1 Tsunami and Disaster Mitigation Research Center, Universitas Syiah Kuala, Banda Aceh, Indonesia

2 Meteorological, Climatological, and Geophysical Agency, Jakarta, Indonesia

3 GFZ German Research Centre for Geosciences, Telegrafenberg, 14473 Potsdam, Germany

4 University Potsdam, Karl-Liebknecht Street, 14476 Potsdam, Germany



BACKGROUND AND OBJECTIVES: Soil or rock types in a region are often interpreted qualitatively by visually comparing various geophysical properties such as seismic wave velocity and vulnerability, as well as gravity data. Better insight and less human-dependent interpretation of soil types can be obtained from a joint analysis of separated and independent geophysical parameters. This paper discusses the application of a neural network approach to derive rock properties and seismic vulnerability from horizontal-to-vertical seismic ratio and seismic wave velocity data recorded in Majalengka-West Java, Indonesia.
METHODS: Seismic microtremors were recorded at 54 locations and additionally multichannel analyses of surface wave experiments were performed at 18 locations because the multichannel analyses of surface wave experiment needs more effort and space. From the two methods, the values of the average shear wave velocity for the upper 30 meters, peak amplitudes and the dominant frequency between the measurement points were obtained from the interpolation of those geophysical data. Neural network was then applied to adaptively cluster and map the geophysical parameters. Four learning model clusters were developed from the three input seismic parameters: shear wave velocity, peak amplitude, and dominant frequency.
FINDINGS: Generally, the values of the horizontal to vertical spectral ratios in the west of the study area were low (less than 5) compared with those in the southeastern part. The dominant frequency values in the west were mostly low at around 0.1–3 Hertz, associated with thick sedimentary layer. The pattern of the shear wave velocity map correlates with that of the horizontal to vertical spectral ratio map as the amplification is related to the soil or rock rigidity represented by the shear wave velocity. The combination of the geophysical data showed new features which is not found on the geological map such as in the eastern part of the study area.
CONCLUSION: The application of the neural network based clustering analysis to the geophysical data revealed four rock types which are difficult to observe visually. The four clusters classified based on the variation of the geophysical parameters show a good correlation to rock types obtained from previous geological surveys. The clustering classified safe and vulnerable regions although detailed investigation is still required for confirmation before further development. This study demonstrates that low-cost geophysical experiments combined with neural network-based clustering can provide additional information which is important for seismic hazard mitigation in densely populated areas.

Graphical Abstract

Neural network based classification of rock properties and seismic vulnerability


  • The Vs30 map of Majalengka correlates relatively well with geology of the study area indicating the lialibility of the results of the research;
  • The neural network-based analysis has defined four different types of rocks of Majalengka-West Java, defining the leve lof vulnerability;
  • The SOM method recognizes the detailed vulnerable zone of the study area which locates in east characterized by high H/V, low dominant frequency and low Vs30 which is difficult to be seen by visual observation;
  • The application of neural network-based SOM method, even provides more detail information of the small clusters situated in large clusters as the cluster 3 is located within the Cluster 4.


Main Subjects


©2023 The author(s). This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit:


GJESM Publisher remains neutral concerning jurisdictional claims in published maps and institutional affliations.


Google Scholar Scopus Web of Science PlumX Metrics Altmetrics Mendeley |


GJESM Publisher

Letters to Editor

GJESM Journal welcomes letters to the editor for the post-publication discussions and corrections which allows debate post publication on its site, through the Letters to Editor. Letters pertaining to manuscript published in GJESM should be sent to the editorial office of GJESM within three months of either online publication or before printed publication, except for critiques of original research. Following points are to be considering before sending the letters (comments) to the editor.

[1] Letters that include statements of statistics, facts, research, or theories should include appropriate references, although more than three are discouraged.
[2] Letters that are personal attacks on an author rather than thoughtful criticism of the author’s ideas will not be considered for publication.
[3] Letters can be no more than 300 words in length.
[4] Letter writers should include a statement at the beginning of the letter stating that it is being submitted either for publication or not.
[5] Anonymous letters will not be considered.
[6] Letter writers must include their city and state of residence or work.
[7] Letters will be edited for clarity and length.