Identification of Multiscale Spatial Structure of Lunar Impact Crater: A Semivariogram Approach

Identifying the spatial structure of lunar impact craters is necessary to increase our understanding of past geologic processes on the Moon. However, detecting multiscale spatial structures of craters in images in appropriate resolutions using optimum scale parameters has not been quantified. This article presents a semivariogram approach for this purpose. The range of the semivariogram model represents the minimum average size of the crater type detected in an image of a spatial resolution. The feature lag distances of the semivariogram model indicate that a series of appropriate spatial resolutions rather than a single spatial resolution are required to address multiscale lunar impact crater structures. The optimum scale parameters for delineating multiscale crater structures in segmentation are constrained by the range and feature lag distances derived from semivariogram of the corresponding image in a certain spatial resolution. This article fills the gap in quantifying multiscale spatial structure of impact craters using semivariogram analysis for optimizing object-based crater mapping.