Use of convolutional neural networks with encoder-decoder structure for predicting the inverse operator in hydraulic tomography

•A deep learning algorithm has been used for the approximation of the inverse operator in HT.•The algorithm is based on convolutional neural networks.•The approach has been successfully applied on the synthetic transmissivity fields. In this manuscript, we discuss the capabilities of a deep learning...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Journal of hydrology (Amsterdam) Ročník 604; s. 127233
Hlavní autoři: Jardani, A., Vu, T.M., Fischer, P.
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier B.V 01.01.2022
Elsevier
Témata:
ISSN:0022-1694, 1879-2707
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract •A deep learning algorithm has been used for the approximation of the inverse operator in HT.•The algorithm is based on convolutional neural networks.•The approach has been successfully applied on the synthetic transmissivity fields. In this manuscript, we discuss the capabilities of a deep learning algorithm implemented with the Conventional Neural Network concept to characterize the hydraulic properties of aquifers. The algorithm called CNN-HT is designed to predict the inverse operator of hydraulic tomography using a synthetic training dataset in which the hydraulic head data associated with pumping tests are linked to hydraulic transmissivity field. This approach relies on an adaptation of the SegNet network that was initially developed to process image segmentation. The SegNet is composed of encoders and decoders networks. In the encoder, sequential operations with multiple filters, as convolution, batch normalization, max-pooling are performed to identify feature maps of the input data. In the decoder, the up-sampling, convolution, batch normalization and regression operations are used to prepare the output by recovering the loss of spatial resolution that occurred in the encoder process. In this adaptation, we used the least-square iterative formulation at the initial iteration with Jacobian matrix to resize the hydraulic head data to match the size of the output (transmissivity field). This protocol was applied to the hydraulic head data computed numerically by solving the groundwater flow equation for a given transmissivity field, generated geostatistically with Gaussian and spherical variograms. A part of this data was used for training the network and the other part to test its performance. The test step confirmed the effectiveness of this tool in reconstructing the main heterogeneities of the hydraulic properties, and its effectiveness is related to the nature and quantity of the training data. Moreover, the CNN-HT method provided inversion results of the same quality than those obtained with the Gauss-Newton algorithm using the finite difference or adjoint state method in the computation of the Jacobian matrix. However, the computational time is longer in CNN-HT but this time can be less or of the same order as that of Gauss-Newton using finite difference method.
AbstractList In this manuscript, we discuss the capabilities of a deep learning algorithm implemented with the Conventional Neural Network concept to characterize the hydraulic properties of aquifers. The algorithm called CNN-HT is designed to predict the inverse operator of hydraulic tomography using a synthetic training dataset in which the hydraulic head data associated with pumping tests are linked to hydraulic transmissivity field. This approach relies on an adaptation of the SegNet network that was initially developed to process image segmentation. The SegNet is composed of encoders and decoders networks. In the encoder, sequential operations with multiple filters, as convolution, batch normalization, max-pooling are performed to identify feature maps of the input data. In the decoder, the up-sampling, convolution, batch normalization and regression operations are used to prepare the output by recovering the loss of spatial resolution that occurred in the encoder process. In this adaptation, we used the least-square iterative formulation at the initial iteration with Jacobian matrix to resize the hydraulic head data to match the size of the output (transmissivity field). This protocol was applied to the hydraulic head data computed numerically by solving the groundwater flow equation for a given transmissivity field, generated geostatistically with Gaussian and spherical variograms. A part of this data was used for training the network and the other part to test its performance. The test step confirmed the effectiveness of this tool in reconstructing the main heterogeneities of the hydraulic properties, and its effectiveness is related to the nature and quantity of the training data. Moreover, the CNN-HT method provided inversion results of the same quality than those obtained with the Gauss-Newton algorithm using the finite difference or adjoint state method in the computation of the Jacobian matrix. However, the computational time is longer in CNN-HT but this time can be less or of the same order as that of Gauss-Newton using finite difference method.
•A deep learning algorithm has been used for the approximation of the inverse operator in HT.•The algorithm is based on convolutional neural networks.•The approach has been successfully applied on the synthetic transmissivity fields. In this manuscript, we discuss the capabilities of a deep learning algorithm implemented with the Conventional Neural Network concept to characterize the hydraulic properties of aquifers. The algorithm called CNN-HT is designed to predict the inverse operator of hydraulic tomography using a synthetic training dataset in which the hydraulic head data associated with pumping tests are linked to hydraulic transmissivity field. This approach relies on an adaptation of the SegNet network that was initially developed to process image segmentation. The SegNet is composed of encoders and decoders networks. In the encoder, sequential operations with multiple filters, as convolution, batch normalization, max-pooling are performed to identify feature maps of the input data. In the decoder, the up-sampling, convolution, batch normalization and regression operations are used to prepare the output by recovering the loss of spatial resolution that occurred in the encoder process. In this adaptation, we used the least-square iterative formulation at the initial iteration with Jacobian matrix to resize the hydraulic head data to match the size of the output (transmissivity field). This protocol was applied to the hydraulic head data computed numerically by solving the groundwater flow equation for a given transmissivity field, generated geostatistically with Gaussian and spherical variograms. A part of this data was used for training the network and the other part to test its performance. The test step confirmed the effectiveness of this tool in reconstructing the main heterogeneities of the hydraulic properties, and its effectiveness is related to the nature and quantity of the training data. Moreover, the CNN-HT method provided inversion results of the same quality than those obtained with the Gauss-Newton algorithm using the finite difference or adjoint state method in the computation of the Jacobian matrix. However, the computational time is longer in CNN-HT but this time can be less or of the same order as that of Gauss-Newton using finite difference method.
ArticleNumber 127233
Author Jardani, A.
Vu, T.M.
Fischer, P.
