The 2025 Kamchatka tsunami (from the Mw 8.8 megathrust earthquake on 29 July 2025 off the Kamchatka Peninsula) produced long- lasting, coherent oscillations across the Pacific that persisted for many hours, and analysis of its dominant periods (its “rhythm”) has potential value for improving tsunami warnings.

On 29 July 2025 (23:24:52 UTC), an Mw 8.8 earthquake ruptured along the Kuril- Kamchatka Trench near Petropavlovsk- Kamchatsky. It generated a trans- Pacific tsunami with local run- ups reaching up to ~19 m near the source (and isolated higher splash values in steep valleys), while far- field waves were generally more moderate than initially feared (often ≤1- 2 m in many places, with some higher local values).

The event reactivated parts of the same margin as the larger 1952 Mw ~9.0 earthquake but with differences in rupture depth and slip distribution that limited the far- field tsunami energy relative to 1952.

A key observational feature was the prolonged “ringing” of the Pacific basin: wave trains continued for many hours (reports of multi- hour to multi- day activity at some coastal and deep- ocean sites due to reflections, refraction, and multi-path arrivals). Deep- ocean DART buoys recorded clear signals that could be analyzed spectrally.

In a numerical modeling and analysis study led by Amin Rashidi with co-authors including Denys Dutykh (and Mohammad Mokhtari), researchers simulated the tsunami using an earthquake source model and compared results to six observed DART waveforms.

The simulations matched arrival times and amplitudes reasonably well. Fourier and wavelet analyses of the records showed a coherent long- period signal in the ~30- 70 min band, with station- specific peaks near 32, 43, and 64 min. These periods relate to the source dimensions and the basin- scale propagation characteristics.

The authors note that identifying such dominant periods supports tsunami hazard assessments for Kamchatka- type megathrust events and highlights the practical value of spectral and period analysis for Pacific warning systems. Knowing the characteristic “rhythm” of the wave train can help refine forecasts of wave heights, arrival of maximum energy (which often comes later than the first wave), and the duration of elevated hazard, complementing traditional amplitude and arrival- time predictions.

Related work on the same event has examined rupture complexity (deeper slip limiting far- field energy), ionospheric and atmospheric signatures detectable via GNSS, SWOT satellite altimetry snapshots, and comparisons with the 1952 tsunami, all contributing to better understanding of source- to- basin effects.

In short, the long- lasting, spectrally coherent oscillations of the 2025 event provide both a validation benchmark for models and a potential operational cue (period content) that warning centers can incorporate to improve the quality and duration of alerts.

DART stands for Deep- ocean Assessment and Reporting of Tsunamis. It is a real- time deep- ocean tsunami detection and monitoring network operated primarily by NOAA (U.S. National Oceanic and Atmospheric Administration), with contributions from other countries.

How a DART system works

Each DART station has two main components:

• A Bottom Pressure Recorder (BPR) anchored on the seafloor (typically at depths of 1,000- 6,000 m). It measures changes in water pressure caused by the passage of a tsunami with very high sensitivity (down to ~1 cm of sea- surface height change).
• A surface buoy moored nearby that communicates with the BPR via acoustic signals and relays the data to shore via satellite (Iridium or GOES).

Operating modes:

• Standard mode: Routine low- rate transmissions (e.g., hourly summaries of 15-minute averages) for system health monitoring.
• Event mode: Automatically triggered when the BPR detects a pressure change exceeding a threshold (or manually triggered by warning centers). It switches to high-frequency sampling and rapid transmission (15- second or 1- minute data) for several hours.

These deep- ocean measurements are free of coastal amplification and shelf effects, so they provide a clean signal of the tsunami as it propagates across the open ocean.

Network size and coverage

• The U.S. operates 39 DART systems, strategically placed mainly in the Pacific, with additional coverage in the Atlantic, Gulf of Mexico, and Caribbean.
• Other countries maintain additional stations, bringing the global network to more than 60 systems.
• Stations are positioned in regions with a history of generating destructive tsunamis to give the earliest possible detection far from shore.

Data flow in near real time to NOAA’s two Tsunami Warning Centers (Pacific Tsunami Warning Center and National Tsunami Warning Center), where they are used to refine source estimates, update forecast models (e.g., MOST/SIFT), and improve warnings of amplitude, arrival time, and duration.

Role in the 2025 Kamchatka event

In the July 2025 Mw 8.8 Kamchatka tsunami, multiple DART stations (including ones relatively near the source such as 21416) recorded clear signals. These waveforms were central to the Rashidi, Dutykh & Mokhtari (2026) study: the authors compared numerical simulations against six DART records and performed Fourier/wavelet analysis that revealed the coherent 30- 70 min dominant periods (with peaks near 32, 43 and 64 min). Early DART detections also helped constrain rupture extent and improve real- time forecasts during the event.

Current status and modernization

NOAA continues to operate and maintain the array. Equipment modernization (funded in part by recent infrastructure investments) is underway to improve reliability and data availability, with replacements planned through the late 2020s. Historical and real-time data are archived and publicly available through NOAA’s National Data Buoy Center (NDBC) and National Centers for Environmental Information (NCEI).

In short, the DART network is the primary deep- ocean “eyes” of modern tsunami warning systems.

It supplies the critical open- ocean measurements that convert seismic information into quantitative forecasts of what is actually heading toward coastlines.

Fourier analysis and wavelet analysis are complementary mathematical tools used to study the frequency content (the “rhythm” or dominant periods) of time-series signals such as the sea- level records from DART buoys.

