What are the best practices and standards for data quality and management in sequence stratigraphy projects?

One of the most overlooked factors in data collection is “the purpose” of data collection, especially at the core scale. It is very difficult to integrate data collected into SS project if particular dataset underwent only a very stringent technical QC. As the Sequence Stratigraphy combines data from a very distant fields of geology and the scale of observation (I.e., seismic and geochemistry), the QC process for each field may be enough for the particular field, but not enough for the “big picture”. The most striking example may be the highly heterogeneous intervals, within which the key data from the Sequence Stratigraphy perspective, may be lumped into one category.

Raw data values and their data acquisition parameters (i.e. locations, offsets, equipment type/model/date, duration of test, etc.) along with the results of each critical step in the processing of that data and the key parameters for each separate processing sequence, need to be preserved. Most important is the raw data! Do not adjust or update raw data fields to reflect filtering or any other processing step for convenience or database design issues. Preserve and maintain the original data and all data acquisition parameters.

In this case the “TITO” (thrash-in-thrash-out) rule applies. If the whole project doesn’t encounter for differences specific for all (not just each) tool used, the final data will not meet the standard requirements.

One of my scientific “guru” said: “ it is a rock science, not the rocket science” and it still holds. If you don’t do your ABC’s properly during the kick-off, you won’t (*unless you’re lucky) get your results right.

This portion of your analyses will almost always involve complex, and often expensive, software. Each software will require specific data input formats and nomenclatures. Manage/design your database and attribute schemes to readily support your data migration strategies.

I wish I could say that I have found a better tool than the mashup of results into powerpoint slides, but I haven't. Good luck.

This is deceptive in that it can be very difficult. There are no complete "earth model" databases that have effectively encompassed this issue. You must determine the primary focus of your particular interest and the extent to which you can identify and measure the key parameters you need for your attribute analyses. Then be sure to incorporate those in your database design and management, whether they are addressed by the software you use or not.

This article format being tested by LinkedIn has been easy to read and comment about. I foresee comments for interesting topics becoming a huge mess of overlapped, linked, and partially linked references. How to manage Best Practices and Lessons Learned? How to manage topics, subtopics and links (taxonomies, ontologies, etc.)? Have keyword approaches worked for you and how does AI impact search and retrieval? How about verification/validation of data? I advocate manage your KID: KID = Knowledge, Information, Data. Knowledge has source/proponent/date as critical attributes, Information is metadata that requires attributes for processes, while Data is raw data, data versioning, and must be accessible for easy formatting and sharing.