Author_xml – sequence: 1
  givenname: A.
  surname: Jardani
  fullname: Jardani, A.
  email: abderrahim.jardani@univ-rouen.fr
  organization: Normandy University, UNIROUEN, UNICEAN, UMR CNRS 6143 M2C Morphodynamique Continentale et Côtière, Université de Rouen, France
– sequence: 2
  givenname: T.M.
  surname: Vu
  fullname: Vu, T.M.
  organization: Normandy University, UNIROUEN, UNICEAN, UMR CNRS 6143 M2C Morphodynamique Continentale et Côtière, Université de Rouen, France
– sequence: 3
  givenname: P.
  surname: Fischer
  fullname: Fischer, P.
  organization: HydroSciences Montpellier, Univ. Montpellier, CNRS, IRD, Montpellier, France
BackLink https://insu.hal.science/insu-03661813$$DView record in HAL
BookMark eNqFkU9r3DAUxEVJoZs0H6GgYyl4qz-OZdNDCaFpCgu9NGchS8-xtl7JfZI37KlfPXYceuglusxBM8NjfufkLMQAhHzgbMsZrz7vt_v-5DAOW8EE33KhhJRvyIbXqimEYuqMbBgTouBVU74j5ynt2fykLDfk730CGjtqYzjGYco-BjPQABM-S36M-DvRR597CsFGB1g4eFaaMk42Twi0i0hHBOdt9uGB5h6oD0fApXoENHn-94EuR5pp8JbmeIgPaMb-9J687cyQ4PJFL8j97bdfN3fF7uf3HzfXu8KWosxFy7rONYq3qlS8464TZa2qVsnaNNyykksr2loB8KYSrHGtdY41bPZIaZnl8oJ8Wnt7M-gR_cHgSUfj9d31TvuQJs1kVfGay-Ni_riaR4x_JkhZH3yyMAwmQJySFpWs5o3rspqtV6vVYkwJoftXzple6Oi9fqGjFzp6pTPnvvyXsz6bZf6Mxg-vpr-uaZgnO3pAnayf-cwIEGzWLvpXGp4Aoh2zlw
CitedBy_id crossref_primary_10_1007_s40808_024_01986_5
crossref_primary_10_1016_j_jhydrol_2022_128167
crossref_primary_10_1016_j_jhydrol_2023_129677
crossref_primary_10_1016_j_jhydrol_2024_130648
crossref_primary_10_1016_j_jhydrol_2022_128562
crossref_primary_10_1007_s00477_025_03051_8
crossref_primary_10_1016_j_jhydrol_2022_127830
crossref_primary_10_1016_j_jhydrol_2022_127752
crossref_primary_10_1016_j_jhydrol_2025_133897
crossref_primary_10_1016_j_jhydrol_2024_132368
crossref_primary_10_3390_en16083320
crossref_primary_10_1029_2022WR033241
crossref_primary_10_1002_hyp_70118
crossref_primary_10_1029_2025JH000703
crossref_primary_10_1016_j_earscirev_2023_104370
crossref_primary_10_1007_s11269_022_03346_3
crossref_primary_10_1016_j_jhydrol_2024_131703
crossref_primary_10_1016_j_jhydrol_2022_127443
crossref_primary_10_3390_w17020255
crossref_primary_10_1016_j_engappai_2024_109888
crossref_primary_10_1029_2023WR035408
crossref_primary_10_1016_j_jhydrol_2024_131524
crossref_primary_10_1016_j_advwatres_2023_104463
crossref_primary_10_1016_j_jhydrol_2024_132295
crossref_primary_10_3390_w15162890
crossref_primary_10_5194_hess_29_4251_2025
crossref_primary_10_1016_j_jhydrol_2025_133584
crossref_primary_10_1016_j_compgeo_2024_106316
Cites_doi 10.1007/s11004-012-9397-2
10.1016/0021-9991(92)90400-S
10.1016/j.procs.2018.05.069
10.1109/5.726791
10.1016/S0149-1970(96)00013-3
10.1007/s11004-008-9206-0
10.1088/0266-5611/11/2/005
10.1029/2011WR010616
10.1190/geo2012-0460.1
10.1029/2000WR900114
10.1002/wrcr.20519
10.1029/2001WR001176
10.1029/RG020i002p00219
10.1016/j.cageo.2020.104681
10.1016/j.jhydrol.2017.05.051
10.1109/TPAMI.2016.2644615
10.1016/j.jhydrol.2012.09.031
10.1111/gwat.12457
10.1093/gji/ggab024
10.1002/2015WR017922
10.1029/2008WR007078
10.1029/WR021i003p00359
10.1016/j.cageo.2008.01.013
10.1029/2004WR003874
10.1029/2018WR022643
10.1016/j.advwatres.2017.11.029
10.1016/j.jhydrol.2008.11.014
10.1002/2017WR022148
10.1029/2018GL080404
10.1016/j.jhydrol.2016.08.061
10.1007/BF02769620
10.1016/j.jcp.2018.04.018
10.1016/j.mcm.2011.07.009
ContentType Journal Article
Copyright 2021 Elsevier B.V.
Attribution - NonCommercial
Copyright_xml – notice: 2021 Elsevier B.V.
– notice: Attribution - NonCommercial
DBID AAYXX
CITATION
7S9
L.6
1XC
VOOES
DOI 10.1016/j.jhydrol.2021.127233
DatabaseName CrossRef
AGRICOLA
AGRICOLA - Academic
Hyper Article en Ligne (HAL)
Hyper Article en Ligne (HAL) (Open Access)
DatabaseTitle CrossRef
AGRICOLA
AGRICOLA - Academic
DatabaseTitleList AGRICOLA


DeliveryMethod fulltext_linktorsrc
Discipline Geography
EISSN 1879-2707
ExternalDocumentID oai:HAL:insu-03661813v1
10_1016_j_jhydrol_2021_127233
S002216942101283X
GroupedDBID --K
--M
-~X
.~1
0R~
1B1
1RT
1~.
1~5
29K
4.4
457
4G.
