Analisis Konstituen Harmonik Dan Karakteristik Sinyal Tsunami Pada Tide gauge Dan Dart Buoy (Studi Kasus: Tsunami Kamchatka 2025)

Dafid Aulia, Ir. Abdul Basith, S.T., M.Si., Ph.D.

2026 | Skripsi | TEKNIK GEODESI

Gempa bumi megathrust berkekuatan Mw 8,8 di lepas pantai Semenanjung Kamchatka pada 30 Juli 2025 pukul 06:24 WIB memicu gelombang tsunami yang merambat melintasi Samudra Pasifik Utara. Selama perambatannya menuju pesisir, gelombang ini mengalami perubahan ketinggian akibat efek pendangkalan dan interaksi batimetri lokal yang mendistorsi rekaman data pasang surut. Meskipun berbagai studi mengenai Tsunami Kamchatka 2025 telah dilakukan, sebagian besar masih terbatas pada pemodelan sumber gempa, simulasi propagasi global makro, atau analisis pada perairan tunggal. Penelitian-penelitian terdahulu umumnya menerapkan metode analisis spektral secara parsial tanpa proses isolasi komponen pasang surut astronomis yang presisi, sehingga sinyal tsunami yang dianalisis rentan terkontaminasi oleh bias data pasut. Selain itu, terdapat keterbatasan mengenai perbandingan sistematis karakteristik transformasi sinyal antara wilayah laut lepas dan pesisir, serta belum tersedianya metodologi yang mampu menentukan waktu akhir gangguan osilasi tsunami secara definitif dan objektif. Oleh karena itu, penelitian ini bertujuan mengidentifikasi karakteristik spektral dan durasi gangguan gelombang tsunami secara komprehensif melalui integrasi metode Analisis Harmonik Kuadrat Terkecil, Fast Fourier Transform (FFT), dan Continuous Wavelet Transform (CWT) Morlet.
Pengolahan data dalam penelitian ini dilakukan menggunakan perangkat lunak MATLAB online melalui tiga tahapan utama. Tahap awal meliputi analisis harmonik menggunakan toolbox t_tide berbasis kuadrat terkecil (least square) untuk mengisolasi sinyal murni tsunami dari pasang surut astronomis. Hasil perhitungan dan pemodelan ini divalidasi menggunakan metrik Root Mean Square Error (RMSE) antara data observasi dan model prediksi. Selanjutnya, metode Fast Fourier Transform (FFT) diterapkan pada data residu untuk mengekstrak periode dominan dan amplifikasi spektral. Hasil ekstraksi FFT ini divalidasi dengan mengkomparasikan spektrum gelombang tsunami terhadap spektrum background (kondisi perairan sebelum kejadian) untuk memastikan signifikansi puncak energi. Sebagai tahap akhir, Continuous Wavelet Transform (CWT) Morlet digunakan untuk memetakan evolusi energi guna menentukan waktu tiba, durasi osilasi, dan waktu pemulihan perairan. Keabsahan analisis peluruhan energi dan resonansi lokal melalui CWT ini kemudian divalidasi secara teoretis dengan membandingkan pola perilakunya terhadap studi terdahulu, guna mengonfirmasi konsistensi korelasi fisis antara skala magnitudo gempa dengan durasi osilasi gelombang di perairan pesisir.
Melalui analisis harmonik, model prediksi pasang surut berhasil dibangun dengan akurasi tinggi (RMSE 0,042 s.d. 0,072 m pada stasiun pesisir dan 0,018 s.d. 0,038 m pada laut lepas). Hasil analisis FFT mengungkap periode dominan tsunami di laut lepas berada pada rentang 36,0 s.d. 41,2 menit, yang kemudian mengalami pemanjangan periode menjadi 42,8 s.d. 65,5 menit saat mencapai pesisir. Amplitudo di pesisir (orde 10-2 m) juga teramplifikasi secara signifikan dibanding laut lepas (orde 10-3 m) akibat efek pendangkalan dan resonansi lokal. Pemetaan CWT Morlet menunjukkan Stasiun Petropavlovsk pulih paling cepat (28 jam 46 menit), sedangkan energi osilasi bertahan paling lama di Stasiun Kushiro (77 jam 21 menit). Pemahaman teknis mengenai karakteristik ini, khususnya dalam menentukan waktu tiba, durasi fase kritis, serta puncak amplitudo maksimum, memiliki manfaat krusial bagi sistem mitigasi bencana. Informasi ini menjadi landasan penting bagi tim evakuasi untuk menentukan jendela waktu yang aman dan efektif dalam operasi penyelamatan, sehingga risiko terhadap korban jiwa dapat diminimalisir melalui tindakan tanggap darurat yang terukur.

