The SCEC Community Models, now collectively known as CXM, have been a key component of the SCEC collaboration since the 1990’s. CXM has significantly expanded in recent years and is now a collection of six models that describe different features of Southern California geology that influence seismic activity. 

CXM currently consists of the Community Fault Model (CFM), Community Geodetic Model (CGM), Community Rheology Model (CRM), Community Stress Model (CSM), Community Thermal Model (CTM), and the Community Velocity Models (CVM and the UCVM platform, which delivers multiple velocity models).

Three-dimensional view of the Geological Framework (GF), (a key component of the CRM) highlighting granodiorites that are present in Southern California

While some of these models have existed since the 1990’s, in 2017, the fifth phase of SCEC (SCEC5) established and formalized the community modeling efforts by establishing the CXM working group. The CXM workgroup aims to advance and unify scientific working groups, members of the research computing staff, and the CXM leadership structure to facilitate improvements to the models. Scott Marshall and Laurent Montesi have been co-leaders of the CXM working group since 2017 and 2019, respectively. Under their leadership and that of former CXM leader, Liz Hearn, SCEC community models have become more detailed, expansive, community-driven, and accessible.

Since their beginning, the SCEC community models have been open-source and have encouraged participation from all parts of the community. The first official SCEC community model was the Community Fault Model (CFM), a three-dimensional representation of active faults in southern California. The success of collaboration on the CFM encouraged the community to expand and include the other earthquake-focused models.

“I’ve been using the Community Fault Model since the beginning of my Ph.D. in 2004,” Marshall said. “This got me familiar with the details of the model, how it was created, and what it is based on. Now, I am a member of the CFM development team. This is common amongst the SCEC Community models: people that begin as curious users often later become model developers"

Over time, the concept of community models became a core part of how SCEC characterizes seismic hazards, because the models compile the knowledge and expertise of thousands of scientists who are working on similar problems within earthquake science. 

What is the purpose of CXM?

The goal of CXM, according to Marshall, is “to provide community-driven, open-source models that reflect the state of our knowledge of various parameters within California. For example, if we didn’t have the Community Fault Model and someone wanted to model or simulate faults in California, they would have to compile 100 plus years of research, read all of those papers, digitize all of their work and turn it into a three-dimensional model, which would take decades,” Marshall said. “The CFM essentially provides this in an open-source format for the community. We aim to keep the model as up to date as possible, but this is obviously a challenge as much new work is constantly being published and not all interpretations of data agree. For this reason, many of the community models also provide alternative model versions.”

The CXMs use data from thousands of scientists around the world who study a wide range of parameters that influence earthquakes in California.

“The models present an integrated view of the southern California region,” Montesi said. “To understand how we get to the state we are in today and what will happen later, you need to have an all-encompassing understanding of what the region is like. That is what CXM allows you to do.”

3D perspective view of a preliminary statewide SCEC Community Fault Model. Faults are colored by the major fault area tentatively assigned in the model, and seismicity (Hauksson et al., 2012 and updates, Waldhauser et al., 2008, 2009) is colored by the year of occurrence. This preliminary statewide model is under revision and will be evaluated by a range of collaborators before an initial statewide version is released by SCEC. Image credit: Andreas Plesch.

Andreas Plesch, who has been involved with the Community Fault Model since it was created, said that having community models is important. “You can compare and contrast the ideas and develop results based on a standard. That’s a big advantage,” he said.

With each CXM, a unique set of challenges emerges. For example, among all the models, the Community Geodetic Model (CGM) is one of the most difficult to maintain because the data changes on a daily basis. 

“Producing CGM is a big challenge because it takes a lot of resources, and people have to put in a lot of time,” said Mike Floyd, the primary developer of the SCEC CGM. “But one of the big challenges is using the CGM data to understand when something odd happens [within the data], something that is off the normal trend because smaller events are tough to detect within the noise of geodetic data.”

These models aren’t just used by geoscientists. For example, earthquake engineers make extensive use of SCEC’s velocity models, according to Marshall. “They need to know how fast seismic waves travel through the subsurface to understand how these waves will impact built structures,” Marshall said. 

The ultimate goal of earthquake science is to better characterize and mitigate earthquake hazards. This is done through large-scale efforts that ultimately produce seismic hazard assessments. One such recent effort, UCERF3 (the Third Uniform California Earthquake Rupture Forecast), was partially based on SCEC’s Community Fault Model and aims to predict ruptures on California’s many faults and their resulting ground motion. This project was a collaboration between the United States Geological Survey, California Geological Survey, and numerous SCEC community members. It was partially funded by the California Earthquake Authority, the largest provider of earthquake insurance in California. Having all of the information in the various CXMs available to people working on seismic hazard estimates like UCERF3, should improve future seismic hazard models.

Participants in the 2023 CRM workshop explore rock outcrops along Cucamonga Canyon to understand the field constraints on the Community Rheology Model (CRM). Photo Credit: Tran Huynh.

Improvements in CXM IT Infrastructure

CXM developments have been facilitated by an effective collaboration between the CXM scientific working groups and the SCEC software development staff. SCEC software developer, Mei-Hui Su, is the lead software developer for the CXM explorers and for the SCEC Unified Community Velocity Model (UCVM) software. The CXM leaders and Mei have established an effective development process in which geoscientists organize the necessary data and identify the essential capabilities required for each CXM explorer (the web-based tools that provide access to the CXM's), and Mei implements the required capabilities using advanced software tools including 3D visualization libraries, open-source mapping tools, and geospatial databases. The CXM explorers are developed using shared software infrastructure, which allows for re-use of software between explorers, reduces the new software development required for each website, and lends the CXM explorers a common look and feel. The modern web-based software libraries used by the explorers enable highly efficient visualization capabilities such as selectable map overlays and 3D visualizations in web browsers even on portable devices such as tablets and phones. By simplifying access to the CXM data and providing visualization tools, the CXM explorers contribute to both the scientific review and use of the models.