5GY
5VS
6TJ
7-5
71M
8P~
9JM
9JN
AABNK
AABVA
AACTN
AAEDT
AAEDW
AAIAV
AAIKJ
AAKOC
AALCJ
AALRI
AAOAW
AAQFI
AAQXK
AATLK
AAXUO
ABEFU
ABFNM
ABGRD
ABJNI
ABMAC
ABQEM
ABQYD
ABTAH
ABXDB
ABYKQ
ACDAQ
ACGFS
ACIUM
ACLVX
ACNCT
ACRLP
ACSBN
ADBBV
ADEZE
ADMUD
ADQTV
AEBSH
AEKER
AENEX
AEQOU
AFFNX
AFKWA
AFTJW
AFXIZ
AGHFR
AGUBO
AGYEJ
AHHHB
AIEXJ
AIKHN
AITUG
AJBFU
AJOXV
ALMA_UNASSIGNED_HOLDINGS
AMFUW
AMRAJ
ASPBG
ATOGT
AVWKF
AXJTR
AZFZN
BKOJK
BLXMC
CBWCG
CS3
D-I
DU5
EBS
EFJIC
EFLBG
EJD
EO8
EO9
EP2
EP3
FA8
FDB
FEDTE
FGOYB
FIRID
FNPLU
FYGXN
G-2
G-Q
GBLVA
HLV
HMA
HVGLF
HZ~
H~9
IHE
IMUCA
J1W
K-O
KOM
LW9
LY3
M41
MO0
N9A
O-L
O9-
OAUVE
OZT
P-8
P-9
P2P
PC.
Q38
R2-
RIG
ROL
RPZ
SAB
SCC
SDF
SDG
SDP
SEP
SES
SEW
SPC
SPCBC
SPD
SSA
SSE
SSZ
T5K
TN5
UQL
VOH
WUQ
Y6R
ZCA
ZMT
ZY4
~02
~G-
~KM
9DU
AAHBH
AATTM
AAXKI
AAYWO
AAYXX
ABUFD
ABWVN
ACLOT
ACRPL
ACVFH
ADCNI
ADNMO
ADVLN
AEIPS
AEUPX
AFJKZ
AFPUW
AGQPQ
AIGII
AIIUN
AKBMS
AKRWK
AKYEP
ANKPU
APXCP
CITATION
EFKBS
~HD
7S9
L.6
1XC
VOOES
ID FETCH-LOGICAL-c424t-b0ffd971b7471f1df24876b738a91c0413c2b87ee196209dbcdd09087633c0c13
ISICitedReferencesCount 28
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000731472900001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0022-1694
IngestDate Tue Oct 14 20:46:20 EDT 2025
Thu Oct 02 09:40:51 EDT 2025
Sat Nov 29 07:30:10 EST 2025
Tue Nov 18 21:51:28 EST 2025
Fri Feb 23 02:40:53 EST 2024
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Keywords Deep learning
Convolutional neural networks
Hydraulic tomography
Inverse problem
Language English
License Attribution - NonCommercial: http://creativecommons.org/licenses/by-nc
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c424t-b0ffd971b7471f1df24876b738a91c0413c2b87ee196209dbcdd09087633c0c13
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ORCID 0000-0003-0358-0728
0000-0002-3401-8140
OpenAccessLink https://insu.hal.science/insu-03661813
PQID 2636723846
PQPubID 24069
ParticipantIDs hal_primary_oai_HAL_insu_03661813v1
proquest_miscellaneous_2636723846
crossref_primary_10_1016_j_jhydrol_2021_127233
crossref_citationtrail_10_1016_j_jhydrol_2021_127233
elsevier_sciencedirect_doi_10_1016_j_jhydrol_2021_127233
PublicationCentury 2000
PublicationDate January 2022
2022-01-00
20220101
2022
PublicationDateYYYYMMDD 2022-01-01
PublicationDate_xml – month: 01
  year: 2022
  text: January 2022
PublicationDecade 2020
PublicationTitle Journal of hydrology (Amsterdam)
PublicationYear 2022
Publisher Elsevier B.V
Elsevier
Publisher_xml – name: Elsevier B.V
– name: Elsevier
References Berg, Illman (b0020) 2011; 47
Castagna, Bellin (b0040) 2009; 45
Jiménez, Mariethoz, Brauchler, Bayer (b0105) 2016; 52
Scales, Smith, Fischer (b0180) 1992; 103
Bohling, Zhan, Butler, Zheng (b0025) 2002; 38
Oliver, Cunha, Reynolds (b0160) 1997; 29
Fu, Gómez-Hernández (b0065) 2009; 41
Laloy, Hérault, Jacques, Linde (b0130) 2018; 54
Yeh, Liu (b0225) 2000; 36
Fischer, Jardani, Lecoq (b0060) 2018; 112
Fischer, Jardani, Soueid Ahmed, Abbas, Wang, Jourde, Lecoq (b0055) 2017; 55
Neuman (b0155) 1987
Fernández-Martínez, Mukerji, García-Gonzalo, Fernández-Muñiz (b0050) 2011; 54
Kitanidis (b0125) 1997
Cardiff, Kitanidis (b0030) 2008; 34
Elsheikh, Jackson, Laforce (b0045) 2012; 44
Wang, Jardani, Jourde (b0215) 2017; 551
Indolia, Goswami, Mishra, Asopa (b0090) 2018; 132
Ioffe S., Szegedy C., 2015. Batch normalization: Accelerating deep network training by reducing internal covariate shift, Proceedings of the 32nd International Conference on Machine Learning, 37, 448-456.
Li, Nowak, Cirpka (b0140) 2005; 41
Vu, Jardani (b0210) 2021; 225
Lochbühler, Doetsch, Brauchler, Linde (b0145) 2013; 78
Y. Wu Y. Lin InversionNet: A Real-Time and Accurate Full Waveform Inversion with CNNs and continuous CRFs 2018 arXiv preprint arXiv:1811.07875.
