Noer, Muhammad Yustar Afif (2026) Aplikasi Autopicking Data Mikrotremor Menggunakan Hvsrpy Dengan Window Rejection Untuk Analisa Interpolasi Parameter f0, A0, Kg, Dan Vs30 Di Wilayah Meulaboh – Johan Pahwalan, Aceh. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5017221089-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5017221089-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy |
Abstract
Kota Meulaboh – Johan Pahlawan, Aceh Barat, merupakan wilayah dengan tingkat kerawanan tinggi terhadap gempa bumi akibat pengaruh zona subduksi Sunda dan aktivitas Sesar Sumatera. Berdasarkan data USGS (2020–2025), wilayah Meulaboh dan sekitarnya tercatat mengalami 39 gempa bumi dangkal (kedalaman < 60 km) dengan magnitudo 4,0 – 6,2, gempa terbesar (M6,2) terjadi pada 2022. Penelitian ini bertujuan menganalisis pengaruh window rejection pada hvsrpy terhadap kualitas kurva HVSR, menentukan nilai frekuensi resonansi tanah (f₀), amplitudo H/V (A₀), dan indeks kerentanan seismik (Kg), memperoleh nilai Vs30 menggunakan pendekatan machine learning, serta menganalisis hasil interpolasi spasialnya. Data mikrotremor sekunder diolah menggunakan hvsrpy dengan parameter window rejection n antara 1,30 – 1,50, divalidasi terhadap kriteria SESAME (2004). Sebagai pembanding, data diolah pula menggunakan Geopsy melalui picking window secara manual, dimana hasil kurva HVSR-nya digunakan sebagai parameter inversi ellipticity curve (Dinver) untuk memperoleh profil Vs30 tiap titik. Nilai Vs30 inversi tersebut selanjutnya digunakan sebagai data train model XGBoost untuk Vs30 predict, sementara nilai Vs30 dari data USGS sebagai pembanding validasi terhadap hasil inversi dan predict. Window rejection efektif meningkatkan kualitas kurva HVSR pada seluruh 20 titik (GL1–GL20), dengan window yang direjeksi rata-rata 89,10%. Seluruh titik memenuhi tiga kriteria reliability, 18 titik memenuhi enam kriteria clarity, dan 2 titik memenuhi lima kriteria akibat amplitudo H/V rendah. Nilai f₀ berkisar 0,88 – 2,21 Hz (Tanah Tipe II/Jenis IV), A₀ antara 1,14 – 2,93, dan Kg antara 1,14 hingga 6,93, dengan 13 titik kelas rendah (Kg < 4,80) dan 7 titik kelas sedang (4,80 ≤ Kg ≤ 15) tanpa satu titik pun kelas tinggi. Estimasi ketebalan sedimen dari f₀ dan Vs30 predict berkisar 25,00 – 62,29 meter. Nilai Vs30 predict XGBoost berkisar 210,70 – 225,80 m/s, menghasilkan Mean Absolute Percentage Error (MAPE) 1,37% terhadap Vs30 inversi dan 9,16% terhadap Vs30 USGS dengan kecenderungan underestimate, sehingga seluruh titik tetap terklasifikasi kelas situs SD (tanah sedang) menurut SNI 1726:2019. Pola sebaran spasial keseluruhan konsisten dengan dominasi endapan aluvium Holosen (Qh) penyusun Meulaboh Embayment.
=================================================================================================================================
Meulaboh City–Johan Pahlawan, West Aceh, is a region highly susceptible to earthquake hazards due to the influence of the Sunda subduction zone and Sumatran Fault activity. Based on USGS data (2020–2025), the Meulaboh area and its surroundings recorded 39 shallow earthquakes (depth < 60 km) with magnitudes ranging from 4,0 to 6,2, the largest (M6,2) occurring in 2022. This study aims to analyze the effect of window rejection in hvsrpy on HVSR curve quality, to determine soil resonant frequency (f₀), H/V amplitude (A₀), and seismic vulnerability index (Kg), to obtain Vs30 values using a machine learning approach, and to analyze the resulting spatial interpolation. Secondary microtremor data were processed using hvsrpy with a window rejection parameter (n) of 1,30–1,50, validated against the SESAME (2004) criteria. As a comparison, the data were also processed using Geopsy through manual window picking, where the resulting HVSR curves were used as the ellipticity curve inversion parameter (Dinver) to obtain the Vs30 profile at each point. These inverted Vs30 values were subsequently used as training data for an XGBoost model to predict Vs30, while USGS Vs30 values served as a comparison for validating both the inversion and prediction results. Window rejection effectively improved HVSR curve quality at all 20 points (GL1–GL20), with an average of 89,10% of windows rejected. All points satisfied the three reliability criteria, 18 points satisfied the six clarity criteria, and 2 points satisfied only five due to low H/V amplitude. f₀ values ranged from 0,88 to 2,21 Hz (Soil Type II/Class IV), A₀ ranged from 1,14 to 2,93, and Kg ranged from 1,14 to 6,93, with 13 points in the low class (Kg < 4,80) and 7 points in the medium class (4,80 ≤ Kg ≤ 15), with no points in the high class. Sediment thickness estimated from f₀ and predicted Vs30 ranged from 25,00 to 62,29 meters. XGBoost-predicted Vs30 values ranged from 210,70 to 225,80 m/s, yielding a Mean Absolute Percentage Error (MAPE) of 1,37% relative to inverted Vs30 and 9,16% relative to USGS Vs30, with a tendency toward underestimation; all points nonetheless remained classified as site class SD (medium soil) according to SNI 1726:2019. The overall spatial distribution pattern is consistent with the dominance of Holocene alluvial deposits (Qh) forming the Meulaboh Embayment.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Meulaboh – Johan Pahwalan, Autopicking, Hvsrpy, Mikrotremor, Vs30, Window Rejection. |
| Subjects: | H Social Sciences > HV Social pathology. Social and public welfare > HV551.5.I4 Hazard mitigation Q Science > QE Geology > QE538.8 Earthquakes. Seismology Q Science > QE Geology > QE539 Microseisms. T Technology > TN Mining engineering. Metallurgy > TN269 Prospecting--Geophysical methods |
| Divisions: | Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Geophysics Engineering > 33201-(S1) Undergraduate Thesis |
| Depositing User: | Muhammad Yustar Afif Noer |
| Date Deposited: | 24 Jul 2026 03:10 |
| Last Modified: | 24 Jul 2026 03:10 |
| URI: | http://repository.its.ac.id/id/eprint/136938 |
Actions (login required)
 |
View Item |