I plan to create a database of mineral optical properties, searchable through a custom web-interface. The facility will be linked to the IMA database created by Bob Downs. The database will begin with data from references, but will be editable by the community of petrographers and mineralogists, so that it can improve over time. Because the system will be designed in the coming academic year (2008-2009), I am calling now for comments on the database design and search interface.
I have a fairly solid background in mineralogy, optical mineralogy and petrography, but over the years I have occasionally been stymied by some unusual mineral in thin section. While I have typically been able to use reference materials to identify the minerals in question, I have often wished for an easily searchable database for mineral optical properties to help in such situations.
The petrographic microscope is one of the most efficient tools in geology, in that it allows the collection of a great deal of information in a short period of time, with little expense. Unfortunately, many students receive little or no education in petrography these days. Geologists who have received insufficient training can use the database to support their petrographic skills. I believe also that such a facility could enhance student learning in Petrography/Optical Mineralogy if used carefully, and may even spur renewed interest in this valuable tool.
Most importantly, there is a wealth of petrographic knowledge in the community of users, most of which never gets written down, and which may pass away in the coming decades. In the past, such knowledge was passed along from teacher to student, but given the state of petrographic education, that process may now be insufficient to preserve the petrographic knowledge contained within the community. While textbooks can be of use in preventing such loss, they are limited in size and thus, scope. I hope to create a system that can record and preserve most of the petrographic knowledge attained with long labor by the best of petrographers.
I envision a way that a user, faced with a mineral he or she cannot identify, can enter some or all of a set of properties measured on the unknown into a web page. I emphasize that not all fields must be filled in order to search the database. After a short wait, the system will provide a list of minerals, ranked by the likelihood of being the unknown. Each result will be linked to a page of detailed information, a page that will also include links to outside sources of information (RRUFF, WebMineral, Google, MinDat, etc.).
The properties determined by the user are generally known imperfectly. The search will be able to accommodate such uncertainty in most determinations, and will use the stated uncertainty in producing results. For example, if the user's best determination for the index of refraction is 1.6, but can only be certain that the true value is between 1.56 and 1.63, then the search system will prefer minerals with an index of refraction range that intersects 1.6, but will also allow results whose range only intersects 1.56 or 1.63. These poorer matches will be moved down in the results list accordingly.
While some users may prefer to search only by optical properties, others may prefer to bias their results based on mineral abundance. Each mineral will have both an estimated global abundance, and estimated abundances in each of a set of rock types. The results list could therefore be biased based upon rock type. As a simple example, plagioclase and cordierite have fairly similar properties, and would both appear in many searches. With this biasing, cordierite is likely to be very low in a search on an mafic igneous rock, but substantially higher in a search on a high-T pelitic metamorphic rock.
I envision an expert adjusting the database to reflect his/her years of experience looking at minerals so as to expand the range of properties to better reflect reality. The optical properties presented with the original mineral descriptions will not, in general, accurately reflect the range of properties for that mineral in all its variations. Correct information about such a range requires many years of experience, and the true information is not typically published. This community editing is a vital part of the database, but with no gatekeepers, the quality of the data will plummet. I envision a team of experts in different mineral groups who would vet the proposed database modifications and approve those that are valuable.
Currently, the resources required for the system are minimal. However the data entry will require a great deal of effort, for which I plan to harness and pay a team of undergraduates. Some would be at my home institution, but many would likely be from elsewhere. I will apply to the NSF for funding to support this part of the project.
-Dave Hirsch, chair (and, currently, sole member) of IMA subcommittee on Mineral Optical Properties
July 24, 2008