Landsat images and artificial intelligence techniques used to map volcanic ashfall and pyroclastic material following the eruption of Mount Agung, Indonesia
- Publication Type:
- Journal Article
- Citation:
- Arabian Journal of Geosciences, 2020, 13 (3)
- Issue Date:
- 2020-02-01
Open Access
Copyright Clearance Process
- Recently Added
- In Progress
- Open Access
This item is open access.
© 2020, Saudi Society for Geosciences. Of the 829 active volcanoes distributed worldwide, 129 are located in Indonesia. Two recent eruptions of Mount Agung on Bali Island, Indonesia, on November 21, 2017, and January 11, 2018, produced massive ash, steam, and gas emissions. Rainwater carried these pyroclastic materials in cold lahars to southwestern parts of the island. Because explosive eruptions of Mount Agung during 1963–1964 produced voluminous ashfall and catastrophic pyroclastic flows, the monitoring of this volcano has been considered essential. Land cover (LC) mapping is one method commonly used to monitor the spread of materials in volcanic areas due to the inaccessibility of field data during ongoing eruptions. We analyzed multispectral data using two different classifiers: an artificial neural network (ANN) and a support vector machine (SVM). Landsat imagery was used to generate a LC map with four feature classes: rock and sand, vegetation, cloud, and shadow. The ANN method was more accurate than the SVM method, with classification accuracies of 94.67% and 97% for the first and second Mount Agung eruptions, respectively. The SVM classifier was better than ANN at classifying images taken prior to eruption, with an overall accuracy of 91.60%. Thus, both classifiers accurately distinguished eruption products and environmental features, and are suitable for LC classification in volcanic regions.
Please use this identifier to cite or link to this item: