Model Space-Time ETAS (Epidemic Type Aftershock Sequence) Untuk Katalog Gempa Bumi Di Sulawesi-Maluku

Faradilla, Sonia (2024) Model Space-Time ETAS (Epidemic Type Aftershock Sequence) Untuk Katalog Gempa Bumi Di Sulawesi-Maluku. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6003221018-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2026. Download (2MB) | Request a copy

Abstract

Pemodelan katalog gempa bumi berdasarkan model point processes telah dikembangkan di Indonesia seperti Poisson point processes, log-Gaussian Cox processes, Neyman–Scott point processes dan ETAS. Namun demikian, pemodelan yang diterapkan di Indonesia masih secara parsial yakni berbasis data titik lokasi koordinat dan data titik waktu. Padahal di luar Indonesia pemodelan katalog gempa bumi yang memanfaatkan data waktu, latitude, longitude dan magnitudo berhasil diaplikasikan dengan model space-time ETAS yang banyak digunakan untuk pemodelan analisis seismisitas. Apabila dilihat dari fungsi intensitas bersyaratnya, model ini memberikan informasi mengenai peluang terjadinya gempa bumi susulan yang dipengaruhi oleh waktu, lokasi koordinat dan magnitudo. Estimasi parameter pada model ini menggunakan metode maximum likelihood estimation, namun tidak semua hasil yang diperoleh close form sehingga memerlukan metode numerik dikarenakan fungsi log-likelihood pada model ini mempunyai bentuk integral lipat yang cukup kompleks. Penelitian ini bertujuan untuk menerapkan model space-time ETAS menggunakan data katalog gempa bumi di Sulawesi-Maluku seperti waktu, latitude, longitude dan magnitudo. Hal ini dicapai dengan proses estimasi parameter dua tahap yakni tahap pertama menggunakan MLE dikarenakan terdapat hasil yang tidak close form, maka dilanjutkan tahap kedua untuk mencapai kecepatan konvergensi menggunakan Davidon-Fletcher-Powell dan stokastik declustering dengan menerapkan fungsi kernel. Berdasarkan hasil estimasi parameter, magnitudo berpengaruh terhadap terjadinya gempa bumi di Sulawesi-Maluku. Hasil dari evaluasi kebaikan model memberikan informasi mengenai performa kualitas model ini baik karena memenuhi distribusi normal, serta melalui peta bahaya seismik model ini memberikan informasi mengenai sebagian besar kejadian gempa bumi terjadi di sekitar laut Maluku dengan clusters kejadian gempa bumi menyebar hampir di seluruh wilayah Sulawesi dan Maluku.
=============================================================================================================================
Earthquake catalog modeling based on point process models such as Poisson point processes, log-Gaussian Cox processes, Neyman-Scott point processes, and ETAS have been developed in Indonesia. However, the modeling applied in Indonesia is still partially based on coordinate location point data and time point data. Meanwhile, outside Indonesia, earthquake catalog modeling using time, latitude, longitude, and magnitude data has been successfully applied with the model space-time ETAS, which is widely used for seismicity analysis modeling. In terms of the conditional intensity function, this model provides information on the probability of aftershocks influenced by time, location coordinates, and magnitude. Parameter estimation in this model uses the maximum likelihood estimation method, but not all results obtained are close form, so it requires numerical methods because the log-likelihood function in this model has a rather complex folding integral form. This study aims to apply the model space-time ETAS using earthquake catalog data in Sulawesi-Maluku such as time, latitude, longitude and magnitude. This is achieved by a two-stage parameter estimation process, namely the first stage using MLE because there are results that are not close form, followed by the second stage to achieve convergence speed using Davidon-Fletcher-Powell and stochastic declustering by applying kernel functions. Based on the parameter estimation results, the magnitude affects the occurrence of earthquakes in Sulawesi-Maluku. The results of the evaluation of the goodness of the model provide information about the quality performance of this model is good because it meets the normal distribution, and through the seismic hazard map of this model provides information about most of the earthquake events occurred around the Maluku Sea with clusters of earthquake events spread almost throughout Sulawesi and Maluku.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Davidon-Fletcher-Powell, Maximum Likelihood Estimation, Model Space-Time ETAS, Seismisitas, Stokastik Declustering, Declustering Stochastic, Seismicity
Subjects:	Q Science > QA Mathematics > QA353.K47 Kernel functions (analysis) Q Science > QA Mathematics > QA401 Mathematical models.
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis
Depositing User:	Sonia Faradilla
Date Deposited:	16 Feb 2024 07:21
Last Modified:	16 Feb 2024 07:21
URI:	http://repository.its.ac.id/id/eprint/107207

Actions (login required)

View Item