Reconstruction of flow conditions from 2004 Indian Ocean tsunami deposits at the Phra Thong island using a deep neural network inverse model

Abstract

Abstract. The 2004 Indian Ocean tsunami caused significant economic losses and a large number of fatalities in the coastal areas. The estimation of tsunami flow conditions using inverse models has become a fundamental aspect of disaster mitigation and management. Here, a case study involving the Phra Thong island, which was affected by the 2004 Indian Ocean tsunami, in Thailand was conducted using inverse modeling that incorporates a deep neural network (DNN). The DNN inverse analysis reconstructed the values of flow conditions such as maximum inundation distance, flow velocity and maximum flow depth, as well as the sediment concentration of five grain-size classes using the thickness and grain-size distribution of the tsunami deposit from the post-tsunami survey around Phra Thong island. The quantification of uncertainty was also reported using the jackknife method. Using other previous models applied to areas in and around Phra Thong island, the predicted flow conditions were compared with the reported observed values and simulated results. The estimated depositional characteristics such as volume per unit area and grain-size distribution were in line with the measured values from the field survey. These qualitative and quantitative comparisons demonstrated that the DNN inverse model is a potential tool for estimating the physical characteristics of modern tsunamis.

Understanding flow characteristics from tsunami deposits at Odaka, Joban Coast, using a deep neural network (DNN) inverse model

“…Reconstructed flow conditions closely approximated the values measured in situ from the Sendai Plain. The method was also applied to the 2004 Indian Ocean tsunami at the island of Phra Thong in Thailand (Mitra et al, 2021a). The reconstructed inundation distance, velocity, and flow depth fell in the range of the actual measurements.…”

“…Several inverse and for-ward models have been proposed to reconstruct flow conditions (Jaffe et al, 2012;Li et al, 2012;Sugawara and Goto, 2012;Johnson et al, 2016;Yoshii et al, 2018). Recently Mitra et al (2020Mitra et al ( , 2021a proposed a new one-dimensional inverse model using the deep neural network (DNN) and predicted reasonable flow conditions of the 2011 Tohoku-oki tsunami and the 2004 Indian Ocean tsunami from measured grain-size distributions and thickness at different locations, such as the Sendai Plain and island of Phra Thong, Thailand. This inverse model utilizes a forward model that incorporates the non-uniform, unsteady transport of suspended sediments with turbulent mixing.…”

Development of black ice prediction model using GIS-based multi-sensor model validation

“…Parameter (w n , b n ) optimisation of the DNN model was performed using the evolution algorithm. The deep learning method based on the existing gradient descent and back propagation is quicker by dynamic programming but it is not suitable for application to the system dynamics environment with dynamic characteristics owing to the time change (Mitra et al, 2021). Therefore, the parameters of the DNN are optimised with the evolution algorithm, which is versatile for models with various characteristics such as dynamic change.…”

A Medium‐Sized Paleo‐Tsunami Reconstruction by a Deep Neural Network Processing Sedimentary Deposits

“…The combination of forward 2D models with deep neural network (DNN) models has been successfully used to reconstruct the characteristics of the 2011 Tohoku‐oki (Mitra et al., 2020; Naruse & Abe, 2017), 2004 Indian Ocean (Mitra et al., 2021) and other tsunamis (e.g., Mitra et al., 2023). The direct observation of the events and the rapid sampling of the tsunami deposits validated the model results.…”