Evaluación de métodos de normalización aplicados a bases de datos minerales de Espectrografía de Infrarrojo Cercano (SWIR) provenientes de múltiples instrumentos y para análisis de vectores de exploración

Over the past decade, short-wave infrared (SWIR) spectroscopy has made significant advances in detecting geochemical variations in minerals like white mica, alunite, and chlorite for exploring hydrothermal ore deposits. These variations provide valuable clues, indicating changes in temperature, pH, and fluid oxidation state towards the mineralized center. However, small calibration differences among devices challenge data integration. This study evaluates the 2200 nm Al-OH absorption feature in four white mica SWIR spectroscopy databases collected by TerraSpec™ and OreXpress™ from samples at the Grasshopper porphyry prospect. It evaluates three normalization methodologies: rescaling, mean normalization, and Z-score, yielding p-values for successful data merging of up to 0.75. Findings suggest effective normalization methods across devices, reducing biases from uncalibrated spectrometers. This research offers a methodology to correct SWIR database biases, facilitating accurate data integration across instruments for vectoring analysis.