Abstract:Understanding the source lithology of basalts can greatly impact our understanding of magmatic processes, which can help us gain insight into mantle heterogeneity and crustal material recycling. However, many uncertainties remain about how to recognize the relationship between major element patterns of basalts and source lithological and/or compositional diversities using typical petrological methods. In this study, discriminant functional analysis and multivariate regression are carried out using major element logratios for literature experimental melts of four primitive mantle‐like lherzolites and six mafic lithologies (five pyroxenites and one hornblendite). A parameter called FCMS (FCMS = ln (FeO/CaO) − 0.058 * (ln (MgO/SiO2))3 − 0.636 * (ln (MgO/SiO2))2 − 1.850 * ln (MgO/SiO2) − 1.170, all the major elements in weight percent) is proposed to identify source lithologies for basaltic melts. When the logratios ln(K2O/Al2O3), ln (TiO2/Na2O), and ln (Na2O/K2O) are incorporated into FCMS, the temperature and pressure effects on the compositional heterogeneity of peridotites melts can be significantly reduced; a more powerful parameter called FCKANTMS (FCKANTMS = ln (FeO/CaO) − 0.08 * ln(K2O/Al2O3) − 0.052 * ln (TiO2/Na2O) − 0.036 * ln (Na2O/K2O) * ln (Na2O/TiO2) − 0.062 * (ln (MgO/SiO2))3 − 0.641 * (ln (MgO/SiO2))2 − 1.871 * ln (MgO/SiO2) − 1.473, all the major elements in weight percent) can be obtained. Olivine fractional crystallization and accumulation, although they can significantly change most major element contents and ratios, usually increase or decrease FCMS and FCKANTMS by only 0–0.15, which is one order of magnitude lower than their variations in the primary experimental melts of different lithologies. Importantly, approximately 80% and 50% low to moderate degree (F < 60%) partial melts (n = 303) of mafic lithology (with or without volatiles in source materials) can be distinguished from melts of peridotite and transitional lithologies when FCKANTMS combined with ln (CaO/TiO2) and ln (SiO2/(CaO + Na2O + TiO2)). Thus, the chemical markers could be used to constrain source lithology and origin of natural basalts.

Amalgamations are valid in compositional data analysis, can be used in agglomerative clustering, and their logratios have an inverse transformation

Proportion-based normalizations outperform compositional data transformations in machine learning applications

Discriminant Analysis Method for Compositional Data Based on An Isometric Logratio Transformation

The chiPower transformation: a valid alternative to logratio transformations in compositional data analysis

A comprehensive workflow for compositional data analysis in archaeometry, with code in R

Three approaches to supervised learning for compositional data with pairwise logratios

Fisher Discriminant Analysis of Compositional Data.

Spatial Hyperspheric Models for Compositional Data

On the distribution of isometric log-ratio transformations under extra-multinomial count data

Error Propagation in Isometric Log-ratio Coordinates for Compositional Data: Theoretical and Practical Considerations

Units Recovery Methods in Compositional Data Analysis

Comparison of log-ratio and log10 chemical elemental data analysis of Central Amazonian pottery and archaeological implications

Empirical observations on the effects of data transformation in machine learning classification of geological domains

Clustering transformed compositional data using K-means, with applications in gene expression and bicycle sharing system data

Using Major Element Logratios to Recognize Compositional Patterns of Basalt: Implications for Source Lithological and Compositional Heterogeneities

Multinomial logit PLS regression of compositional data

Multiple linear regression modeling for compositional data

Compositional transformations can reasonably introduce phenotype-associated values into sparse features

Simple and flexible sign and rank-based methods for testing for differential abundance in microbiome studies

Partial correlations in compositional data analysis

Multivariate analysis for geochemical process identification using stream sediment geochemical data: A perspective from compositional data