From Core to Crust: An Engineer and Geophysicist Explains the Impact of Earth Models in Understanding Hazard Risks and Seismic Activities

Photo: 3D Visualization of Earth's Interior: Cross-Sectional Slices Highlighting the Aleutian-Alaskan Subduction Slab

Accurate earthquake predictions can save countless lives and prevent catastrophic damage. Although it may seem like a distant dream, the work of geophysicists like Lili Feng brings this reality closer every day.

As a senior machine learning research engineer and geophysicist specializing in acoustic and audio wave signal processing, machine learning (ML), and model building, Feng focuses on developing sophisticated Earth models, revolutionizing the industry's understanding and modeling of the Earth's complex structure.

Drawing upon years of expertise, the geophysicist provides invaluable insights into the workings of these models, how they unravel the planet's mysteries, and their role in safeguarding the environment and benefiting people worldwide.

Earth Modeling in a Nutshell

Earth models are complex computational representations of the Earth's structure, composition, and dynamics, integrating geophysical, geological, and geochemical data to understand the planet comprehensively.

Feng explains that these models incorporate information from seismic waves, gravity measurements, magnetic fields, and other geophysical observations to create detailed 3D representations of the Earth, from crust to core. These findings allow Feng and other geophysicists to study the Earth's internal structure, identify potential earthquake hazards, and explore the Earth's evolution over geological timescales.

Feng further emphasizes that developing earth models employs technologies like acoustic and audio wave signal processing, machine learning (ML), high-performance computing (HPC), and cloud computing. By leveraging these technologies, researchers have achieved unprecedented accuracy and detail in Earth models. Integrating these advanced methods not only refines the understanding of the Earth's structure but also enables informed decision-making in fields such as oil and gas exploration, geothermal energy development, and carbon sequestration. This signifies a significant advancement in the ability to interpret the Earth's intricate systems.

The Evolution of Earth Model Building

Geophysicists like Feng have long strived to have a detailed understanding of the Earth's structures. Feng acknowledges that Earth building models have come a long way since their inception. Traditionally, these models were constructed using limited data sets and simplified Earth structure assumptions. They were simplistic and often relied on physical representations and basic mathematical calculations.

"The old models provided a foundational understanding but lacked the precision and depth required to predict seismic activities accurately," Feng notes.

However, with advanced sensing technologies and computational methods, geophysicists can now create highly detailed and accurate representations of Earth models. For instance, Feng's innovation, SurfPy, an acoustic data processing software, harnesses Statistical Machine Learning to achieve unprecedented accuracy in subsurface imaging. Developed with sophisticated algorithms and signal-processing techniques, Feng's design has enabled the software to process quantities of seismic data, resulting in high-resolution models that reveal the complex intricacies of the Earth's subsurface.

Feng further elaborates, "In the past, we relied on sparse data points and interpolation techniques to fill in the gaps. But now, with ML and digital signal processing integration, we can process vast amounts of seismic data and create detailed models that reveal the nuances of the Earth's subsurface."

Earth Models and Seismic Activities

Accurate prediction and understanding of seismic activities through Earth models find significant applications in pressing global concerns, such as climate change and natural disasters. Advanced Earth modeling and analysis of patterns and anomalies within seismic data have significantly improved the predictive abilities of experts, enhancing the preparedness and mitigation of the impacts of earthquakes and other hazards on communities.

Beyond prediction, these models provide insights into the Earth's inner workings. By simulating different scenarios, geophysicists gain a deeper understanding of the causes and effects of seismic activities, uncovering new dynamic processes that influence the Earth's behaviour.

Feng's work exemplifies the interdisciplinary approach necessary for modern geophysics, combining geophysics, ML, and software development expertise. His integrated acoustic data processing and advanced software development for the Earth Model Building Project have shed light on the geological intricacies of regions such as Alaska, Mongolia, and Spain.

Through developing an innovative data processing and model-building workflow, Feng has harnessed the potential of vast acoustic datasets to train high-resolution Earth models. This breakthrough has led to the release of a publicly available software package, revealing critical geological and tectonic details that enhance our understanding of the Earth's dynamic processes in these regions.

Feng adds that the project has significant implications for comprehending the mechanisms behind earthquakes, volcanic eruptions, and tsunamis, particularly in Alaska. His research indicates that the Alaskan-Aleutian subduction zone is a major hazard, with tsunamis originating from this area posing a serious threat to the Western coast of the United States.

Feng's insights into natural hazards have underscored the critical role of advanced Earth models in strengthening risk assessment and mitigation practices. Such advancements are pivotal to safeguarding the lives of countless individuals in America and across the globe, showcasing the vital link between Earth science and public safety.

"With each advancement, we not only gain a clearer picture of the Earth's seismic behaviour but also a greater responsibility to use this knowledge wisely for the benefit of all," he mentions.

The Work Beyond Earth Modeling

As the challenges of climate change and natural disasters become increasingly severe, the demand for sophisticated Earth modeling techniques grows. Yet, for Feng, perfecting these models is just the beginning. He sees them as tools to drive substantive change, acknowledging their true value lies beyond their creation.

Effective hazard preparedness and resilience can only be achieved through collaboration involving scientists, policymakers, and communities. This entails merging scientific knowledge with policy development and actively engaging communities, ensuring the findings from Earth models inform actionable strategies that safeguard lives, infrastructure, and ecosystems.