ABSTRAKSeismik menjadi salah satu metode yang menggunakan akustik kelautan untuk mengetahui struktur di bawah permukaan laut. Penampang seismik hasil akuisisi data biasanya terdapat noise yang merupakan gelombang atau sinyal yang tidak dikehendaki dan harus dibuang dari data utama. Untuk mengidentifikasi noise pada data seismik tersebut, dalam penelitian ini digunakan fungsi FFT atau Fast Fourier Transform. Penelitian ini bertujuan untuk memisahkan sinyal akustik objek dan noise pada data seismik dengan menggunakan fungsi FFT dan metode Bandpass Filter pada data SeismicSingle Channel 2D di Perairan Krakatau. Akuisisi data seismik yang digunakan pada penelitian ini menggunakan peralatan single-channel high resolution Seismic berupa EGG Multi Electrode Sparker dan hidrofon 16 elemen merek EGG sebagai receiver. Pengolahan data seismik dalam bentuk .segy pada penelitian ini menggunakan perangkat lunak Seisee dan Matlab. Frekuensi tepi optimal pada penelitian ini berkisar antara 7 Hz hingga 29 Hz, dan frekuensi 2400 Hz untuk tepi atas. Frekuensi cut-off optimal bergantung pada nilai frekuensi pada tiap trace dengan nilai frekuensi cut-off bawah menggunakan frekuensi pada kisaran 200 Hz hingga 400 Hz dan nilai frekuensi cut-off atas menggunakan frekuensi 1500 Hz. Kata kunci: Bandpass Filter, FFT, Frekuensi, Noise, Seismik.ABSTRACTSeismic is one method that uses marine acoustics principle to determine the structure below the sea surface. Seismic cross-sections resulting from data acquisition usually contain noise which is an unwanted wave or signal and must be removed from the main data. To identify noise in the seismic data, this study used the Fast Fourier Transform. This study aimed to separate the object's acoustic signal and noise in seismic data by using the FFT function and Bandpass Filter method on single channel 2D seismic data in Krakatau waters. Seismic data acquisition used in this study was a single-channel high resolution seismic form of EGG Multi Electrode Sparker and EGG 16 elements hydrophone as receiver. Seismic data processing in the form of *.segy in this study were Seisee and Matlab software. The optimal edge frequency in this study ranges from 7 Hz to 29 Hz, and the frequency was 2400 Hz for the top edge. The optimal cut-off frequency depends on the frequency value in each trace, with the lower cut-off frequency value using a frequency in the range of 200 Hz to 400 Hz and the upper cut-off frequency value using a frequency of 1500 Hz.Keywords: Bandpass Filter, FFT, Frequency, Noise, Seismic

The finless porpoise (Neophocaena asiaeorientalis) is a species frequently observed in Korean coastal waters that remains highly vulnerable to bycatch and habitat disturbance. To develop effective conservation strategies, it is essential to understand their acoustic behavior and environmental preferences. This study utilized Passive Acoustic Monitoring to investigate the acoustic characteristics and activity patterns of finless porpoises in coastal waters near cage aquaculture farms from September to October 2021. A total of 372,707 clicks and 175,119 click trains were identified. Mean acoustic parameters were peak frequency 122.0 ± 11.1 kHz, 3 dB bandwidth 15.4 ± 12.0 kHz, 10 dB bandwidth 45.3 ± 16.1 kHz, and ICI 39.0 ± 34.8 ms. Click activity exhibited a distinct diel pattern, with significantly higher activity during the early morning and nighttime. Generalized Additive Model analysis revealed significant non-linear relationships between click activity and tide, temperature, salinity, and hour. Specifically, click activity decreased with rising temperatures and lower salinity, while the effect of tide was relatively limited. These findings provide critical baseline data for the development of acoustic deterrent devices tailored to Korean marine environments and contribute to the management of bycatch mitigation and coastal ecosystem conservation.