Gottlieb, Dietrich (b0085) 1995; 11
Zio (b0240) 1997; 31
Tarantola, Valette (b0205) 1982; 20
Badrinarayanan, Kendall, Cipolla (b0015) 2017; 39
Zhao, Illman, Berg (b0230) 2016; 542
Cardiff, Barrash, Kitanidis (b0035) 2013; 49
Kingma D.P., Ba J., 2014. Adam: A method for stochastic optimization, International Conference on Learning Representations (ICLR).
Sykes, Wilson, Andrews (b0200) 1985; 21
Kendall, Cipolla (b0110) 2016; 2016
LeCun, Bottou, Bengio, Haffner (b0135) 1998; 86
Puzyrev, Swidinsky (b0165) 2021; 149
Sun (b0195) 2018; 45
Gal, Ghahramani (b0075) 2016
Jardani, Dupont, Revil, Massei, Fournier, Laignel (b0100) 2012; 472-473
Fu, Jaime Gómez-Hernández (b0070) 2009; 364
Zhu, Zabaras (b0235) 2018; 366
Remy, Boucher, Wu (b0170) 2009
Shen (b0185) 2018; 54
Badrinarayanan (10.1016/j.jhydrol.2021.127233_b0015) 2017; 39
Fernández-Martínez (10.1016/j.jhydrol.2021.127233_b0050) 2011; 54
Zio (10.1016/j.jhydrol.2021.127233_b0240) 1997; 31
Castagna (10.1016/j.jhydrol.2021.127233_b0040) 2009; 45
Neuman (10.1016/j.jhydrol.2021.127233_b0155) 1987
Kendall (10.1016/j.jhydrol.2021.127233_b0110) 2016; 2016
Fischer (10.1016/j.jhydrol.2021.127233_b0055) 2017; 55
Gottlieb (10.1016/j.jhydrol.2021.127233_b0085) 1995; 11
Lochbühler (10.1016/j.jhydrol.2021.127233_b0145) 2013; 78
Li (10.1016/j.jhydrol.2021.127233_b0140) 2005; 41
Shen (10.1016/j.jhydrol.2021.127233_b0185) 2018; 54
10.1016/j.jhydrol.2021.127233_b0220
Oliver (10.1016/j.jhydrol.2021.127233_b0160) 1997; 29
Cardiff (10.1016/j.jhydrol.2021.127233_b0035) 2013; 49
Zhu (10.1016/j.jhydrol.2021.127233_b0235) 2018; 366
Gal (10.1016/j.jhydrol.2021.127233_b0075) 2016
Zhao (10.1016/j.jhydrol.2021.127233_b0230) 2016; 542
Wang (10.1016/j.jhydrol.2021.127233_b0215) 2017; 551
Scales (10.1016/j.jhydrol.2021.127233_b0180) 1992; 103
Fischer (10.1016/j.jhydrol.2021.127233_b0060) 2018; 112
Tarantola (10.1016/j.jhydrol.2021.127233_b0205) 1982; 20
Fu (10.1016/j.jhydrol.2021.127233_b0070) 2009; 364
Jardani (10.1016/j.jhydrol.2021.127233_b0100) 2012; 472-473
LeCun (10.1016/j.jhydrol.2021.127233_b0135) 1998; 86
Bohling (10.1016/j.jhydrol.2021.127233_b0025) 2002; 38
Cardiff (10.1016/j.jhydrol.2021.127233_b0030) 2008; 34
Sykes (10.1016/j.jhydrol.2021.127233_b0200) 1985; 21
Elsheikh (10.1016/j.jhydrol.2021.127233_b0045) 2012; 44
Berg (10.1016/j.jhydrol.2021.127233_b0020) 2011; 47
Kitanidis (10.1016/j.jhydrol.2021.127233_b0125) 1997
Sun (10.1016/j.jhydrol.2021.127233_b0195) 2018; 45
Laloy (10.1016/j.jhydrol.2021.127233_b0130) 2018; 54
Vu (10.1016/j.jhydrol.2021.127233_b0210) 2021; 225
Jiménez (10.1016/j.jhydrol.2021.127233_b0105) 2016; 52
Fu (10.1016/j.jhydrol.2021.127233_b0065) 2009; 41
10.1016/j.jhydrol.2021.127233_b0095
Remy (10.1016/j.jhydrol.2021.127233_b0170) 2009
Yeh (10.1016/j.jhydrol.2021.127233_b0225) 2000; 36
Puzyrev (10.1016/j.jhydrol.2021.127233_b0165) 2021; 149
Indolia (10.1016/j.jhydrol.2021.127233_b0090) 2018; 132
10.1016/j.jhydrol.2021.127233_b0115
References_xml – volume: 2016
  start-page: 4762
  year: 2016
  end-page: 4769
  ident: b0110
  article-title: Modelling uncertainty in deep learning for camera relocalization
  publication-title: IEEE international conference on Robotics and Automation (ICRA)
– volume: 45
  start-page: 11137
  year: 2018
  end-page: 11146
  ident: b0195
  article-title: Discovering State-Parameter Mappings in Subsurface Models Using Generative Adversarial Networks
  publication-title: Geophys. Res. Lett.
– volume: 86
  start-page: 2278
  year: 1998
  end-page: 2324
  ident: b0135
  article-title: Gradient-based learning applied to document recognition
  publication-title: Proc. IEEE
– volume: 364
  start-page: 328
  year: 2009
  end-page: 341
  ident: b0070
  article-title: Uncertainty assessment and data worth in groundwater flow and mass transport modeling using a blocking Markov chain Monte Carlo method
  publication-title: J. Hydrol.
– volume: 551
  start-page: 29
  year: 2017
  end-page: 46
  ident: b0215
  article-title: A hybrid inverse method for hydraulic tomography in fractured and karstic media
  publication-title: J. Hydrol.
