's most dangerous fault lines are linked - and that's very bad news
International Business Times (UK), October 2016

Rodgers Creek Fault Traced through Santa Rosa
USGS News Release, April 4, 2016

Special Report: “Fault impact on Central Coast
KION News Channel, February 2016

Discovery of possible connection between two earthquake faults in the San Francisco Bay area featured in the San Francisco Chronicle and other newspaper, radio, and television outlets
USGS News, January 2016

Quake danger grows: Two Bay Area faults linked in new research
San Jose Mercury News, January 2016

New data on 2 Bay Area faults cause worry about next big quake”
San Francisco Chronicle, January 2016

Alarming Discovery Shows Bay Area’s 2 Most Dangerous Earthquake Faults May Be Connected
CBS SF Bay Area, KPIX 5, January 2016

CALIFORNIA: Major fault near reactors links to 2nd crack
The Washington Times, November 2015

Mapping project reveals ancient faults in changing seafloor
San Francisco Chronicle, July 2015

New Maps Reveal California’s Sensational Seafloor Geography
Wired, May 2015

Using 3D Visualization, Geologists Explore the Complex Areas Where Faults Join and Split
KQED Science blog, November 2014

To see some of the faults on land and a few offshore, such as the Hosgri:


Looking at coastal bluffs from the ocean, with red arrows drawn on the photo to show the locations of uplifted marine terraces.

A view of the spectacular central California coastline from research vessel (R/V) Parke Snavely. Note the emergent marine terraces (arrows), which are ancient coastlines and evidence for coastal uplift likely caused by interactions and movements on the faults.
Credit: Sam Johnson, USGS Pacific Coastal and Marine Science Center.