Search

Tsunamis from landslide and volcanic sources have been responsible for significant destruction and fatalities worldwide as evidenced most recently during the January 2022 Tonga volcanic tsunami (Heidarzadeh et al., 2022: https://doi.org/10.1016/j.oceaneng.2022.112165). Indonesia is a hot spot for such tsunamis from landslide and volcanic sources as the region suffered from destructive events in the past, such as the 1883 tsunami following the Krakatau eruption which costed at least 36,000 lives. More recently the region was struck by the 2018 Anak Krakatau volcanic tsunami with approximately 450 deaths, and the 2018 Palu (Sulawesi) tsunami with more than 4,000 casualties. Therefore, it is vital to further study the generation potential and mechanisms of such tsunamis and to improve hazard knowledge base.Here, we study three recent tsunamis in Indonesia, two of which occurred following an earthquake while the other one occurred following a volcanic eruption. All three have a landslide component in their sources: the June 2021 Seram Island tsunami (earthquake), the December 2018 Palu tsunami (earthquake), and the December 2018 Anak Krakatau tsunami (volcanic eruption).A tsunami was observed on 16th June 2021 in Seram Island following an Mw 5.9 earthquake. The tsunami amplitude was approximately 50 cm at Tehoru tide gauge whereas two other stations showed amplitudes of less than 4 cm. Such a relatively large tsunami (50 cm) is unexpected from a normal-faulting Mw 5.9 earthquake. We hypothesize that that a secondary source (i.e., a landslide) was involved. We applied tsunami modelling and source analysis to examine this hypothesis. Tsunami simulations confirmed that that the earthquake could only have contributed to a few centimeters of the tsunami and thus cannot reproduce the 50 cm waves. However, we could reproduce the tsunami observations using a landslide source. For more information see here: https://doi.org/10.1785/0120210274. Regarding the September 2018 Palu tsunami, it is now commonly accepted that a submarine landslide should have most likely contributed to the tsunami generation in addition to the earthquake. However, the nature of the landslide whether submarine or subaerial, and the contribution of the two sources are not clear. We propose a novel dual landslide-earthquake source that explains most of the observation of the 2018 Palu event. Our dual model comprises the USGS earthquake model (length = 264 km, width = 37 km, slip = 0 – 8.5 m) combined with a submarine landslide with a length of 1.0 km, a width of 2.0 km, and a thickness of 80.0 m. For more information see here: https://doi.org/10.1080/21664250.2022.2122293. For the December 2018 Anak Krakatau tsunami, we present the results of our field surveys. We surveyed 29 locations and measured tsunami runups from 0.9 m to 5.2 m, tsunami heights from 1.4 to 6.3 m, and inundation distances from 18 to 212 m. For more information, see here: https://doi.org/10.1007/s00024-020-02587-w.We also discuss future directions towards expanding our limited understanding of tsunamis from landslide and volcanic sources in Indonesia which are often unpredictable and deadly. This research is funded by The Royal Society (UK), grant number CHL/R1/180173.

The Seafloor Observation Network for Earthquakes and Tsunamis along the Japan Trench (S-net) is presently the world’s largest network of ocean bottom pressure sensors for real-time tsunami monitoring. This paper analyzes the efficacy of such a vast system in tsunami forecasting through exhaustive synthetic experiments. We consider 1500 hypothetical tsunami scenarios from megathrust earthquakes with magnitudes ranging from Mw 7.7–9.1. We employ a stochastic slip model to emulate heterogeneous slip patterns on specified 240 subfaults over the plate interface of the Japan Trench subduction zone and its vicinity. Subsequently, the associated tsunamis in terms of maximum coastal tsunami heights are evaluated along the 50-m isobath by means of a Green’s function summation. To produce tsunami forecasts, we utilize a tsunami inversion from virtually observed waveforms at the S-net stations. Remarkably, forecasts accuracy of approximately 99% can be achieved using tsunami data within an interval of 3 to 5 min after the earthquake (2-min length), owing to the exceedingly dense observation points. Additionally, we apply an optimization technique to determine the optimal combination of stations with respect to earthquake magnitudes. The results show that the minimum requisite number of stations to maintain the accuracy attained by the existing network configuration decreases from 130 to 90 when the earthquake size increases from Mw 7.7 to 9.1.

