Opinions expressed by Digital Journal contributors are their own.
In an era where much of the Earth’s mysteries have been studied, one might assume that the planet has yielded all its secrets. However, for Lili Feng, a Senior Machine Learning Research Engineer at a leading data analytics firm in the media sector, what was once shrouded in the Earth’s mystery is now unveiled by technology. The emergence of machine learning (ML) and seismic data analysis marks the beginning of a new era of discovery, revealing possibilities we’ve only just begun to explore.
Unlocking earth’s secrets with machine learning
As a former Research Geophysicist holding a Doctor of Philosophy from the University of Colorado Boulder, ML has become Feng’s key to uncharted territories, revealing hidden patterns and insights in seismic data and audience behaviours, illustrating how technology can be a beacon for exploration and understanding in diverse fields.
Feng’s methodology involves refining existing processes to discover new potentials. His work transcends applying advanced ML algorithms and signal processing techniques to reveal the Earth’s secrets and leveraging these new technologies to reshape the landscape of audience analytics in the media industry.
Seismic waves: The new frontier in earth science
Seismic wave analysis, a process involving acoustic signal interpretation, is traditionally conducted through seismic surveys. These techniques have been applied in diverse industries, including media, healthcare, and communication technology. In geophysical exploration, however, seismic waves are repercussions from the Earth’s core, offering critical insights into its concealed strata. Feng compares this process to “listening to the Earth’s heartbeat,” enabling scientists to ‘see’ beneath the surface across various geological layers, unravelling the mysteries of the Earth’s inner architecture.
However, the conventional method of analyzing seismic signals faced significant challenges due to large data volumes and complexity. Feng’s adoption of ML techniques in this realm presents a significant shift in approach, enhancing the efficiency and depth of seismic data analysis.
Revolutionizing seismic analysis with integrated acoustic data processing
Feng’s pioneering study, “Integrated Acoustic Data Processing and Advanced Software Development for Earth Model Building,” represents a comprehensive approach to acoustic data processing. Utilizing ML algorithms, Feng and his team have adeptly processed and interpreted seismic data with unparalleled precision and detail, bringing patterns and structures that previously eluded detection to light.
Feng’s work has culminated in creating intricate three-dimensional models of the Earth’s interior, pivotal in deciphering geological phenomena, including the genesis of natural resources and the mechanics of seismic activities. Integrating ML with seismic analysis not only enhances the clarity and resolution of subsurface models but also accelerates the exploration process. This enables quicker and more effective investigations into Earth’s subterranean mysteries and provides deeper insights into geological events like earthquakes.
Spearheading tools for practical impact
The Senior ML Research Engineer’s journey in redefining seismic data analysis spans several groundbreaking projects. Aside from his research, he also developed ‘SurfPy,’ a state-of-the-art software at the forefront of statistical ML. According to Feng, the ingenuity of SurfPy lies in its ability to utilize sound waves to create detailed models or pictures, unravelling the mysteries of the Earth’s layers and offering a vital tool for predicting natural disasters and understanding geological structures.
The novelty of SurfPy lies in its multifaceted approach. “We’re combining various types of sound waves that traverse through the Earth, providing us with a more comprehensive view of the subsurface,” Feng explains. The software employs Bayesian Monte Carlo inference, a sophisticated statistical ML technique, to enhance the clarity of its underground images. This method not only improves image quality but also quantifies the confidence level in these visualizations. Engineered for efficiency, SurfPy harnesses Python’s programming prowess to operate across multiple computers simultaneously, significantly speeding up data processing.
SurfPy has yielded significant geological insights globally. For instance, in Alaska, Feng’s software rendered 3D models that illuminated the Earth’s crustal layers, revealing the movements over millions of years and the critical boundaries of tectonic plates. Shifting to Mongolia, the software uncovered details in the Khövsgöl rift, including variations in crustal thickness and underlying hot spots, which are crucial to understanding surface elevations. Similarly, in Spain, SurfPy’s insights extended to the alignment of underground rocks and their relationship with the Earth’s surface, offering valuable perspectives on landscape formation.
“SurfPy is a gateway to a new understanding of our planet. By revealing hidden geological and tectonic features beneath Alaska, Mongolia, and Spain, we’re advancing scientific knowledge and enhancing our ability to mitigate and assess natural hazards. This has profound implications, especially in the United States, where understanding seismic activities in Alaska can be key to protecting countless lives on the Western coast from potential tsunamis,” Feng elaborates.
An approach to balanced innovation
While instrumental in advancing the understanding of the Earth’s subsurface, the technology spearheaded by Feng raises significant ethical concerns. The detailed data generated by SurfPy, which can create precise and comprehensive subsurface models, could be misused for potential natural resource exploitation. Its ability to pinpoint resource-rich areas with unprecedented accuracy might lead to intensified mining or drilling activities, leading to environmental impacts, such as habitat destruction, water contamination, and increased carbon emissions.
Such insight underscores the necessity for experts like Feng to advocate for stringent ethical guidelines in geophysical exploration and data management, particularly in the media industry. The careful handling, storage, and protection of sensitive data are crucial to prevent privacy breaches. It emphasizes the need for a balanced approach to technology, where its use aligns with the greater good.
In this pursuit of technological excellence, it is vital to remember that the accurate measure of success lies not just in the power of these tools but in how responsibly and ethically they are wielded.
