Skip to main content Skip to secondary navigation
View of the front of Stanford Campus

Stanford Earth imaging Project (SEP)

is an industry-funded academic consortium whose purpose is to improve the theory and practice of estimating 3-D and time-lapse earth models from active and passive seismic data.

Main content start

SEP has pioneered innovations in 3-D seismic imaging and processing for more than 50 years. Today, we are continuing our research on data analysis and imaging with an increasing focus on novel applications to support the ongoing energy transition. The fundamental knowledge and the new methods we are developing will increase the safety, reduce the environmental impact and cost of geophysical monitoring of large-scale CO2 geologic sequestration and the discovery and production of hydrocarbons necessary to support human society during the energy transition and beyond. We are pioneering the use of modern data-acquisition technologies (e.g., fiber sensing) and computational tools (e.g., high-performance and cloud computing) for increasing the resiliency of urban environments to geologic and other natural hazards, often amplified by climate change.

Our students and postdoctoral scholars acquire broadly applicable skills by tackling problems related to imaging huge 3-D seismic datasets. Ph.D. students are required to apply their thesis main methods to a 3-D, or 3-C (three components) field dataset. We also undertake small 2-D data-analysis projects with geophysical data of all kinds. The diversity of applications exercises our judgment and creativity at combining fundamentals of statistical signal theory, optimization theory, numerical analysis, and wave propagation theory, and this has led us to numerous improvements and some breakthroughs. Current active research projects include robust and sparse elastic full waveform inversion from data acquired with geophones and fiber-optic sensing. We are also working on the applications of machine learning to seismic data processing and imaging that complement and enhance conventional physics-based methods.

We organize our research to facilitate technology transfer and immediate large-scale impact by using high-efficiency computational hardware such as Graphic Processing Units (GPUs) and modern software tools such as containers and cloud computing. We aim to facilitate the verification of our research results by distributing our publications and software to the wider community. All PhD theses, and research progress reports that are at least three years old, are made available to the public on our website.

Upcoming Events