the impact of big data in geoscience

Steve Garrett, manager of Chevron

Chevron’s Steve Garrett, Global Technology Centre manager in Aberdeen.

During a recent conference in London, organised by The Geological Society, Chevron’s Steve Garrett, Global Technology Centre manager in Aberdeen, discussed the impact of “Big Data” in geoscience.

The three-day event brought together early career geoscientists and leading industry and academic experts to discuss the opportunities and challenges of Big Data and showcase advances in data collection and interpretation technology. The event presented an opportunity to encourage learning and collaboration between geoscience and computer science practitioners.

data science for earth science: an industry perspective

data science for earth science: an industry perspective

Talking “Big Data” and geoscience.

After setting the broader industry context around technology development past and present, Garrett’s keynote speech, entitled “Data Science for Earth Science – Perspectives from Industry”, focused on key messages that subsurface is usually under-sampled and measurements are diverse and indirect.

“We measure properties such as acoustic impedance and resistivity rather than direct measurements of oil and gas, and have tens or hundreds – rather than millions or billions – of wells as data points,” explained Garrett.

“To build a sufficiently large and resilient data set to apply such methods, there is potential for the industry to collaborate to address these challenges and opportunities. This can be seen by the current U.K. continental shelf project championed by the Technology Leadership Board and Exploration Task Force, which is applying machine learning to seek missed pay using well logs from the Northern North Sea.”

A frequently asked question is what these technology advances could mean for professional development for individuals in the oil and gas industry? In answering, Garrett adds that some of our more numerate earth scientists may choose to re-train as data scientists.

“While the role of the domain expert remains key,” he adds, “we all need to learn enough about data science to frame the projects with computer science and statistical experts.”