Tsunami travel time prediction using neural networks

Barman, Rahul; Kumar, B. Prasad; Pandey, P. C.; Dube, S. K.

doi:10.1029/2006gl026688

Cited by 31 publications

(20 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The model is not intended for application to the Indian Ocean, for which travel time atlases have already been published (Barman et al 2006;Kumar et al 2006). …”

mentioning

confidence: 99%

Method of calculating tsunami travel times in the Andaman Sea region

et al. 2007

View full text Add to dashboard Cite

A new model to calculate tsunami travel times in the Andaman Sea region has been developed. The model specifically provides more accurate travel time estimates for tsunamis propagating to Patong Beach on the west coast of Phuket, Thailand. More generally, the model provides better understanding of the influence of the accuracy and resolution of bathymetry data on the accuracy of travel time calculations. The dynamic model is based on solitary wave theory, and a lookup function is used to perform bilinear interpolation of bathymetry along the ray trajectory. The model was calibrated and verified using data from an echosounder record, tsunami photographs, satellite altimetry records, and eyewitness accounts of the tsunami on 26 December 2004. Time differences for 12 representative targets in the Andaman Sea and the Indian Ocean regions were calculated. The model demonstrated satisfactory time differences (<2 min/h), despite the use of low resolution bathymetry (ETOPO2v2). To improve accuracy, the dynamics of wave elevation and a velocity correction term must be considered, particularly for calculations in the nearshore region.

show abstract

“…The model is not intended for application to the Indian Ocean, for which travel time atlases have already been published (Barman et al 2006;Kumar et al 2006). …”

mentioning

confidence: 99%

Method of calculating tsunami travel times in the Andaman Sea region

et al. 2007

View full text Add to dashboard Cite

show abstract

“…Prediction of ocean parameters using ANN has been carried out by a number of researchers over last two decades (Deo and Naidu 1999;Deo et al 2001;Arena and Puca 2004;Zamini et al 2008;Makarynskyy 2004;Mandal and Prabaharan 2006;Londhe and Panchang 2006;Browne et al 2007;Nitsure and Londhe 2012;Barman et al 2006;Bhaskaran et al 2010;Deshmukh et al 2016). Generally using a simple network architecture, namely, feed-forward or recurrent many of these authors mainly predicted waves in short and long terms based on uni-variate (wave height only) or multi-variate (wave height and wind) time series.…”

Section: Introductionmentioning

confidence: 99%

Review of Applications of Neuro-Wavelet Techniques in Water Flows

Dixit

Londhe²,

Deo

2016

INAE Lett

View full text Add to dashboard Cite

The flow of water over land and oceans is governed by the natural process of hydrological cycle and hence it is highly random and complex. An estimation or prediction of behavior of such flows is thus very difficult and requires continuous research with latest tools of analysis and forecasting. The methods adopted for this purpose range from earlier empirical equations and numerical methods to modern data driven techniques which follow advances in computer technology. The latter coupled with progress in neuro-science had given rise to the method of artificial neural networks (ANNs) which has been applied abundantly since late 1980s to carry out system modeling and prediction type of exercises in water resources and ocean engineering. Despite wide spread applications of ANNs issues such as non-linearity, non-homogenity and non-stationarity of given information continue to elude high accuracy in the results and thus have prompted extension of conventional ANNs to their combination with other pre-or post-processing tools. The wavelet neural network or neuro-wavelet transform (NWT) is a recent effort in this regard to accurately model or predict different types of parameters characterizing river or ocean processes. This paper presents a review of applications of NWT in hydrologic and Oceanic modeling.

show abstract

“…For the Indian coast, major research on tsunami modelling and observations, changes in coastal morphology, changes in coastal ecosystems and palaeo-tsunamis began after the 26 December 2004 event. Various workers studied on tsunamis of the Indian coast [9][10][11][24][25][26][27][28][29] . More details on tsunamis and their effects on coastal morphology and ecosystems are available in ref.…”

mentioning

confidence: 99%

“…Tsunamis, another cause of coastal flooding, are less frequent and mostly occur due to underwater earthquakes or abrupt seafloor slumping 8 . Tsunamis can result in the generation of waves reaching as high as 10-15 m, and inundating a wider swath of the coastal zone, causing total destruction of the coast, for example, the Indonesian tsunami of December 2004 wiped out vast tracts of land along the coasts of Indonesia, Thailand, Sri Lanka and India [9][10][11] . Sea-level rise is another major concern, albeit affecting slowly and steadily, which would result in large-scale submergence of low-lying coastal areas and flat islands; for example, Entire deltaic plains (GangesBrahmaputra Delta) of Bangladesh and adjoining coasts of India; Everglades along south Florida, USA; low-lying islands like Maldives and numerous tropical Pacific islands.…”

mentioning

confidence: 99%

Coastal Inundation Research:An overview of the Process

2017

View full text Add to dashboard Cite

Coastal inundation is the flooding of coastal zone resulting from increased river discharge, spring tides, severe storms, or generation of powerful waves from tectonic activity (tsunami). This article discusses the critical factors that contribute to coastal inundation. Among the probable factors that cause coastal flooding and destruction, storm surge is the most frequent, and hence this article provides a detailed evaluation of the progress made in storm inundation research. Recent advances in coastal inundation modelling include efforts to understand the nonlinear dynamic interaction of near-shore waves, wind and atmospheric pressure with still water sea level and coastal currents, and their combined effects on storm surge along the coast and interaction with coastal morphology. An advanced storm-surge model comprises different modules, viz. an atmospheric component, and two ocean components for surge and wave simulations; these modules are coupled with each other. The nesting of regional coastal model with an ocean-wide model captures the far-field boundary forcing of extreme events that usually originate from the warm open ocean. Even though significant advancements reported on the efficiency and accuracy of storm surge and inundation prediction, further studies are required to understand the nonlinear interaction of storm surge with coastal landforms and their vegetation (land cover). In the context of rising sea level, increased tropical cyclone activity and rapid shoreline change, it is pertinent to evaluate the future flooding risk associated with landfall of tropical cyclones in densely populated coastal cities.

show abstract

Tsunami travel time prediction using neural networks

Cited by 31 publications

References 10 publications

Method of calculating tsunami travel times in the Andaman Sea region

Method of calculating tsunami travel times in the Andaman Sea region

Review of Applications of Neuro-Wavelet Techniques in Water Flows

Coastal Inundation Research:An overview of the Process

Contact Info

Product

Resources

About