Calculating slope gradient variations in the submarine landforms by R and Python statistical libraries

Abstract

This research focuses on the analysis of the submarine geomorphology in the Mariana Trenchlocated in west Pacific Ocean. The research question is to identify variations in the geomorphicform and bathymetry in different segments of the trench. Technically, the paper applies Pythonand R programming statistical libraries for geospatial modelling of the data sets. Themethodological approach of the statistical data analysis by scripting libraries aimed to visualizegeomorphic variations in the 25 transect profiles of the trench. Multiple factors affect submarinegeomorphology causing variations in the gradient slope: geological settings (rock composition,structure, permeability, erodibility of the materials), submarine erosion, gravity flows of waterstreams, tectonics, sediments from the volcanic arcs, transported by transverse submarinecanyons. Understanding changes in geomorphic variations is important for the correct geospatialanalysis. However, modelling such a complex structure as hadal trench requires numericalcomputation and advanced statistical analysis of the data set. Such methods are proposed by Rand Python programming languages. Current research presented usage of statistical libraries forthe data processing: Matplotlib, NumPy, SciPy, Pandas, Seaborn, StatsModels by Python. Theresearch workflow includes following steps: Partial least squares regression analysis; OrdinaryLeast Square (OLS); Violin plots and Bar plots for analysis of ranges of the bathymetric data;Isotonic Regression by StatsModels library; Data distribution analysis by Bokeh and Matplotliblibraries; Circular bar plots for sorting data by R; Euler-Venn diagrams for visualizingoverlapping of attributes and factors by Python. As a result of the data analysis, thegeomorphology of the trench slopes in 25 transecting profiles was modelled. The resultsachieved by the statistical data modelling show differences in the gradient slope in varioussegments of the trench depending on its spatial location. This shows complex geologicalstructure of the trench. The paper contributes towards the methodological development of thedata analysis in marine geology through the stepwise workflow explanations with a case studyof Python and R applications