Aisah, Aisah (2020) Pemodelan Intensitas Gempabumi di Sulawesi-Maluku dengan Menggunakan Bayesian Mixture Poisson dan Log-Gaussian Cox Processes. Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06211850010026-Master_Thesis.pdf Download (2MB) | Preview |
Abstract
Indonesia merupakan salah satu negara yang rawan terjadi gempa karena terletak pada wilayah cincin api Pasifik aktif. Sulawesi dan Maluku merupakan dua daerah di Indonesia yang paling sering terjadi gempabumi karena hampir 50% gempabumi selama tahun 2009 – 2018 di Indonesia terjadi di wilayah Sulawesi-Maluku. Poisson Process merupakan model dasar pada spatial point process yang dapat digunakan untuk memodelkan gempabumi jika memenuhi kriteria homogenitas, independen, dan berdistribusi Poisson. Namun, umumnya tidak semua kriteria terpenuhi khususnya pada pemodelan gempabumi. Mixture Poisson dan Log-Gaussian Cox Process adalah model yang dikembangkan karena tidak ketidakhomogenan pola penyebaran peristiwa gempabumi. Pemodelan intensitas gempabumi dengan window observasi di wilayah Sulawesi-Maluku dilakukan dengan melibatkan tiga covariates yaitu jarak episentrum gempabumi ke sesar aktif, zona subduksi, dan gunung api terdekat. Metode estimasi parameter model menggunakan pendekatan Bayes dengan Markov Chain Monte Carlo (MCMC) dan Intergrated Nested Laplace Approximation (INLA), sedangkan metode pembangkitan sampel pada metode MCMC dengan menggunakan gibss sampling. Hasil analisis menunjukkan bahwa intensitas gempabumi di Sulawesi-Maluku secara signifikan diperoleh oleh jarak episentrum gempabumi terhadap sesar aktif, zona subduksi, dan gunung api terdekat pada model Poisson dan Mixture Poisson Process, sedangkan jarak episentrum gempabumi terhadap zona subduksi terdekat tidak berpengaruh signifikan terhadap intensitas gempabumi pada model LGCP. Model terbaik untuk memodelkan intensitas gempabumi signifikan di Sulawesi-Maluku adalah LGCP tanpa faktor geologis zona subduksi yang dipilih berdasarkan nilai WAIC terkecil.
========================================================================================================================
Indonesia is one of the countries prone to earthquakes because it is located in the active Pacific ring of fire. Sulawesi and Maluku are the two areas in Indonesia that have the most frequent earthquakes because nearly 50% of earthquakes during 2009 - 2018 in Indonesia occurred in the Sulawesi-Maluku region. The Poisson Process is a basic model of the spatial point process that can be used to model earthquakes if it meets the criteria for homogeneity, independence, and has a Poisson distribution. However, generally not all criteria are met, especially in earthquake modeling. Mixture Poisson and Log-Gaussian Cox Process is a model developed because there is no inhomogeneity of the distribution pattern of earthquake events. Earthquake intensity modeling with observation windows in the Sulawesi-Maluku region is carried out by involving three covariates, namely the distance of the earthquake epicenter to the active fault, the subduction zone and the nearest volcano. The method of estimating model parameters uses the Bayes approach with Markov Chain Monte Carlo (MCMC) and Intergrated Nested Laplace Approximation (INLA), while the sample generation method in the MCMC method uses gibss sampling. The results of the analysis show that the earthquake intensity in Sulawesi-Maluku is significantly obtained by the distance of the earthquake epicenter to the active fault, subduction zone, and the nearest volcano in the Poisson and Mixture Poisson Process models, while the distance of the earthquake epicenter to the closest subduction zone has no significant effect on the intensity of the earthquake. on LGCP models. The best model for modeling significant earthquake intensity in Sulawesi-Maluku is LGCP without subduction zone geological factors which is selected based on the smallest WAIC value.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | Gempabumi, Gibss Sampling, INLA, MCMC, WAIC, Earthquake |
| Subjects: | Q Science Q Science > QA Mathematics > QA274.2 Stochastic analysis Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression Q Science > QA Mathematics > QA279.5 Bayesian statistical decision theory. Q Science > QA Mathematics > QA76.9 Computer algorithms. Virtual Reality. Computer simulation. |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis |
| Depositing User: | Aisah Aisah |
| Date Deposited: | 27 Aug 2020 08:32 |
| Last Modified: | 23 Dec 2023 15:58 |
| URI: | http://repository.its.ac.id/id/eprint/81485 |
Actions (login required)
 |
View Item |