– reference: Ioffe S., Szegedy C., 2015. Batch normalization: Accelerating deep network training by reducing internal covariate shift, Proceedings of the 32nd International Conference on Machine Learning, 37, 448-456.
– volume: 38
  start-page: 60-1
  year: 2002
  end-page: 60-15
  ident: b0025
  article-title: Steady shape analysis of tomographic pumping tests for characterization of aquifer heterogeneities
  publication-title: Water Resour. Res.
– volume: 366
  start-page: 415
  year: 2018
  end-page: 447
  ident: b0235
  article-title: Bayesian deep convolutional encoder–decoder networks for surrogate modeling and uncertainty quantification
  publication-title: J. Comput. Phys.
– year: 2016
  ident: b0075
  article-title: Dropout as a Bayesian approximation: Representing model uncertainty in deep learning
  publication-title: Proceedings of the 33rd International Conference on Machine Learning
– volume: 34
  start-page: 1480
  year: 2008
  end-page: 1491
  ident: b0030
  article-title: Efficient solution of nonlinear, underdetermined inverse problems with a generalized PDE model
  publication-title: Comput. Geosci.
– volume: 41
  year: 2005
  ident: b0140
  article-title: A Geostatistical inverse modeling of transient pumping tests using temporal moments of drawdown
  publication-title: Water Resour. Res.
– start-page: 533
  year: 1987
  end-page: 561
  ident: b0155
  article-title: Stochastic continuum representation of fractured rock permeability as an alternative to the R.E.V
  publication-title: and fracture network concepts,
– volume: 55
  start-page: 208
  year: 2017
  end-page: 218
  ident: b0055
  article-title: Application of large-scale inversion algorithms to hydraulic tomography in an alluvial aquifer
  publication-title: Groundwater
– volume: 472-473
  start-page: 287
  year: 2012
  end-page: 300
  ident: b0100
  article-title: Geostatistical inverse modeling of the transmissivity field of a heterogeneous alluvial aquifer under tidal influence
  publication-title: J. Hydrol.
– year: 2009
  ident: b0170
  article-title: Applied Geostatistics with SGeMS: A User’s Guide
– volume: 21
  start-page: 359
  year: 1985
  end-page: 371
  ident: b0200
  article-title: Sensitivity analysis for steady state groundwater flow using adjoint operators
  publication-title: Water Resour. Res.
– volume: 39
  start-page: 2481
  year: 2017
  end-page: 2495
  ident: b0015
  article-title: Segnet: A deep convolutional encoder-decoder architecture for image segmentation
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 41
  start-page: 105
  year: 2009
  end-page: 128
  ident: b0065
  article-title: A blocking Markov Chain Monte Carlo method for inverse stochastic hydrogeological modeling
  publication-title: Math. Geosci.
– volume: 49
  start-page: 7311
  year: 2013
  end-page: 7326
  ident: b0035
  article-title: Hydraulic conductivity imaging from 3-D transient hydraulic tomography at several pumping/observation densities
  publication-title: Water Resour. Res.
– volume: 31
  start-page: 303
  year: 1997
  end-page: 315
  ident: b0240
  article-title: Approaching the inverse problem of parameter estimation in groundwater models by means of artificial neural networks
  publication-title: Prog. Nucl. Energy
– volume: 45
  start-page: W04410
  year: 2009
  ident: b0040
  article-title: A Bayesian approach for inversion of hydraulic tomographic data
  publication-title: Water Resour. Res.
– reference: Kingma D.P., Ba J., 2014. Adam: A method for stochastic optimization, International Conference on Learning Representations (ICLR).
– reference: Y. Wu Y. Lin InversionNet: A Real-Time and Accurate Full Waveform Inversion with CNNs and continuous CRFs 2018 arXiv preprint arXiv:1811.07875.
– volume: 78
  start-page: ID1
  year: 2013
  end-page: ID14
  ident: b0145
  article-title: Structure-coupled joint inversion of geophysical and hydrological data
  publication-title: Geophysics
– volume: 149
  start-page: 104681
  year: 2021
  ident: b0165
  article-title: Inversion of 1D frequency-and time-domain electromagnetic data with convolutional neural networks
  publication-title: Comput. Geosci.
– volume: 54
  start-page: 2889
  year: 2011
  end-page: 2899
  ident: b0050
  article-title: Uncertainty assessment for inverse problems in high dimensional spaces using particle swarm optimization and model reduction techniques
  publication-title: Math. Comput. Modell.
– volume: 11
  start-page: 353
  year: 1995
  end-page: 360
  ident: b0085
  article-title: Identification of the permeability distribution in soil by hydraulic tomography
  publication-title: Inverse Prob.
– volume: 36
  start-page: 2095
  year: 2000
  end-page: 2105
  ident: b0225
  article-title: Hydraulic tomography: Development of a new aquifer test method
  publication-title: Water Resour. Res.
– volume: 132
  start-page: 679
  year: 2018
  end-page: 688
  ident: b0090
  article-title: Conceptual understanding of convolutional neural network - a deep learning approach
  publication-title: Procedia Comput. Sci.
– volume: 54
  start-page: 8558
  year: 2018
  end-page: 8593
  ident: b0185
  article-title: A transdisciplinary review of deep learning research and its relevance for water resources scientists
  publication-title: Water Resour. Res.
– volume: 20
  start-page: 219
  year: 1982
  end-page: 232
  ident: b0205
  article-title: Generalized nonlinear inverse problems solved using the least squares criterion
  publication-title: Rev. Geophys.
– volume: 103
  start-page: 258
  year: 1992
  end-page: 268
  ident: b0180
  article-title: Global optimization methods for multimodal inverse problems
  publication-title: J. Comput. Phys.
– volume: 112
  start-page: 83
  year: 2018
  end-page: 94
  ident: b0060
  article-title: Hydraulic tomography of discrete networks of conduits and fractures in a karstic aquifer by using a deterministic inversion algorithm
  publication-title: Adv. Water Resour.
