Pemanfaatan Eqtransformer & Pyocto Dalam Deteksi Kejadian Seismik, Studi Kasus Jawa Timur

Wicaksono, Galih Bagus (2026) Pemanfaatan Eqtransformer & Pyocto Dalam Deteksi Kejadian Seismik, Studi Kasus Jawa Timur. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5017221003-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (5MB) | Request a copy

Abstract

Penelitian ini bertujuan mengidentifikasi kinerja model machine learning EQTransformer dan PyOcto dalam mendeteksi kejadian seismik di Jawa Timur serta membandingkan katalog gempa yang dihasilkan dengan katalog regional. Data continuous waveform dari 27 stasiun AUSPASS periode September–November 2015 diproses melalui automatic picking (EQTransformer), phase association (PyOcto), penentuan hiposenter (NonLinLoc), relokasi (HypoDD), dan perhitungan magnitudo lokal. EQTransformer menghasilkan 2.023.622 picks yang diasosiasikan PyOcto menjadi 17.421 event. Setelah quality control ketat, diperoleh 205 event berkualitas tinggi dan 73 event hasil relokasi dengan magnitudo ML 0,59–2,96. Distribusi hiposenter menunjukkan konsentrasi di zona patahan aktif (Kendeng, Waru-Cepu, Pasuruan) dan kompleks vulkanik (Gunung Wilis) dengan dominasi gempa dangkal (<15 km). Perbandingan dengan katalog ISC mengonfirmasi workflow ini berhasil mendeteksi mikro-gempa (ML <3,0) yang tidak tercatat katalog regional, meningkatkan kelengkapan katalog gempa Jawa Timur pada rentang magnitudo rendah.
===================================================================================================================================
This study aims to identify the performance of machine learning models EQTransformer and PyOcto in detecting seismic events in East Java and compare the resulting catalog with regional catalogs. Continuous waveform data from 27 AUSPASS stations during September–November 2015 were processed through automatic picking (EQTransformer), phase association (PyOcto), hypocenter determination (NonLinLoc), relocation (HypoDD), and local magnitude calculation. EQTransformer produced 2,023,622 picks that PyOcto associated into 17,421 events. After rigorous quality control, 205 high-quality events and 73 relocated events with magnitude ML 0.59–2.96 were obtained. Hypocenter distribution shows concentration in active fault zones (Kendeng, Waru-Cepu, Pasuruan) and volcanic complexes (Mount Wilis) with dominance of shallow earthquakes (<15 km). Comparison with ISC catalog confirms this workflow successfully detected micro-earthquakes (ML <3.0) not recorded in regional catalogs, improving East Java earthquake catalog completeness in the low magnitude range.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	EQTransformer, PyOcto, deteksi gempa otomatis, katalog gempa, NonLinLoc, HypoDD, magnitudo lokal. EQTransformer, PyOcto, automatic earthquake detection, earthquake catalog, NonLinLoc, HypoDD, local magnitude.
Subjects:	Q Science > QE Geology > QE538.8 Earthquakes. Seismology
Divisions:	Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Geophysics Engineering > 33201-(S1) Undergraduate Thesis
Depositing User:	Galih Bagus Wicaksono
Date Deposited:	23 Jul 2026 05:41
Last Modified:	23 Jul 2026 05:41
URI:	http://repository.its.ac.id/id/eprint/136390

Actions (login required)

View Item