Digital mineralogy: from soil to oil ...and beyond!
X-ray powder diffraction (XRPD) remains at the forefront of methods that are employed to make quantitative mineralogical analyses of geological materials such as rocks, soils and mineral deposits. Indeed, when clay minerals are prominent components of the samples of interest the utility of XRPD methods to provide quantitative data is unsurpassed. Accurate quantification of large numbers (100’s-1000’s) of samples, however, can still be a challenging undertaking particularly if there is large variation in the range of minerals that may be encountered. A complimentary emerging approach is that of ‘digital mineralogy’. In tune with the current growth in ‘data science’ digital mineralogy focuses first and foremost on the XRD patterns that are acquired from samples, considering them simply as digital data containing encoded mineralogical information. Viewed like this, conventional expert analysis and interpretation involving the steps of mineral identification and quantification become somewhat redundant, at least in the early stages of processing large data sets. Instead approaches such as data mining and cluster analysis among others may be employed to explore and understand relationships between variation in mineralogy and other soil or rock properties of interest. Examples of these data-based approaches to unravelling and advancing understanding of soil property - mineralogy relationships will be illustrated using data from the African Soil Information Service (AfSiS). Additionally, the potential for the application of ‘digital mineralogy’ to the oil industry, a prolific user of XRPD data, will be illustrated. A final example will demonstrate how the ‘digital mineralogy’ approach has been used to identify soils in Scotland as potential analogues for samples from Mars that have been analysed by the ChemMin instrument on the Curiosity Rover, Mars Science Laboratory (MSL).
https://www.munich-geocenter.org/events/seminars/frontiers-in-earth-sciences-23/frontiers-in-earth-sciences-2019-05-24
https://www.munich-geocenter.org/logo.png
Digital mineralogy: from soil to oil ...and beyond!
Abstract
X-ray powder diffraction (XRPD) remains at the forefront of methods that are employed to make quantitative mineralogical analyses of geological materials such as rocks, soils and mineral deposits. Indeed, when clay minerals are prominent components of the samples of interest the utility of XRPD methods to provide quantitative data is unsurpassed. Accurate quantification of large numbers (100’s-1000’s) of samples, however, can still be a challenging undertaking particularly if there is large variation in the range of minerals that may be encountered. A complimentary emerging approach is that of ‘digital mineralogy’. In tune with the current growth in ‘data science’ digital mineralogy focuses first and foremost on the XRD patterns that are acquired from samples, considering them simply as digital data containing encoded mineralogical information. Viewed like this, conventional expert analysis and interpretation involving the steps of mineral identification and quantification become somewhat redundant, at least in the early stages of processing large data sets. Instead approaches such as data mining and cluster analysis among others may be employed to explore and understand relationships between variation in mineralogy and other soil or rock properties of interest. Examples of these data-based approaches to unravelling and advancing understanding of soil property - mineralogy relationships will be illustrated using data from the African Soil Information Service (AfSiS). Additionally, the potential for the application of ‘digital mineralogy’ to the oil industry, a prolific user of XRPD data, will be illustrated. A final example will demonstrate how the ‘digital mineralogy’ approach has been used to identify soils in Scotland as potential analogues for samples from Mars that have been analysed by the ChemMin instrument on the Curiosity Rover, Mars Science Laboratory (MSL).