– volume: 54
  start-page: 381
  year: 2018
  end-page: 406
  ident: b0130
  article-title: Training-image based geostatistical inversion using a spatial generative adversarial neural network
  publication-title: Water Resour. Res.
– volume: 29
  start-page: 61
  year: 1997
  end-page: 91
  ident: b0160
  article-title: Markov chain Monte Carlo methods for conditioning a permeability field to pressure data
  publication-title: Math. Geol.
– volume: 542
  start-page: 156
  year: 2016
  end-page: 171
  ident: b0230
  article-title: On the importance of geological data for hydraulic tomography analysis: laboratory sandbox study
  publication-title: J. Hydrol.
– volume: 44
  start-page: 515
  year: 2012
  end-page: 543
  ident: b0045
  article-title: Bayesian Reservoir History Matching Considering Model and Parameter Uncertainties
  publication-title: Math. Geosci.
– year: 1997
  ident: b0125
  article-title: Introduction to geostatistics: applications in hydrogeology
– volume: 47
  start-page: W10507
  year: 2011
  ident: b0020
  article-title: Three-dimensional transient hydraulic tomography in a highly heterogeneous glaciofluvial aquifer-aquitard system
  publication-title: Water Resour. Res.
– volume: 52
  start-page: 3966
  year: 2016
  end-page: 3983
  ident: b0105
  article-title: Smart pilot points using reversible-jump Markov-chain Monte Carlo
  publication-title: Water Resour. Res.
– volume: 225
  start-page: 1319
  year: 2021
  end-page: 1331
  ident: b0210
  article-title: Convolutional neural networks with SegNet architecture applied to three-dimensional tomography of subsurface electrical resistivity: CNN-3D-ERT
  publication-title: Geophys. J. Int.
– volume: 44
  start-page: 515
  issue: 5
  year: 2012
  ident: 10.1016/j.jhydrol.2021.127233_b0045
  article-title: Bayesian Reservoir History Matching Considering Model and Parameter Uncertainties
  publication-title: Math. Geosci.
  doi: 10.1007/s11004-012-9397-2
– year: 2009
  ident: 10.1016/j.jhydrol.2021.127233_b0170
– volume: 103
  start-page: 258
  issue: 2
  year: 1992
  ident: 10.1016/j.jhydrol.2021.127233_b0180
  article-title: Global optimization methods for multimodal inverse problems
  publication-title: J. Comput. Phys.
  doi: 10.1016/0021-9991(92)90400-S
– volume: 132
  start-page: 679
  year: 2018
  ident: 10.1016/j.jhydrol.2021.127233_b0090
  article-title: Conceptual understanding of convolutional neural network - a deep learning approach
  publication-title: Procedia Comput. Sci.
  doi: 10.1016/j.procs.2018.05.069
– volume: 86
  start-page: 2278
  issue: 11
  year: 1998
  ident: 10.1016/j.jhydrol.2021.127233_b0135
  article-title: Gradient-based learning applied to document recognition
  publication-title: Proc. IEEE
  doi: 10.1109/5.726791
– ident: 10.1016/j.jhydrol.2021.127233_b0220
– volume: 31
  start-page: 303
  issue: 3
  year: 1997
  ident: 10.1016/j.jhydrol.2021.127233_b0240
  article-title: Approaching the inverse problem of parameter estimation in groundwater models by means of artificial neural networks
  publication-title: Prog. Nucl. Energy
  doi: 10.1016/S0149-1970(96)00013-3
– ident: 10.1016/j.jhydrol.2021.127233_b0115
– volume: 41
  start-page: 105
  year: 2009
  ident: 10.1016/j.jhydrol.2021.127233_b0065
  article-title: A blocking Markov Chain Monte Carlo method for inverse stochastic hydrogeological modeling
  publication-title: Math. Geosci.
  doi: 10.1007/s11004-008-9206-0
– volume: 11
  start-page: 353
  issue: 2
  year: 1995
  ident: 10.1016/j.jhydrol.2021.127233_b0085
  article-title: Identification of the permeability distribution in soil by hydraulic tomography
  publication-title: Inverse Prob.
  doi: 10.1088/0266-5611/11/2/005
– volume: 47
  start-page: W10507
  year: 2011
  ident: 10.1016/j.jhydrol.2021.127233_b0020
  article-title: Three-dimensional transient hydraulic tomography in a highly heterogeneous glaciofluvial aquifer-aquitard system
  publication-title: Water Resour. Res.
  doi: 10.1029/2011WR010616
– volume: 78
  start-page: ID1
  issue: 3
  year: 2013
  ident: 10.1016/j.jhydrol.2021.127233_b0145
  article-title: Structure-coupled joint inversion of geophysical and hydrological data
  publication-title: Geophysics
  doi: 10.1190/geo2012-0460.1
– volume: 36
  start-page: 2095
  issue: 8
  year: 2000
  ident: 10.1016/j.jhydrol.2021.127233_b0225
  article-title: Hydraulic tomography: Development of a new aquifer test method
  publication-title: Water Resour. Res.
  doi: 10.1029/2000WR900114
– volume: 49
  start-page: 7311
  issue: 11
  year: 2013
  ident: 10.1016/j.jhydrol.2021.127233_b0035
  article-title: Hydraulic conductivity imaging from 3-D transient hydraulic tomography at several pumping/observation densities
  publication-title: Water Resour. Res.
  doi: 10.1002/wrcr.20519
– start-page: 533
  year: 1987
  ident: 10.1016/j.jhydrol.2021.127233_b0155
  article-title: Stochastic continuum representation of fractured rock permeability as an alternative to the R.E.V
– volume: 38
  start-page: 60-1
  issue: 12
  year: 2002
  ident: 10.1016/j.jhydrol.2021.127233_b0025
  article-title: Steady shape analysis of tomographic pumping tests for characterization of aquifer heterogeneities
  publication-title: Water Resour. Res.
