Clustering within rootless cone groups on Iceland and Mars: Effect of nonrandom processes

Abstract

This study aims at quantifying the spatial distribution of cones within rootless cone groups (RCGs). Our data set consists of (1) seven Icelandic RCGs (identified through field investigations), (2) seven candidate RCGs on Mars (identified through Mars Orbiter Camera (MOC) and Thermal Emission Imaging System (THEMIS) images), and (3) four groups of impact craters on Mars (also identified through MOC and THEMIS images) to determine if they can be remotely distinguished from the RCGs purely on the basis of their spatial distribution. Several independent statistical techniques are used, including nearest neighbor analysis, analysis of variance (ANOVA), and linear alignment detection analysis. Our results indicate that the spatial distribution of each RCG is not random: Within a cone group, cones preferentially form near existing cones. The presence of at least one significant linear cone alignment in each RCG (and strong linear alignments in some groups) suggests that rootless cones form as the surface signature of preferred lava pathways. An ANOVA on mean nearest neighbor distances reveals the Martian cone groups to be statistically indistinguishable from the Icelandic RCGs, supporting existing interpretations that they represent rootless cone groups. A similar ANOVA showed that the Martian cone groups do not resemble the Martian impact crater clusters studied: The impact craters have significantly greater nearest neighbor distances and show no evidence of aggregation within a crater group.