TY - JOUR
T1 - Estimation of groundwater storage from seismic data using deep learning
AU - Lähivaara, Timo
AU - Malehmir, Alireza
AU - Pasanen, Antti
AU - Kärkkäinen, Leo
AU - Huttunen, Janne M.J.
AU - Hesthaven, Jan S.
PY - 2019
Y1 - 2019
N2 - Convolutional neural networks can provide a potential framework to characterize groundwater storage from seismic data. Estimation of key components, such as the amount of groundwater stored in an aquifer and delineate water table level, from active-source seismic data are performed in this study. The data to train, validate and test the neural networks are obtained by solving wave propagation in a coupled poroviscoelastic–elastic media. A discontinuous Galerkin method is applied to model wave propagation, whereas a deep convolutional neural network is used for the parameter estimation problem. In the numerical experiment, the primary unknowns estimated are the amount of stored groundwater and water table level, while the remaining parameters, assumed to be of less of interest, are marginalized in the convolutional neural network-based solution. Results, obtained through synthetic data, illustrate the potential of deep learning methods to extract additional aquifer information from seismic data, which otherwise would be impossible based on a set of reflection seismic sections or velocity tomograms.
AB - Convolutional neural networks can provide a potential framework to characterize groundwater storage from seismic data. Estimation of key components, such as the amount of groundwater stored in an aquifer and delineate water table level, from active-source seismic data are performed in this study. The data to train, validate and test the neural networks are obtained by solving wave propagation in a coupled poroviscoelastic–elastic media. A discontinuous Galerkin method is applied to model wave propagation, whereas a deep convolutional neural network is used for the parameter estimation problem. In the numerical experiment, the primary unknowns estimated are the amount of stored groundwater and water table level, while the remaining parameters, assumed to be of less of interest, are marginalized in the convolutional neural network-based solution. Results, obtained through synthetic data, illustrate the potential of deep learning methods to extract additional aquifer information from seismic data, which otherwise would be impossible based on a set of reflection seismic sections or velocity tomograms.
KW - Inverse problem
KW - Modelling
KW - Monitoring
KW - Wave
UR - http://www.scopus.com/inward/record.url?scp=85069896604&partnerID=8YFLogxK
U2 - 10.1111/1365-2478.12831
DO - 10.1111/1365-2478.12831
M3 - Article
AN - SCOPUS:85069896604
VL - 67
SP - 2115
EP - 2126
JO - Geophysical Prospecting
JF - Geophysical Prospecting
SN - 0016-8025
IS - 8
ER -