Pemodelan Risiko Gempabumi Di Sumatera Menggunakan Generalized Additive Poisson Model Untuk Model Thomas Cluster Process

Susanto, Tabita Yuni (2021) Pemodelan Risiko Gempabumi Di Sumatera Menggunakan Generalized Additive Poisson Model Untuk Model Thomas Cluster Process. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

Text 06211740000026-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Sumatera merupakan salah satu kawasan rawan gempabumi karena kondisi geografisnya yang dilalui oleh sesar, zona subduksi, dan gunung berapi. Pada penelitian ini, kejadian gempabumi di wilayah Sumatera dimodelkan dengan Thomas cluster process karena persebaran titik gempabumi di pulau Sumatera cenderung mengelompok. Terdapat kecenderungan bahwa gempabumi terjadi di daerah yang dekat dengan gunung berapi, zona subduksi, dan sesar aktif, sehingga fungsi intensitas pada model Thomas cluster dimodelkan menggunakan Generalized Additive Poisson Model (GAPM). Estimasi nonparametrik dari intensitas gempabumi menunjukkan bahwa masing-masing kovariat mempengaruhi intensitas gempabumi dalam bentuk non log-linear. Hasil eksplorasi data menunjukkan bahwa gempabumi di Sumatera tidak homogen dan membentuk cluster yang disebabkan karena faktor geologis seperti keberadaan gunung berapi, zona subduksi, dan sesar aktif. Pemodelan intensitas gempabumi dengan GAPM yang menghasilkan AIC terkecil adalah dengan menggunakan kovariat jarak gempabumi ke gunung dan subduksi, basis function cubic regression spline, 3 titik knot pada setiap kovariat, dan smoothing parameter 0,001 (subduksi) dan 0,011886 (gunung berapi). Dilihat dari envelope K-function dan nilai AIC terkecil, Thomas cluster process dengan intensitas gempabumi dimodelkan menggunakan Generalized Additive Poisson Model merupakan model terbaik dengan AIC sebesar -1.909.492.
========================================================================================================
Sumatra is one of the areas prone to earthquakes because of its geographical conditions that are traversed by faults, subduction zones, and volcanoes. In this study, the occurrence of earthquakes in the Sumatra region is modeled by the Thomas cluster process, because the distribution of earthquake points on the island of Sumatra tends to be clustered. There is a tendency that earthquakes occur in areas close to volcanoes, subduction zones, and active faults, so that the intensity function in the Thomas cluster process model is modeled using the Generalized Additive Poisson Model (GAPM). Nonparametric estimates of the earthquake intensity show that each covariate affects the earthquake intensity in a non log-linear form. The results of data exploration show that the earthquake in Sumatra is not homogeneous and forms clusters due to geological factors such as the presence of volcanoes, subduction zones, and active faults. Earthquake intensity modeling with GAPM that produces the smallest AIC is to use covariate distance from earthquake to volcano and earthquake distance to subduction, using the basis function cubic regression spline, using three point knots for each covariate, and using smoothing parameter 0.001 on subduction covariates and 0.011886 on the volcano covariate. Based on the envelope K-function and the smallest AIC value, the Thomas cluster process with the earthquake intensity modeled using the Generalized Additive Poisson Model is the best model with an AIC of -1.909.492.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Gempabumi, Gunung Berapi, Sesar, Thomas Cluster, Process, Zona Subduksi. Cluster, Earthquake, Fault, Subduction, Thomas Cluster Process, Volcano.
Subjects:	Q Science > QE Geology > QE538.8 Earthquakes. Seismology
Divisions:	Faculty of Mathematics, Computation, and Data Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Tabita Yuni Susanto
Date Deposited:	05 Mar 2021 03:05
Last Modified:	05 Mar 2021 03:05
URI:	http://repository.its.ac.id/id/eprint/83493

Actions (login required)

View Item