Data Types - General Comments

There are two fundamental types of data that will be stored in MetPetDB
  • Sample data (information, images, analyses, mineralogy, etc of a specific sample)
  • Derivative data (information derived from calculations based on samples, e.g. P-T conditions, P-T paths, etc.

Sample data

We are currently working to design the specific requirements of the sample data.

The basic unit of information is based on the physical sample, which is typically the hand sample collected in the field. Individual samples are typically subdivided into subsamples for further work. A subsample can be a thin section, a polished thin section, or a piece of the sample ground up for bulk chemical analysis. There can be numerous subsamples for an individual sample, and all chemical work must refer to a subsample.

For any individual sample, there are four general types of information that will be stored in the database:

There are minimum requirements for each data type and details of each can be found in other pages.

Derivative data

There are many types of derivative data that we would like to include in the database, because these may be of tremendous value to non-petrologists who wish to learn what is available from our science. For example, a structural geologist working in an area may be interested in the P-T conditions that are constrained by the petrology. He or she may not wish to examine the raw data, but only the resulting P-T calculations. While we know that this approach is fraught with innumerable pitfalls, we envision a system where the data would be presented with a sense of its uncertainty so that unwise use could be kept to a minimum.