Integrating WorldView-2 imagery and terrestrial LiDAR point clouds to extract dyke swarm geometry: Implications for magma emplacement mechanisms

Abstract

Dyke geometries are useful indicators of the palaeostress field during magma emplacement. In this paper, we present a multi-scale extraction method of dyke geometries by integrating WorldView-2 (WV2) imagery and terrestrial light detection and ranging (LiDAR) data. Color composite and fusion WV2 images with 0.5-m resolution were generated by using the Gramm–Schmidt Spectral Sharpening approach, which facilitates the discrimination of dyke swarms and provides the ability to measure the orientation, exposed length, and thickness of dykes in sub-horizontal topographic exposures. A terrestrial laser scanning survey was performed on a sub-vertical exposure of dykes to obtain LiDAR data with point spacing of ~0.02m at 30m. The LiDAR data were transformed to images for extracting dyke margins based on image segmentation, then the dyke attitudes, thicknesses, and irregularity of dyke margins were measured according to the points on dyke margins. This method was applied at Sijiao Island, Zhejiang, China where late Cretaceous mafic dyke swarms are widespread. The results show that integrating WV2 imagery and terrestrial LiDAR improves the accuracy, efficiency, and objectivity in determining dyke geometries in two and three dimensions. The ENE striking dykes are dominant, and intruded the host rock (mainly granite) with sub-vertical dips. Based on the aspect ratios of the dykes, the magmatic overpressure was estimated to be less than 11.5MPa, corresponding to a magma chamber within 6.6km in the lithosphere.