1. Fourier Analysis

Fourier analysis decomposes a signal into a sum of pure sinusoidal waves of different frequencies (or periods).

• Core idea: Any reasonably well- behaved time series can be expressed as a combination of sines and cosines of various frequencies, each with its own amplitude and phase.
• The result is a power spectrum (or periodogram) that shows how much energy is present at each frequency and period.
• In tsunami studies it answers the question: “What are the dominant periods present in the entire record?”

Strengths:

• Excellent for identifying the overall dominant periods in a stationary (statistically constant) signal.
• Computationally efficient (via the Fast Fourier Transform- FFT).

Limitations:

• It loses all time information. You know which periods are strong, but not when they occurred or how they evolved.
• It assumes the signal’s statistical properties do not change much over the whole record.

In the 2025 Kamchatka paper, Fourier analysis of the DART records revealed a coherent energy band roughly between 30 and 70 minutes, with clear peaks near 32, 43 and 64 min.

2. Wavelet Analysis

Wavelet analysis is a time-frequency method. It examines the signal with a set of “wavelets”, localized oscillating functions that can be stretched (to capture long periods) or compressed (to capture short periods) and slid along the time axis.

• Core idea: Instead of one global spectrum, you obtain a scalogram (or time- frequency map) showing how the strength of each period changes over time.
• Common choice for geophysical signals is the Continuous Wavelet Transform (CWT), often using a Morlet wavelet (a complex sinusoid modulated by a Gaussian envelope).

Strengths:

• Reveals when particular periods appear, strengthen, or decay.
• Handles non- stationary signals (tsunamis are highly non- stationary: the first arrival is different from later reflected or trapped waves).
• Useful for tracking the evolution of the wave train as it propagates and interacts with bathymetry.

Limitations:

• Slightly more computationally intensive.
• There is a trade- off between time resolution and frequency resolution (controlled by the choice of wavelet and scales).

In the same Kamchatka study, the wavelet analysis confirmed that the 30- 70 min energy was coherent and persistent, and showed how the station- specific peaks evolved after the main arrival, information that a pure Fourier spectrum cannot provide.

How they are typically combined for tsunami records

1. Pre-process the DART time series (remove tides, de- spike, filter if needed).
2. Apply Fourier analysis → global dominant periods.
3. Apply wavelet analysis → time- frequency evolution of those periods.
4. Interpret the results physically (source dimensions, dispersion, reflections, shelf resonances, etc.).

This combination is why the paper could confidently state that the tsunami “generated a coherent 30- 70 min signal” with specific peaks, and why such spectral information is valuable for improving tsunami warnings: knowing the characteristic rhythm helps forecast how long elevated water levels will last and when the largest waves are likely to arrive.

Primarily tsunami modeling of the M_w 8.8 Kamchatka Peninsula earthquake on July 29, 2025

An Mw 8.8 megathrust earthquake along the Kuril- Kamchatka Trench near the eastern coast of Petropavlovsk- Kamchatsky, Russia, on 29 July 2025 triggered a Pacific- wide tsunami with a reported maximum wave height of 19 m.

The authors:

• Performed numerical modeling of the tsunami.
• Analyzed six observed tsunami waveforms recorded at DART (Deep- ocean Assessment and Reporting of Tsunamis) gauges.
• Conducted Fourier and wavelet analyses to identify the dominant period range of the tsunami.

Main findings:

• Simulated waveforms show a good fit with the recorded sea- level data.
• Wavelet, Fourier analysis reveals a coherent 30- 70 min signal, with station- specific peaks at 32, 43, and 64 min across the DART stations.
• Comparison of maximum simulated amplitudes and arrival times shows reasonable agreement with observations, supporting the reliability of the earthquake source model used.
• The results confirm that Kamchatka megathrust events can generate basin- wide tsunamis carrying coherent long-period energy.
• This supports existing tsunami hazard assessments and highlights the value of dominant- period (“rhythm”) analysis for Pacific warning systems.

The long- lasting coherent oscillations (the “Pacific ringing” for many hours) and the characteristic periods are presented as useful diagnostic information that can help refine forecasts of wave- train duration, timing of maximum energy, and overall hazard persistence beyond first- arrival predictions.

The full text is behind a Springer paywall and subscription; only the abstract and some front- matter (figures list, references) is freely visible on the page.

Journal information: Natural Hazards, Volume 122, article 450 (2026)

DOI: 10.1007/s11069-026-08198-3

Published: 14 May 2026

Authors: Amin Rashidi,
Denys Dutykh &
Mohammad Mokhtari 

Abstract

An M_w 8.8 megathrust earthquake along the Kuril–Kamchatka Trench near the eastern coast of Petropavlovsk-Kamchatsky, Russia, on July 29, 2025, triggered a tsunami across the Pacific Ocean with a reported maximum wave height of 19 m. In this study, we perform numerical modeling of the tsunami generated by this event and analyze six observed tsunami waveforms recorded at DART tide gauges. Additionally, Fourier and wavelet analyses are conducted on these records to identify the dominant period range of the tsunami. Our simulated waveforms show a good fit with the recorded sea level data. Wavelet–Fourier analysis shows the tsunami generated a coherent 30–70 min signal, with station-specific peaks at 32, 43, and 64 min across the DART stations. The comparison of maximum simulated tsunami wave amplitudes and arrivals shows reasonable agreement with recorded data, implying that the earthquake source model is capable of accurately reproducing the tsunami. These findings confirm that Kamchatka megathrust events can generate basin-wide tsunamis with coherent long-period energy, supporting existing tsunami hazard assessments, and highlighting the value of dominant period analysis for Pacific warning systems.