  doi: 10.1029/2001WR001176
– volume: 20
  start-page: 219
  issue: 2
  year: 1982
  ident: 10.1016/j.jhydrol.2021.127233_b0205
  article-title: Generalized nonlinear inverse problems solved using the least squares criterion
  publication-title: Rev. Geophys.
  doi: 10.1029/RG020i002p00219
– volume: 149
  start-page: 104681
  year: 2021
  ident: 10.1016/j.jhydrol.2021.127233_b0165
  article-title: Inversion of 1D frequency-and time-domain electromagnetic data with convolutional neural networks
  publication-title: Comput. Geosci.
  doi: 10.1016/j.cageo.2020.104681
– volume: 2016
  start-page: 4762
  year: 2016
  ident: 10.1016/j.jhydrol.2021.127233_b0110
  article-title: Modelling uncertainty in deep learning for camera relocalization
  publication-title: IEEE international conference on Robotics and Automation (ICRA)
– volume: 551
  start-page: 29
  year: 2017
  ident: 10.1016/j.jhydrol.2021.127233_b0215
  article-title: A hybrid inverse method for hydraulic tomography in fractured and karstic media
  publication-title: J. Hydrol.
  doi: 10.1016/j.jhydrol.2017.05.051
– volume: 39
  start-page: 2481
  issue: 12
  year: 2017
  ident: 10.1016/j.jhydrol.2021.127233_b0015
  article-title: Segnet: A deep convolutional encoder-decoder architecture for image segmentation
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
  doi: 10.1109/TPAMI.2016.2644615
– volume: 472-473
  start-page: 287
  year: 2012
  ident: 10.1016/j.jhydrol.2021.127233_b0100
  article-title: Geostatistical inverse modeling of the transmissivity field of a heterogeneous alluvial aquifer under tidal influence
  publication-title: J. Hydrol.
  doi: 10.1016/j.jhydrol.2012.09.031
– volume: 55
  start-page: 208
  issue: 2
  year: 2017
  ident: 10.1016/j.jhydrol.2021.127233_b0055
  article-title: Application of large-scale inversion algorithms to hydraulic tomography in an alluvial aquifer
  publication-title: Groundwater
  doi: 10.1111/gwat.12457
– volume: 225
  start-page: 1319
  issue: 2
  year: 2021
  ident: 10.1016/j.jhydrol.2021.127233_b0210
  article-title: Convolutional neural networks with SegNet architecture applied to three-dimensional tomography of subsurface electrical resistivity: CNN-3D-ERT
  publication-title: Geophys. J. Int.
  doi: 10.1093/gji/ggab024
– volume: 52
  start-page: 3966
  issue: 5
  year: 2016
  ident: 10.1016/j.jhydrol.2021.127233_b0105
  article-title: Smart pilot points using reversible-jump Markov-chain Monte Carlo
  publication-title: Water Resour. Res.
  doi: 10.1002/2015WR017922
– year: 1997
  ident: 10.1016/j.jhydrol.2021.127233_b0125
– year: 2016
  ident: 10.1016/j.jhydrol.2021.127233_b0075
  article-title: Dropout as a Bayesian approximation: Representing model uncertainty in deep learning
– volume: 45
  start-page: W04410
  issue: 4
  year: 2009
  ident: 10.1016/j.jhydrol.2021.127233_b0040
  article-title: A Bayesian approach for inversion of hydraulic tomographic data
  publication-title: Water Resour. Res.
  doi: 10.1029/2008WR007078
– volume: 21
  start-page: 359
  issue: 3
  year: 1985
  ident: 10.1016/j.jhydrol.2021.127233_b0200
  article-title: Sensitivity analysis for steady state groundwater flow using adjoint operators
  publication-title: Water Resour. Res.
  doi: 10.1029/WR021i003p00359
– volume: 34
  start-page: 1480
  issue: 11
  year: 2008
  ident: 10.1016/j.jhydrol.2021.127233_b0030
  article-title: Efficient solution of nonlinear, underdetermined inverse problems with a generalized PDE model
  publication-title: Comput. Geosci.
  doi: 10.1016/j.cageo.2008.01.013
– volume: 41
  issue: 8
  year: 2005
  ident: 10.1016/j.jhydrol.2021.127233_b0140
  article-title: A Geostatistical inverse modeling of transient pumping tests using temporal moments of drawdown
  publication-title: Water Resour. Res.
  doi: 10.1029/2004WR003874
– volume: 54
  start-page: 8558
  issue: 11
  year: 2018
  ident: 10.1016/j.jhydrol.2021.127233_b0185
  article-title: A transdisciplinary review of deep learning research and its relevance for water resources scientists
  publication-title: Water Resour. Res.
  doi: 10.1029/2018WR022643
– volume: 112
  start-page: 83
  year: 2018
  ident: 10.1016/j.jhydrol.2021.127233_b0060
  article-title: Hydraulic tomography of discrete networks of conduits and fractures in a karstic aquifer by using a deterministic inversion algorithm
  publication-title: Adv. Water Resour.
  doi: 10.1016/j.advwatres.2017.11.029
– volume: 364
  start-page: 328
  issue: 3-4
  year: 2009
  ident: 10.1016/j.jhydrol.2021.127233_b0070
  article-title: Uncertainty assessment and data worth in groundwater flow and mass transport modeling using a blocking Markov chain Monte Carlo method
  publication-title: J. Hydrol.
  doi: 10.1016/j.jhydrol.2008.11.014
– ident: 10.1016/j.jhydrol.2021.127233_b0095
– volume: 54
  start-page: 381
  issue: 1
  year: 2018
  ident: 10.1016/j.jhydrol.2021.127233_b0130
  article-title: Training-image based geostatistical inversion using a spatial generative adversarial neural network
  publication-title: Water Resour. Res.