We analyze the regional tsunami hazard along the Sea of Japan coast associated with 60 active faults beneath the eastern margin of the Sea of Japan. We generate stochastic slip distribution using a Monte Carlo approach at each fault, and the total number of required earthquake samples is determined based on convergence analysis of maximum coastal tsunami heights. The earthquake recurrence interval on each fault is estimated from observed seismicity. The variance parameter representing aleatory uncertainty for probabilistic tsunami hazard analysis is determined from comparison with the four historical tsunamis, and a logic-tree is used for the choice of the values. Using nearshore tsunami heights at the 50 m isobath and an amplification factor by the Green’s law, hazard curves are constructed at 154 locations for coastal municipalities along the Sea of Japan coast. The highest maximum coastal tsunamis are expected to be approximately 3.7, 7.7, and 11.5 m for the return periods of 100-, 400-, and 1000-year, respectively. The results indicate that the hazard level generally increases from southwest to northeast, which is consistent with the number and type of the identified fault systems. Furthermore, the deaggregation of hazard suggests that tsunamis in the northeast are predominated by local sources, while the southwest parts are likely affected by several regional sources.

SUMMARYWe propose a method for defining the optimal locations of a network of tsunameters in view of near real-time tsunami forecasting using sea surface data assimilation in the near and middle fields, just outside of the source region. The method requires first the application of the empirical orthogonal function analysis to identify the potential initial locations, followed by an optimization heuristic that minimizes a cost-benefit function to narrow down the number of stations. We apply the method to a synthetic case of the 2015 Mw8.4 Illapel Chile earthquake and show that it is possible to obtain an accurate tsunami forecast for wave heights at near coastal points, not too close to the source, from assimilating data from three tsunameters during 14 min, but with a minimum average time lag of nearly 5 min between simulated and forecasted waveforms. Additional tests show that the time lag is reduced for tsunami sources that are located just outside of the area covered by the tsunameter network. The latter suggests that sea surface data assimilation from a sparse network of stations could be a strong complement for the fastest tsunami early warning systems based on pre-modelled seismic scenarios.

This study proposes the detection of large offshore tsunamis by airborne radar. Altimeter measurements were carried out south of Japan by an observational aircraft along the Jason satellite tracks crossing over the Kuroshio Current. A nadir-pointing frequency-modulated continuous wave (FM-CW) radar and a global navigation satellite system (GNSS) based positioning system were concurrently used to measure the sea surface height (SSH). The SSH was validated against the Jason-2 and Jason-3 satellite altimeter SSH. After the removal of the geoid and tidal signals, the observation error was less than 10 cm. This is sufficiently small to capture offshore tsunamis within 10 min after the genesis. This was demonstrated by virtual airborne radar-detected SSH from a numerically simulated hypothetical large offshore tsunami south of Japan. The feasibilities and limitations of utilizing commercial airplanes with altimeters to create a dense tsunami observation network are discussed. The suggested methodology can be adapted to various emerging airborne platforms such as unmanned aerial vehicles. Moreover, a dense coverage of SSH observations is invaluable from an oceanographic point of view.

An efficient and cost‐effective near‐field tsunami warning system is crucial for coastal communities. The existing tsunami forecasting system is based on offshore Deep‐Ocean Assessment and Reporting of Tsunamis and Global Navigation Satellite System (GNSS) buoys which are not affordable for many countries. A potential cost‐effective solution is to utilize position data from ships traveling in coastal and offshore regions. In this study, we examine the feasibility of using ship‐borne GNSS data in tsunami forecasting. We carry out synthetic experiments by applying a data assimilation (DA) method with ship position (elevation and velocity) data. Our findings show that the DA method can recover the reference model with high accuracy if a dense network of ship elevation data is used. However, the use of ship velocity data alone is unable to recover the reference model. In addition, we carried out sensitivity studies of the DA method to the ship spatial distribution. We find that a 20 km gap between the ships works well in terms of accuracy and computational time for the example source model that we explored. The highest accuracy is obtained when data from a sufficient number of ships traveling in and around the tsunami source area are available.

