Methods used by large companies to target marketing to individual consumers are being applied in the search for energy and mineral resources.
The sheer amount of data of all kinds being generated and stored is unprecedented. Researchers are exploring ways to sift through massive amounts of information to tease out patterns that may have been overlooked or not considered before, explained Robert Hazen, executive director of the Deep Carbon Observatory at the Carnegie Institution for Science in Washington, D.C.
Hazen is co-author of a new study published in American Minerologist that examines the use of network analysis in the search for new mineral species. The mineral hazenite is named for him.
“It’s similar to when you order something from Amazon or Netflix,” Hazen said.
The companies have access to your recent purchases, plus other information like ZIP code and credit use. By analyzing shopping patterns, “They see there’s a high probability” you will like this product, said Hazen.
“There may be some subtle connections, like activities you’ve shared with friends on social media. It’s a very powerful driver for targeted marketing,” he said.
“The techniques apply equally well to the natural world,” he added.
Application to Oil
“Mining and petroleum companies have been doing this for decades, but computers can tease out correlations we didn’t see before,” he said.
Explorationists have tons of data relating to distribution of resources, depth, stratigraphy, microbial and biological information, trace elements and much more.
“You look for very high dimensional relationships,” he said.
“If I’m looking for oil and I recognize sediment, that’s a one-to-one correlation. We can look at many more variables simultaneously. There may be some complicated non-linear combination that can help maximize our chances of discovery.”
New network analysis visualization methods “can be very powerful in showing trends that might not be obvious to the
human mind,” Hazen said.
“The idea is clearly applicable to energy exploration.”
Hazen said computerized analysis techniques are being used by medical professionals to gain more accurate diagnoses, and by governments to track possible terrorist activity.
The technique goes beyond traditional geology by amassing data about how and where minerals have formed and using that information to help find other deposits, according to the report by Hazen, lead author Shaunna Morrison, also of the Carnegie Institution, and several other colleagues.
Mineral Challenge
The researchers have launched a project to find and identify new carbon-bearing minerals whose existence has been predicted. Ten new minerals already have been discovered through the project, which invites researchers and amateur collectors worldwide to join the search.
Hazen said models indicate at least 145 “missing” species are yet to be discovered.
Hazen said the project, which can be explored at mineralchallenge.net, has been a “transformation in the mineral world. The mineralogy community seems to love it.”
Network visualization techniques can help point to specific locations where rare, undiscovered minerals are likely to be found.
In addition to locating new minerals, the Mineral Challenge website reports that the research into carbon mineral evolution is providing a fresh perspective of Earth’s history by addressing suites of new questions, which could influence Earth materials research and education:
- What were the earliest carbon-bearing minerals on Earth?
- Did carbon-bearing minerals play a role in the origin of life?
- How did the evolution of life affect the evolution of carbon minerals (and vice versa)?
- Are there carbon-bearing minerals on the moon and Mars?
- Are humans affecting the diversity and distribution of carbon-bearing minerals?
Hazen said a new undertaking, the Deep Time Infrastructure Project, involves some 50 collaborators and is aimed at gaining new insights regarding Earth’s evolving oxidation states of the atmosphere, oceans and near-surface environments.
He said analytics promise a “very dynamic future ... it will transform science.”
While some discoveries may not have economic applications, “If we understand distribution and diversity, we can make better predictions and better husband the resources we have. It’s a tremendously important part of any approach to earth resources.”