Abstract
Three-dimensional modeling of geoscientific data of limited amounts and strongly biased locations is difficult and impractical using almost any method. To obtain a correct spatial model from data under such constraints, this study systematically demonstrated the effectiveness of neural kriging (NK), which is based on a deep neural network (DNN) with a semivariogram learning criterion. As a novel case study of resource geology, NK was applied to clarify the three-dimensional deposit structure of an active seafloor hydrothermal vent area, Izena Hole in the middle Okinawa Trough, using Cu, Zn, Pb, Ag, and Ba content data, geological columns, and resistivity data from drill core samples obtained at six drill sites aligned nearly E–W and two drill sites far from these six. Two high-content zones clearly appeared for all five elements, in the sulfide mound and underlying sediment, and about 30 m below the seafloor as a stratiform shape. Advantages of NK were demonstrated as follows. The NK results showed the locations of these high-content zones and their horizontal extent along the pathway of hydrothermal fluids more clearly than the result of typical geostatistical simulation, turning bands simulation (TBS), by enabling content estimations even in areas without data. NK also estimated the uppermost high-content zones more accurately than the ordinal DNN methods or TBS. Next, NK was tested for a region 20 times wider than the first target, and the resultant Cu and Zn content models showed indistinct horizontal linear features and vertical continuity, which may have been caused by the fault structure and small amounts of metal precipitation in the deep part, respectively. Thus, one advantage of NK is that its spatial model can contribute to constructing hypotheses of transport and enrichment mechanisms of target metals.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Abbaszadeh Shahri A, Larsson S, Renkel C (2020) Artificial intelligence models to generate visualized bedrock level: a case study in Sweden. Model Earth Syst Environ 6:1509–1528. https://doi.org/10.1007/s40808-020-00767-0
Bengio Y, Courville AC, Vincent P (2013) Representation learning: a review and new perspectives. IEEE Trans Pattern Anal Mach Intell 35(8):1798–1828. https://doi.org/10.1109/TPAMI.2013.50
Bhattacharya BB, Jain SC, Mallick K (1974) Geophysical prospecting for barite. Geophys Prospect 22(3):421–429. https://doi.org/10.1111/j.1365-2478.1974.tb00096.x
Cao B, Luo X, Zhang L, Lei Y, Zhou J (2020) Petrofacies prediction and 3-D geological model in tight gas sandstone reservoirs by integration of well logs and geostatistical modeling. Mar Pet Geol 114:104202. https://doi.org/10.1016/j.marpetgeo.2019.104202
Chatterjee S, Bandopadhyay S, Ganguli R, Bhattacherjee A, Samanta B, Pal SK (2007) General regression neural network residual estimation for ore grade prediction of limestone deposit. Min Technol 116(3):89–99. https://doi.org/10.1179/174328607X228875
Chen L, Ren C, Li L, Wang Y, Zhang B, Wang Z, Li L (2019) A comparative assessment of geostatistical, machine learning, and hybrid approaches for mapping topsoil organic carbon content. ISPRS Int J Geoinf 8(4):174. https://doi.org/10.3390/ijgi8040174
Chilès JP, Desassis N (2018) Fifty years of kriging. In: Daya Sagar B, Cheng Q, Agterberg F (eds) Handbook of mathematical geosciences. Springer, Cham
de Sá VR, Koike K, Goto T-N, Nozaki T, Takaya Y, Yamasaki T (2021a) 3D geostatistical modeling of metal contents and lithofacies for mineralization mechanism determination of a seafloor hydrothermal deposit in the middle Okinawa Trough, Izena Hole. Ore Geol Rev 135:104194. https://doi.org/10.1016/j.oregeorev.2021.104194
de Sá VR, Koike K, Goto T-N, Nozaki T, Takaya Y, Yamasaki T (2021b) A combination of geostatistical methods and principal components analysis for detection of mineralized zones in seafloor hydrothermal systems. Nat Resour Res 30(4):2875–2887. https://doi.org/10.1007/s11053-020-09705-4
Dowd PA, Saraç C (1994) A neural network approach to geostatistical simulation. Math Geol 26(4):491–503. https://doi.org/10.1007/BF02083491
Doyle MG, Allen RL (2003) Subsea-floor replacement in volcanic-hosted massive sulfide deposits. Ore Geol Rev 23(3–4):183–222. https://doi.org/10.1016/S0169-1368(03)00035-0
Erdogan Erten G, Yavuz M, Deutsch CV (2022) Combination of machine learning and kriging for spatial estimation of geological attributes. Nat Resour Res 31:191–213. https://doi.org/10.1007/s11053-021-10003-w
