CorStruth - Automated Analysis of HyLogger Data from the NVCL

October 2018
 1.   Improving the Interpretation of HyLogger Data Using XRF
2. A new round of uploads started.

I have been looking at how we can improve the interpretation of HyLogging data with XRF measurements. There is a note describing the results here. The main conclusions are :

  • Element concentrations determined from HyLogging TIR unmixing results compare quite well with Assay XRF results in situations where we are confident about our unmixing.
  • Where we have known problems in unmixing the comparison breaks down.
  • There is considerable potential to improve the conversion from estimated mineralogy to geochemistry by using scalar measurements from the SWIR. 
  • There is a big opportunity to use XRF results to better control the unmixing mineralogy we produce from HyLogger data
  • Routine acquisition of high-quality XRF data with HyLogger data should be an important objective for an expanded NVCL scope.
  • The Minalyse noise levels for the lighter elements indicate that it will be difficult to use this data for controlling HyLogger interpretation at the 0.1m sample interval.
  • If high quality XRF data is not available in a routine manner we should investigate real-time automated methods of selecting pXRF sampling locations while the core still laid out.

There will be a new set of holes uploaded to CorStruth in the next few months.

The current CorStruth products are:
Interpretations (like the one opposite) suitable for A4 presentation with variable bin sizes.
Browser-based scrollable images with a standard bin size of 0.25 m. (e.g. here)
Downloadable CSV files of TSA+ and jCLST interpretations binned at 1 m intervals.
Mineral Occurrence Maps (MOMA's) of various minerals identified with the NVCL data

There are three ways to access the interpretation images. (See here for more details)

CorStruth's purpose is to provide automatic interpretations of the HyLogging data from every HyLogger-3 dataset acquired as part of the AuScope National Virtual Core Library (NVCL) program. The results are free for you to copy but remember that they have been produced by automatic algorithms which will make mistakes. In this regard make sure you read the notes under Cautions.

The NVCL program is a team effort and thanks are due to everyone involved  - both individuals and organizations. And here most of them are at the 2017 NVCL Workshop in Sydney.

                                                                                                          Andy Green

Useful information

Improving the Interpretation of HyLogger Data Using XRF
Understanding the plots
Why Are the SWIR and TIR Results Different?
What is TSA+
What is jCLST?
Active Mineral List
USGS Resampled Spectra
State HyLogger Links: NSW, NT, Qld, SA, Tas, WA