Ground-based visible–near infrared hyperspectral imaging for monitoring cliff weathering of a volcanic island in Dokdo, South Korea

• DOI: 10.1016/j.enggeo.2022.106854
• Publication Date: 2022-09-15T06:36:42Z
• Authors: Young-Sun Son; Sang-Gun Noh; Eun-Seok Bang; Kwang-Eun Kim; Seong-Jun Cho; Hyunseob Baik
• Keywords: VNIR; Cliff; Outcrop

Coastal cliffs undergo erosion and weathering more rapidly under the influence of strong waves sea winds, leading to stability environmental conservation issues. Ground-based hyperspectral imaging is useful for identification geological interpretation minerals or rocks in vertical outcrops that are difficult confirm from an aerial view through situ investigation safety reasons. High spatial spectral resolutions visible–near infrared (VNIR) sensors can be advantageous detecting made volcanic rocks; however, their potential not well known. In this study, two classification techniques, mixture-tuned matched filtering (MTMF) support vector machine (SVM), were applied VNIR data cliff face a island Dokdo, South Korea, results compared. Results show SVM superior MTMF minerals. The distinction between with similar compositions textures deteriorated using both methods. shading surface owing due unevenness stratification also affected accuracy classification. This study shows ground-based image analysis powerful effective approach predict possible geomorphological changes on islands, as it explore highlight potentially vulnerable locations.