  doi: 10.1002/2017WR022148
– volume: 45
  start-page: 11137
  issue: 20
  year: 2018
  ident: 10.1016/j.jhydrol.2021.127233_b0195
  article-title: Discovering State-Parameter Mappings in Subsurface Models Using Generative Adversarial Networks
  publication-title: Geophys. Res. Lett.
  doi: 10.1029/2018GL080404
– volume: 542
  start-page: 156
  year: 2016
  ident: 10.1016/j.jhydrol.2021.127233_b0230
  article-title: On the importance of geological data for hydraulic tomography analysis: laboratory sandbox study
  publication-title: J. Hydrol.
  doi: 10.1016/j.jhydrol.2016.08.061
– volume: 29
  start-page: 61
  issue: 1
  year: 1997
  ident: 10.1016/j.jhydrol.2021.127233_b0160
  article-title: Markov chain Monte Carlo methods for conditioning a permeability field to pressure data
  publication-title: Math. Geol.
  doi: 10.1007/BF02769620
– volume: 366
  start-page: 415
  year: 2018
  ident: 10.1016/j.jhydrol.2021.127233_b0235
  article-title: Bayesian deep convolutional encoder–decoder networks for surrogate modeling and uncertainty quantification
  publication-title: J. Comput. Phys.
  doi: 10.1016/j.jcp.2018.04.018
– volume: 54
  start-page: 2889
  issue: 11-12
  year: 2011
  ident: 10.1016/j.jhydrol.2021.127233_b0050
  article-title: Uncertainty assessment for inverse problems in high dimensional spaces using particle swarm optimization and model reduction techniques
  publication-title: Math. Comput. Modell.
  doi: 10.1016/j.mcm.2011.07.009
SSID ssj0000334
Score 2.5189185
Snippet •A deep learning algorithm has been used for the approximation of the inverse operator in HT.•The algorithm is based on convolutional neural networks.•The...
In this manuscript, we discuss the capabilities of a deep learning algorithm implemented with the Conventional Neural Network concept to characterize the...
SourceID hal
proquest
crossref
elsevier
SourceType Open Access Repository
Aggregation Database
Enrichment Source
Index Database
Publisher
StartPage 127233
SubjectTerms algorithms
Convolutional neural networks
data collection
Deep learning
equations
groundwater flow
Hydraulic tomography
image analysis
Inverse problem
Sciences of the Universe
tomography
Title Use of convolutional neural networks with encoder-decoder structure for predicting the inverse operator in hydraulic tomography
URI https://dx.doi.org/10.1016/j.jhydrol.2021.127233
https://www.proquest.com/docview/2636723846
https://insu.hal.science/insu-03661813
Volume 604
WOSCitedRecordID wos000731472900001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVESC
  databaseName: Elsevier SD Freedom Collection Journals 2021
  customDbUrl:
  eissn: 1879-2707
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0000334
  issn: 0022-1694
  databaseCode: AIEXJ
  dateStart: 19950101
  isFulltext: true
  titleUrlDefault: https://www.sciencedirect.com
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1bb9MwFLZKhwQviKsoNxnBW5WQOFmcPEaw0Y1RVWJDfbOS2NFarWnVm8YTv4j_yDmxnbYbaOOBl6RKGifp99U-Pj7nO4S8l3HOI84iRwXKc8IoUk7GktJRpZ_FGS9hyMrrYhO834-Hw2TQav2yuTDrC15V8eVlMvuvUMMxABtTZ_8B7qZROACfAXTYAuywvRXwZ9o7j-Hk5j6AAspW1rs66NuktKGGpVRzR6p639VasriigLGHszmu4SxtOtWowgAOaHqm6pV5dJSc_5DzbIUy2cvpZFv7-rq1i1_Vck9g0qYTlGeQyMXGD3EMVNX1pbppDk8DneBoYk9-X9W8crtf3YZwo4Xl2wCGdhPDa_wXbDPPtSk1290zzIz9SFc9dpXukWOeYNIc3-6yI12y-Fr3rz0RY3esXwqm_8x3fcaZFtvYldvupd_E4NOhODnqf9k9uxWj2EtPYIt5AQ6M9RHYQ8EaJtt7jO8ncZvspUcHw-PN0B8EoZWnx_fYpIx9-OND_c0YunOOUblXjIPa4jl9SB4Y8GiqKfaItFT1mNz7rAzQT8hPoBqdlnSHalRTjVqqUaQavUI12lCNAtXohmoUqEYN1ailGhygDdXohmpPydnhwenHnmMKejhFyMKlk3tlKRPu5-gJKX1ZMpguRzkP4izxCw_sqYLlMVcKhgXmJTIvpPSSWjQxKLzCD56RdjWt1HNCEw6WOw-womQegt2ZSb6vYHJSqsLL_JB1SGh_WVEYtXssunIhbFjjWBhABAIiNCAd4jaXzbTcy00XxBY2YWxWbYsK4ORNl74DmJvboM47sE0g18SGax3y1tJAQM-Py3lZpaarhWBREGHJwDB6cauWXpL7-P_T_sNXpA1Aq9fkbrFejhbzN4bKvwGU29ZI
linkProvider Elsevier
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Use+of+convolutional+neural+networks+with+encoder-decoder+structure+for+predicting+the+inverse+operator+in+hydraulic+tomography&rft.jtitle=Journal+of+hydrology+%28Amsterdam%29&rft.au=Jardani%2C+Abderrahim&rft.au=Vu%2C+T.+M.&rft.au=Fischer%2C+Pierre&rft.date=2022&rft.pub=Elsevier&rft.issn=0022-1694&rft.eissn=1879-2707&rft.volume=604&rft_id=info:doi/10.1016%2Fj.jhydrol.2021.127233&rft.externalDBID=HAS_PDF_LINK&rft.externalDocID=oai%3AHAL%3Ainsu-03661813v1
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0022-1694&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0022-1694&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0022-1694&client=summon