The Mw 8.8 megathrust earthquake that occurred off the coast of the Kamchatka Peninsula on July 30, 2025, at 06:24 WIB triggered a tsunami that propagated across the North Pacific Ocean. During its propagation toward the coast, the wave underwent significant height alterations due to shoaling effects and local bathymetric interactions, which distorted the sea-level records. Although various studies on the 2025 Kamchatka Tsunami have been conducted, most remain limited to seismic source modeling, macro-scale global propagation simulations, or analyses of isolated water bodies. Previous research has generally employed spectral analysis methods partially, without a precise prior isolation of astronomical tidal components, rendering the analyzed tsunami signals susceptible to tidal bias. Furthermore, there is a lack of systematic comparisons regarding signal transformation characteristics between offshore and coastal regions, as well as a deficiency in methodologies capable of definitively and objectively determining the end time of tsunami-induced oscillations. Therefore, this study aims to comprehensively identify the spectral characteristics and the duration of tsunami wave disturbances through the integration of Least Squares Harmonic Analysis, Fast Fourier Transform (FFT), and Morlet Continuous Wavelet Transform (CWT).
Data processing in this study was carried out using MATLAB Online software through three main stages. The initial stage involved harmonic analysis using the t_tide toolbox based on the least squares method to isolate pure tsunami signals from astronomical tides. The calculation and modeling results were validated using the Root Mean Square Error (RMSE) metric between the observation data and the prediction model. Subsequently, the Fast Fourier Transform (FFT) method was applied to the residual data to extract the dominant periods and spectral amplification. The results of the FFT extraction were validated by comparing the tsunami wave spectrum against the background spectrum (pre-event water conditions) to ensure the significance of the energy peaks. In the final stage, the Morlet Continuous Wavelet Transform (CWT) was utilized to map the energy evolution to determine the arrival time, oscillation duration, and water recovery time. The validity of the energy decay and local resonance analysis using CWT was then theoretically validated by comparing its behavioral patterns with previous studies, to confirm the consistency of the physical correlation between the earthquake magnitude scale and the duration of wave oscillation in coastal waters.
Through harmonic analysis, a tidal prediction model was successfully developed with high accuracy (RMSE of 0.042–0.072 m at coastal stations and 0.018–0.038 m at offshore stations). The FFT analysis results revealed that the dominant period of the tsunami in offshore waters ranged from 36.0 to 41.2 minutes, which then experienced a period lengthening to 42.8–65.5 minutes upon reaching the coast. The amplitudes at the coast (on the order of 10-2 m) were also significantly amplified compared to those offshore (on the order of 10-3 m) due to the shoaling effect and local resonance. The Morlet CWT mapping showed that Petropavlovsk Station recovered the fastest (28 hours and 46 minutes), whereas the oscillation energy persisted the longest at Kushiro Station (77 hours and 21 minutes). A technical understanding of these characteristics, particularly in determining the arrival time, critical phase duration, and maximum amplitude peak, provides crucial benefits for disaster mitigation systems. This information serves as an important foundation for evacuation teams to determine a safe and effective time window for rescue operations, thereby minimizing the risk to human lives through measured emergency response actions.

Kata Kunci : tsunami, tide gauge, DART buoy, analisis harmonik, Fast Fourier Transform (FFT), Continous Wavelet Transform (CWT) morlet