GRSG Conference 2022: Orbit to Outcrop

Title: VIS-NIR spectroscopy in geology and mining: How to make the data work for you

Author: Evelien Rost


Chemometric models that utilize large numbers of spectral points to predict sample composition have been widely accepted for quantitative and qualitative analysis within agricultural, pharmaceutical and industrial markets. Traditionally, geological remote sensing applications and mineral identification programs use univariate peak ratios or peak recognition software to relate spectral reflectance characteristics to the spectral features of the spectrum.

However, overlapping spectral features limit the effectiveness of simple ratio- and peak recognition routines and prevent adequate quantitative predictions from samples with an unknown composition.

This session will introduce the chemometric solutions of the SummitCAL team of Malvern Panalytical. This team develops chemometric models for NIR spectroscopy, performs feasibility studies, and provides training to your team to build internal expertise on chemometric modelling. With several case studies in the field of minerals and mining we will introduce quantitative and qualitative mineral models and show how chemometric models improve interpretation of mineral composition.

The chemometric models can be continuously calibrated and monitored for model improvement. Our services bring value to various applications in remote sensing and mining, for example by monitoring and maintaining fertilizer levels in soil, or quantitative prediction of muscovite mineralogy for improved ore processing.