Being situated on the major subduction zones in conjunction with a considerable number of submarine active faults and coastal volcanoes, Japan has a long history of catastrophic tsunami events. Consequently, enormous efforts in disaster mitigation, particularly in relation with tsunami hazards, have been made across the country. It is of our interest to review the developments of tsunami observing systems in Japan, which may lead to a global implication beyond national boundaries. In this paper, we first discuss, in general, the evolution of past to present tsunami observing systems available around the territory of Japan. More specifically, we identify the existing offshore observational networks that are mainly consisted of cabled ocean bottom pressure gages and global navigation satellite system buoys, and briefly analyze their performance and viability in the long-term future. In that context, we also appraise the potential of emerging technologies in the offshore tsunami detection leveraging unconventional platforms such as commercial ships and airplanes, which have recently been introduced by several studies in Japan.

Sustaining an accurate, timely, and global tsunami forecast system remains a challenge for scientific communities. To this end, various viable geophysical monitoring devices have been deployed. However, it is difficult to implement new observation networks in other regions and maintaining the existing systems is costly. This study proposes a new and complementary approach to monitoring the tsunami using existing moving platforms. The proposed system consists of a radar altimeter, Global Navigation Satellite Systems receiver, and an adequate communication link on airborne platforms such as commercial airplanes, drones, or dedicated high-speed aircraft, and a data assimilation module with a deterministic model. We demonstrated, through twin-data experiment, the feasibility of the proposed system in forecasting tsunami at the Nankai Trough of Japan. Our results demonstrated the potential of an airborne tsunami observation as a viable future technology through proxy observations and rigorous numerical experiments. The wide coverage of the tsunamigenic regions without a new observation network is an advantage while various regulatory constraints need to be overcome. This study offered a novel perspective on the developments in tsunami detection and forecasting technology. Such multi-purpose observation using existing platforms provides a promising and practical solution in establishing sustainable observational networks.

© 2021 The Authors. The July 2020 Mw7.8 Alaska tsunamigenic earthquake was a fresh call for potential large tsunamis associated with the Aleutian subduction zone. The second largest ever-recorded earthquake worldwide (Mw 9.2) occurred in this zone in 1964 indicating its massive earthquake and tsunami risk. Our analysis of the July 2020 tsunami revealed that it involves very long period waves (51–64 min) which is unusual for an Mw7.8 earthquake. The tsunami coastal amplitude was small (~0.5 m) which is much smaller than that usually expected from a tsunamigenic earthquake of this size. Here, through numerical simulations and spectral analyses, we explain the ultra-long period and small amplitude waves of the tsunami. Our analysis using an analytical equation showed that the ultra-long period of the tsunami (51–64 min) can be reproduced using the shallow water depth around the source region (100–200 m) and the length of the coseismic deformation area (~100 km). By comparing the coastal amplitude of this event with four other similar-size and similar-mechanism earthquakes (thrust Mw7.8), we attribute the relatively small coastal amplitude of this tsunami to the deep focal depth of the earthquake (28 km) and the extremely shallow water around the source region. Royal Society (the United Kingdom), grant number CHL\R1\180173

The future Nankai Trough tsunami is one of the imminent threats to the Japanese coastal communities that could potentially cause a catastrophic event. As a part of the countermeasure efforts for such an occurrence, this study analyzes the efficacy of combining tsunami data assimilation (DA) and waveform inversion (WI). The DA is used to continuously refine a wave field model whereas the WI is used to estimate the tsunami source. We consider a future scenario of the Nankai Trough tsunami recorded at various observational systems, including ocean bottom pressure (OBP) gauges, global positioning system (GPS) buoys, and ship height positioning data. Since most of the OBP gauges are located inside the source region, the recorded tsunami signals exhibit significant offsets from surface measurements due to coseismic seafloor deformation effects. Such biased data are not applicable to the standard DA, but can be taken into account in the WI. On the other hand, the use of WI for the ship data may not be practical because a considerably large precomputed tsunami database is needed to cope with the spontaneous ship locations. The DA is more suitable for such an observational system as it can be executed sequentially in time and does not require precomputed scenarios. Therefore, the combined approach of DA and WI allows us to concurrently make use of all observational resources. Additionally, we introduce a bias correction scheme for the OBP data to improve the accuracy, and an adaptive thinning of observations to determine the efficient number of observations.

