Get started
Work

A Framework for Detecting Hidden Terrestrial Impact Structures: Integrating Geophysics, Remote Sensing, and Machine Learning

Year: 2026
Type: preprint
Abstract: The terrestrial impact record is sparse compared with the cratered surfaces of the Moon, Mars, Mercury, and other solid bodies. This incompleteness reflects not only Earth's active geology but also the difficulty of detecting eroded, buried, forest-covered, ice-covered, and marine impact structures. This paper presents a framework for detecting hidden terrestrial impact structures through the integrated use of geophysics, remote sensing, and machine learning. It synthesizes evidence from impact databases, gravity and magnetic anomaly mapping, seismic reflection profiling, digital elevation model (DEM) morphometry, LiDAR bare-earth mapping, multispectral and hyperspectral imaging, and automated crater detection algorithms. The central thesis is that individual methods can identify candidate structures, but only a staged, multi-modal workflow can reduce false positives and move candidates from morphological suspicion toward geological confirmation. The proposed workflow proceeds from global gravity and magnetic screening to DEM and LiDAR analysis, spectral anomaly detection, machine-learning-assisted candidate detection, seismic imaging where appropriate, and final field validation through drilling, petrography, shock metamorphic evidence, and geochemistry. The framework is designed to be testable: it can be validated on known impact structures, calibrated against rejected pseudo-craters, and then applied to poorly surveyed regions. The paper does not claim to identify a new impact structure. Its contribution is methodological: it provides a roadmap for integrating techniques that are often used separately in geology, planetary science, remote sensing, and computational analysis.
DOI: 10.5281/zenodo.21231351
Open Access status: green
OpenAlex ID: W7167638866
Cites:
Cited by:
Related to: