Cyber-Geology on the NSF EC3 Field Trip

Last week, I had the opportunity to participate in the Earthcube EC3 Geology Field Trip [1]. The field trip was organized exceptionally well by Matty Mookerjee with support by Basil Tikoff and sponsored by the National Science Foundation. The purpose of the trip was to develop communications between the cyberinfrastructure community and those involved in field-based, solid earth geoscience.  The participants represented a mix of geologists and information technologists. Both groups included a range of age and experience from esteemed professors to PhD candidates and two high school students! EC3 consisted of day trips in the hot desert of Owens Valley California, evenings in the research station classroom, and lodging in the WMRC dorms. As a computer engineer it was a chance for me to think about how cyber can be applied to geology - an example of Soichiro Honda’s method of "Going to the spot.”

Geology is a fascinating science based on field observations, laboratory testing and extensive debate on conflicting theories. Field studies are central to the current practice of geology. Information technology can incrementally increase the efficiency of a scientist in the field, for example, using voice recognition to record as text a spoken observation of rock formation “strike and dip”. Because observations in the field are highly dependent on a specific scientist’s interests, using cyberinfrastructure to advance field observation is difficult. Currently strike/dip measurements and hand drawn sketches require the scientist to be in the field.  If, for example, it were possible to automate unit identification of an outcrop matching the hand drawn sketch, then large scale collection campaigns could result in opportunities to apply big data analytics.

An advance for field geology education could be based on Augmented Reality. I gave a short talk about AR to an evening session during the field trip [2]. We discussed examples and possible benefits of applying the OGC ARML open standard to geology. In discussions, participants agreed that AR could aid in explaining layers or other geological information to students in a more effective manner than current practice of pointing. There was skepticism that AR could aid the process of sketching an outcrop. Again there is much reliance of geologists on sketches in their notebooks. I look forward to further discussions with Magdalena and John Donahue who are developing the Field Play AR application [3].

One key step toward a cyberinfrastructure in geology is to connect digital mapping tools and field observations to databases in the cloud. An evening presentation by Terry Pavlis [4] focused on requirements for field apps. Many desirable functions for such field apps were identified and might result in NSF funding of development of field apps to benefit field research. For me the key requirement was automatic publishing from the app - an idea suggested by bunk room mate, Joe Andrew. Publishing must be integral and transparent to the scientist’s use of the tool, for example as with an EarthTime tool as suggested by Mike Williams. Understanding the scientist’s workflow and connecting the tools to that workflow are key elements, as highlighted in an evening session by Yolanda Gill. Understanding workflow provenance is key to reproducibility.

Connecting field tools to databases and connecting databases should be a key point of development in EarthCube. In the near term, tools and databases will be closely coupled with private agreements on syntax and semantics. To achieve interoperability of multiple tools with discipline databases will require community agreement on schemas for the geological observations and interpretations.  Large scale sharing of field data will require discussion and agreement on community specifications. To be most effective, community agreements should be based on open international standards such as GeoSciML [5] - a standard for the exchange of digital geoscientific information. For example there are alternate conventions for strike and dip measurements. It is not necessary and probably impossible to converge to a single convention, but it is necessary that the convention for a specific measurement be unambiguously identified in metadata that accompanies the measurement data. Discussion about strike and dip measurement conventions is highly reminiscent of discussion about conventions for latitude/longitude. A recent ad hoc OGC summit on lat/lon conventions resulted in agreement on the need for  “truth in advertising” [6]. Also needed is community agreement on schemas for rock and layer definitions - this is similar to agreement on application schemas for geographic feature types.

In the longer term its interesting to speculate how information technology could advance "integrative geology.”  Certainly the advance of discipline databases is on the road to cross-discipline sharing. That will require development of semantic agreement across geological fields, which, considering  the field trip discussion of the diversity and divisions of geology disciplines, may be awhile in coming. The situation may be similar to that in the Earth Observation domain where development of Climate Data Records and Environmental Data Records proceeds through  verification of the match between physical estimates coming from sensors and  different climatological and environmental phenomena. Another example of cyber advancements in the future would be a peer reviewed 4D model in a geographic coordinate reference system of the earth’s geological history. This would be similar to Global Circulation Models developed for climate prediction. In an evening discussion Marjorie Chan showed the video for lunar mapping that would be a great project for earth geology [7]. An example that can be achieved today with open standards like those from OGC. Discussion on this topic, while we sat looking out at Bishop Tuff, focused on the concern that cross-discipline advances are precluded by significant conceptual differences in the domains. Examples of integrative geology projects in the past have been very costly and have thus come at the expense of more domain specific research.

One topic occasionally discussed in the field trip was that despite the amazing understanding of the Earth that geology has produced, the social relevance seems less appreciated in less academic discussions. What is the social benefit of geology? This is not a question about the value of the science, it is a question of how increasing awareness of the relevance of geology could result in better funding. For example, there could be more awareness of the role of geology in such topics as tracking how “Man-Made Earthquakes Are Changing the Seismic Landscape.” [8]

My  EC3 experience of the excellent discussions provoked by “walking the contacts”leads me to predict that NSF EarthCube will indeed result in  cyberinfrastructure for geology that yields significant advances in geological research, increased awareness of the relevance of geology, and increased societal benefit..