We apply a genetic algorithm (GA) to find the optimized unit sources using dispersive tsunami synthetics to estimate the tsunami source of the 2012 Haida Gwaii earthquake. The optimal number and distribution of unit sources gives the sea surface elevation similar to that from our previous slip distribution on a fault using tsunami data, but different from that using seismic data. The difference is possibly due to submarine mass failure in the source region. Dispersion effects during tsunami propagation reduce the maximum amplitudes by up to 20% of conventional linear long wave propagation model. Dispersion effects also increase tsunami travel time by approximately 1 min per 1,300 km on average. The dispersion effects on amplitudes depend on the azimuth from the tsunami source reflecting the directivity of tsunami source, while the effects on travel times depend only on the distance from the source.

This study applied an extreme learning machine to produce rapid forecasts of tsunami waveforms in coastal areas using tsunami signals recorded at specified locations. The remarkable training speed of the algorithm means that it can run in real-time, and therefore it is suitable for early warning systems in near-field tsunami events. Additionally, as a universal function approximator, the proposed method can capture nonlinearities exhibited by the tsunami. Therefore, it provides advantages over the standard inversion analysis used in many existing studies, which is typically developed under a linear assumption. We applied the proposed method to the 2011 Tohoku earthquake tsunami. Our results demonstrate that the proposed method is more accurate and does not significantly increase the computing time, when compared with the standard method. Furthermore, our model uncertainty analysis proves that the method is robust and reliable, despite its dependency on the random input weights and biases (the forecasts from several consecutive runs showed insignificant variability).

We propose a method for accurately estimating the initial tsunami source. Our technique is independent of the earthquake parameters, because we only use recorded tsunami waveforms and an auxiliary basis function, instead of a fault model. We first use the measured waveforms to roughly identify the source area using backward propagated travel times, and then infer the initial sea surface deformation through inversion analysis. A computational intelligence approach based on a genetic algorithm combined with a pattern search was used to select appropriate least squares model parameters and time delays. The proposed method significantly reduced the number of parameters and suppressed the negative effect of regularization schemes that decreased the plausibility of the model. Furthermore, the stochastic approach for deriving the time delays is a more flexible strategy for simulating actual phenomena that occur in nature. The selected parameters and time delays increased the accuracy, and the model's ability to reveal the underlying physics associated with the tsunami-generating processes. In this paper, we applied the method to the 2011 Tohoku-Oki tsunami event and examined its effectiveness by comparing the results to those using the conventional method.

Recently, there are numerous tsunami observation networks deployed in several major tsunamigenic regions. However, guidance on where to optimally place the measurement devices is limited. This study presents a methodological approach to select strategic observation locations for the purpose of tsunami source characterizations, particularly in terms of the fault slip distribution. Initially, we identify favorable locations and determine the initial number of observations. These locations are selected based on extrema of empirical orthogonal function (EOF) spatial modes. To further improve the accuracy, we apply an optimization algorithm called a mesh adaptive direct search to remove redundant measurement locations from the EOF-generated points. We test the proposed approach using multiple hypothetical tsunami sources around the Nankai Trough, Japan. The results suggest that the optimized observation points can produce more accurate fault slip estimates with considerably less number of observations compared to the existing tsunami observation networks.

A measurement buoy with attached sensors has been deployed at our study area to monitor hydrodynamics, water properties, and water quality conditions. High-resolution temporal data have been collected and streamed into an online system that is accessible in nearly real-time. However, in certain circumstances the sensors may fail to provide continuous and high quality data. This results in gaps or corrupted values. The aim of this study was to reconstruct the faulty values. This paper proposes a method based on a data-driven model, using an Artificial Neural Network combined with a Genetic Algorithm to generate a synthetic data series. The generated data can be used as a patch for the incomplete measured data. Additional improvements were achieved by removing seasonal patterns from the original time series using a wavelet decomposition prior to the data-driven model training process. Comparisons with a standard missing-data imputation method using the Kohonen self-organizing map were made to further asses the performance of the proposed data-driven model. The algorithm was applied to water temperature data, but the same approach is applicable to other parameters of interest.

