Plotting Ternary Diagrams by R Library ggtern for Geological Modelling

Authors : Polina LEMENKOVA

Pages : 16-25

View : 32 | Download : 14

Publication Date : 2019-12-15

Article Type : Research Paper

Abstract :Data modelling and visualization are interdependent parallel research methods in Earth sciences critical to the proper understanding of the geological phenomena: processes, structure and dynamics. Various approaches of data analysis provided by R programming language satisfy goal of the geodata processing. Selected way of data representation and analysis of the factors provided by ternary diagrams is important way of data visualization and modelling in applied geology. Visualizing geological variables by ternaries enables to highlight, which factors are dependent on and affect others causing variations in the system. Sediment thickness, location on certain tectonic plates, closeness of the volcanic areas, steepness and depths are the geologic, tectonic and geomorphic variables tested using R based modelling in {ggtern} library. Visualizing attributes as a three-component interrelation by triangular plots gives an insight to the geological properties of the system. Traditional techniques for visualization of pairwise linear correlations between two factors on the plots fail to recognize triple variations of the factors within the system. The approach proposed by ternaries identify and handle data correlations by triple factors. Other graphical models include circular plot and Euler-Venn diagram showing quantitative and qualitative geospatial data modelling.
Keywords : ternary diagrams, geologic modelling, R programming, ggtern library, data analysis, machine learning