SCEC’s Associate Director for Information Technology Philip Maechling said, “Our experience is that when we can get geoscientists and computer scientists to work together closely, we make the most rapid progress developing new research tools. The close collaboration between the CXM working group and the research computing group has resulted in an efficient software development process, with little wasted effort, and it has ensured that the resulting software tools are truly useful to SCEC’s research community.”

Future of CXM

One of the long-term goals of CXM is to develop and integrate these models into a self-consistent suite.

“The ultimate goal is to create a physics-based simulation of California and its faults based on the current state of our knowledge,” Marshall said. “But it is important to emphasize that no community model will ever be finished because there is always more data coming in. Models can always be improved. New and improved data and methods are constantly being published. No model can ever exactly represent the geologic complexity of California.”

The SCEC community has an optimistic and challenging vision: to make these community models self-consistent and interoperable so they can be collectively used to simulate California’s complex fault system over time. To reach this goal, all models must be extensively tested to evaluate their performance and to highlight any potential shortcomings. The various SCEC community models are tested in various independent studies by earthquake scientists all around the world. When the models do not fit the existing data well, they are revised.

Cross-section of a seismic velocity model with detailed descriptions of sedimentary basins in Southern California. From left to right: Santa Maria Basin, Ventura Basin, Los Angeles Basin. and Salton Sea.

At the 2023 SCEC Annual Meeting, two new important resources were announced: The Earthquake Geology Information Warehouse and the SCEC Geologic Slip Rate Database. The Earthquake Geology Information Warehouse is intended to provide an organizational umbrella for future important datasets related to earthquake geology. For now, the Earthquake Geology Information Warehouse hosts the Precariously Balanced Rocks Project and the SCEC Geologic Slip Rate Database. The SCEC Geologic Slip Rate Database comprises a collection of geologic slip rate estimates (from hundreds of studies) for faults in California, Nevada, and a small portion of northernmost Mexico.This simplifies the process of identifying existing geologic slip rate estimates for a given region or fault(s) and provides direct web links to the relevant publications (where available) so that users can find, read, and gain an understanding of the relevant work. Importantly, the SCEC Geologic Slip Rate Database is meant to be a living archive with an approximately annual update cycle.

“This (Geologic Slip Rate Database) is going to be hugely useful because people won’t have to search through thousands of papers to find the fault information they need,” Marshall said. 

In the end, Marshall and Montesi both emphasized that the goal of CXM is to make these models openly available and easily accessible to anyone who may be interested in these products. “This kind of information should be publicly available because it’s interesting and important work and people shouldn’t have to spend hundreds of hours just searching through papers and other people’s research to find the information they need,” Marshall said. “We hope that the CXMs and the new Earthquake Geology Information Warehouse will help to facilitate advances in earthquake science.”

SCEC has been an open collaboration for more than twenty years. The vision of SCEC is to encourage people with different expertise to work together on the most challenging and important problems in earthquake science. “None of this would work if people weren’t willing to collaborate,” Marshall said. “That’s how we’ve gotten where we are.”

Explore the models here:

• Community Models (CXM) Overview
• Community Fault Model (CFM)
• Community Geodetic Model (CGM)
• Community Rheology Model (CRM)
• Community Stress Model (CSM)
• Community Thermal Model (CTM)
• Community Velocity Model (CVM)
• Unified Community Velocity Model (UCVM)
• The Earthquake Geology Information Warehouse
• The Geologic Slip Rate Database

References

Hauksson, E., Yang, W., and Shearer, P. M., "Waveform Relocated Earthquake Catalog for Southern California (1981 to 2011)"; Bull. Seismol. Soc. Am., Vol. 102, No. 5, pp.2239-2244, October 2012, doi: 10.1785/0120120010 SCEC Contribution 1528

Waldhauser, F. and D.P. Schaff, Large-scale relocation of two decades of Northern California seismicity using cross-correlation and double-difference methods, J. Geophys. Res.,113, B08311, doi:10.1029/2007JB005479, 2008.

Waldhauser, F., Near-real-time double-difference event location using long-term seismic archives, with application to Northern California, Bull. Seism. Soc. Am., 99, 2736-2848, doi:10.1785/0120080294, 2009.

About the Authors

	Shreya Agrawal is an Earth scientist and journalist focusing on climate change, environmental and social issues and politics. She graduated from USC in 2023 with a dual bachelor's degree in Geological Sciences and English, and a master's degree in journalism. She hopes to better communicate science to the public and bridge gaps in science communication.
	Scott T. Marshall is a professor in the Department of Geological and Environmental Sciences at Appalachian State University. His research focuses on computational modeling of fault mechanics and satellite geodesy. He is currently co-leader of the Community Modeling Group (CXM) of the SCEC Science Planning Committee and is a developer of the SCEC Community Fault Model.
	Philip J. Maechling is the Associate Director for Information Technology for the SCEC where he develops and performs large-scale seismic hazard calculations using high-performance computing. He also develops and releases open-source seismological and engineering software used in ground motion modeling and simulations. He is a member of the SCEC Executive Committee and the Science Planning Committee.