Determination of sea surface deformation generated by earthquakes is crucial to the success of tsunami modeling. Using waveforms recorded at measurement stations and assuming that the rupture velocity is much faster than the tsunami wave celerity, sea surface deformation caused by a tsunamigenic earthquake can be inferred through an inversion operation using the Green's function technique. However, this inversion method for tsunami waveforms possesses a limitation, in that the inverse matrix does not always exist because of the non-uniqueness of the solution. In addition to the large number of unknown parameters, which might produce many local optima on the misfit function measure, the search towards optimality is confined by the uniform distance of unit sources used in the regular Green's function. This study proposes a new method to both optimize the determination of the unknown parameters and introduce a global optimization method for tsunami waveform inversion. The method has been tested using an artificial tsunami source with real bathymetry data. A significant improvement is achieved by stochastically searching for an optimal distribution of unit source locations prior to the inversion.

The transport and fate of admixtures at coastal zones are driven, or at least modulated, by currents. In particular, in tide-dominated areas due to higher near-bottom shear stress at strong currents, sediment concentration and turbidity are expected to be at maximum during spring tide, while algal growth rate likely is peaking up at slack currents during neap tide. Varying weather and atmospheric conditions might modulate the said dependencies, but the water quality pattern still is expected to follow the dominant tidal cycle. As tidal cycling could be predicted well ahead, there is a possibility to use water quality and hydrodynamic high-resolution data to learn past dependencies, and then use tidal hydrodynamic model for nowcasting and forecasting of selected water quality parameters. This paper develops data driven models for nowcasting and forecasting turbidity and chlorophyll-a using Artificial Neural Network (ANN) combined with Genetic Algorithm (GA). The use of GA aims to automate and enhance ANN designing process. The training of the ANN model is done by constructing input–output mapping, where hydrodynamic parameters act as an input for the network, while turbidity and chlorophyll-a are the corresponding outputs (desired target). Afterward, the prediction is carried out only by employing computed water surface elevation as an input for the trained ANN model. The proposed data driven model has successfully revealed complex relationships and utilized its experiential knowledge acquired from the training process for facilitating the subsequent use of the data driven model to yield an accurate prediction.

Tsunamis are extraordinary occurrences that are difficult to identify; most of the incidents have no recorded predictions and tsunamis are generally infrequent events with poor data acquisition. The development of a tsunami database system has become important for improving the management of information with regard to a tsunami early warning system for vulnerable communities along coastline areas. Numerical modelling is usually employed to simulate the wave height and travel time of a wave arriving at a coastline area. However, numerical modelling for tsunami prediction is too time-consuming to be useful as an early warning system for the mitigation of tsunami-related damage and loss. Therefore, this model was used to develop a tsunami database system, based on hypothetical data, in order to develop a recognition pattern for a neural network learning process that will improve the speed and accuracy of tsunami prediction. To improve the accuracy of numerical modelling and observation, an adjustment was established for an advanced training process which used a generalised regression neural network. In other words, the training and testing datasets were obtained by correcting near-field tsunami numerical models from hypothetical earthquakes. The case study was performed on part of the Southern Pangandaran coastline in West Java, Indonesia.

The characteristics of a tsunami is largely determined by its initial source, which can be defined as a perturbation to the equilibrium state of the ocean hydrodynamic. For tsunamis generated by earthquakes, the initial source (as a form of sea surface deformation) is driven by a coseismic vertical displacement of the seabed. Assuming that the rupture process occurs instantaneously, the sea surface deformation can be inferred from recorded tsunami waveforms. Tsunami waveform inversion using the Green’s function technique with least squares optimization is the most widely used method. However, this method can sometimes suffer from ill-posed problems, because of the non-uniqueness of the inversion results. This study proposes a genetic algorithm that helps the traditional least squares method find an optimum solution. We applied the method to an artificial tsunami source with a complex profile. We achieved significant improvements by determining the optimum spatial distribution of the unit sources (the Green’s function) prior to the inversion. In contrast with the regular Green’s function that has equidistant unit sources, our method generates a random spatial distribution. This random characteristic leads to a better approximation of the initial sea surface deformation.