Inverse Analysis of Experimental Scale Turbidity Currents Using Deep Learning Neural Networks

Abstract

AbstractDespite the importance of turbidity currents in environmental and resource geology, their flow conditions and mechanisms are not well understood. This study proposes and verifies a novel method for the inverse analysis of turbidity currents using a deep learning neural network (DNN) with numerical and flume experiment data sets. Numerical data sets of turbidites were generated with a forward model. Then, the DNN model was trained to find the functional relationship between flow conditions and turbidites by processing the numerical data sets. The performance of the trained DNN model was evaluated with 2,000 numerical test data sets and five experiment data sets. Inverse analysis results on numerical test data sets indicated that flow conditions can be reconstructed from depositional characteristics of turbidites. For experimental turbidites, spatial distributions of grain size and thickness were consistent with the sample values. Concerning hydraulic conditions, flow depth, layer‐averaged velocity, and flow duration were reconstructed with a certain level of deviation. The reconstructed flow depth and duration had percent errors less than 36.0% except for one experiment, which had an error of 193% in flow duration. The flow velocity was reconstructed with percent errors 2.38%–73.7%. Greater discrepancies between the measured and reconstructed values of flow concentration (1.79%–300%) were observed relative to the former three parameters, which may be attributed to difficulties in measuring the flow concentration during experiments. Although the DNN model did not provide perfect reconstruction, it proved to be a significant advance for the inverse analysis of turbidity currents.