Farkhutdinov A, Goblet P, de Fouquet C, Cherkasov S (2016) A case study of the modeling of a hydrothermal reservoir: Khankala deposit of geothermal waters. Geothermics 59:56–66. https://doi.org/10.1016/j.geothermics.2015.10.005
Ghaderi A, Abbaszadeh Shahri A, Larsson S (2019) An artificial neural network based model to predict spatial soil type distribution using piezocone penetration test data (CPTu). Bull Eng Geol Environ 78:4579–4588. https://doi.org/10.1007/s10064-018-1400-9
Glasby GP, Notsu K (2003) Submarine hydrothermal mineralization in the Okinawa Trough, SW of Japan: an overview. Ore Geol Rev 23(3–4):299–339. https://doi.org/10.1016/j.oregeorev.2003.07.001
Glorot X, Bordes A, Bengio Y (2011) Deep sparse rectifier neural networks. In: Proc 14th Int Conf Artif Intell Stat, pp 315–323
Heung B, Hob HC, Zhang J, Knudby A, Bulmer CE, Schmidt MG (2016) An overview and comparison of machine-learning techniques for classification purposes in digital soil mapping. Geoderma 265:62–77. https://doi.org/10.1016/j.geoderma.2015.11.014
Ishibashi J-I, Ikegami F, Tsuji T, Urabe T (2015) Hydrothermal activity in the Okinawa Trough back-arc basin: geological background and hydrothermal mineralization. In: Ishibashi J-I, Okino K, Sunamura M (eds) Subseafloor biosphere linked to hydrothermal systems: TAIGA concept. Springer, Tokyo
Ishitsuka K, Mogi T, Sugano K, Yamaya Y, Uchida T, Kajiwara T (2018) Resistivity-based temperature estimation of the Kakkonda geothermal field, Japan, using a neural network and neural kriging. IEEE Geosci Remote Sens Lett 15(8):1154–1158. https://doi.org/10.1109/LGRS.2018.2832647
Ishizu K, Goto T, Ohta Y, Kasaya T, Iwamoto H, Vachiratienchai C, Siripunvaraporn W, Tsuji T, Kumagai H, Koike K (2019) Internal structure of a seafloor massive sulfide deposit by electrical resistivity tomography, Okinawa Trough. Geophys Res Lett 46(20):11025–11034. https://doi.org/10.1029/2019GL083749
Jafrasteh B, Fathianpou N, Suárez A (2018) Comparison of machine learning methods for copper ore grade estimation. Comput Geosci 22:1371–1388. https://doi.org/10.1007/s10596-018-9758-0
Kanevski M, Parkin R, Pozdnukhov A, Timonin V, Maignan M, Demyanov V, Canu S (2004) Environmental data mining and modeling based on machine learning algorithms and geostatistics. Environ Model Softw 19:845–855. https://doi.org/10.1016/j.envsoft.2003.03.004
Keaney A, McKinley J, Graham C, Robinson M, Ruffell A (2013) Spatial statistics to estimate peat thickness using airborne radiometric data. Spat Stat 5:3–24. https://doi.org/10.1016/j.spasta.2013.05.003
Kim H-S, Ji Y (2022) Three-dimensional geotechnical-layer mapping in Seoul using borehole database and deep neural network-based model. Eng Geol 297:106489. https://doi.org/10.1016/j.enggeo.2021.106489
Kingma DP, Ba JL (2015) Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980
Kinoshita M, Yamano M (1997) Hydrothermal regime and constraints on reservoir depth of the Jade site in the Mid-Okinawa Trough inferred from heat flow measurements. J Geophys Res Solid Earth 102(B2):3183–3194. https://doi.org/10.1029/96JB03556
Koike K, Ichikawa Y (2006) Spatial correlation structures of fracture systems for identifying a scaling law and modeling fracture distributions. Comput Geosci 32(8):1079–1095. https://doi.org/10.1016/j.cageo.2006.02.013
Koike K, Matsuda S, Gu B (2001) Evaluation of interpolation accuracy of neural kriging with application to temperature-distribution analysis. Math Geol 33(4):421–448. https://doi.org/10.1023/A:1011084812324
Koike K, Matsuda S, Suzuki T, Ohmi M (2002) Neural network-based estimation of principal metal contents in the Hokuroku District, Northern Japan, for exploring Kuroko-type deposits. Nat Resour Res 11(2):135–156. https://doi.org/10.1023/A:1015520204066
Koike K, Matsuda S (2003) Characterizing content distributions of impurities in a limestone mine using a feedforward neural network. Nat Resour Res 12(3):209–223. https://doi.org/10.1023/A:1025180005454
Koike K, Matsuda S (2006) New indices for characterizing spatial models of ore deposits by the use of a sensitivity vector and influence factor. Math Geol 38(5):541–564. https://doi.org/10.1007/s11004-006-9030-3
Komori S, Masaki Y, Tanikawa W, Torimoto J, Ohta Y, Makio M, Maeda L, Ishibashi J-I, Nozaki T, Tadai O, Kumagai H (2017) Depth profiles of resistivity and spectral IP for active modern submarine hydrothermal deposits: a case study from the Iheya North Knoll and the Iheya Minor Ridge in Okinawa Trough, Japan. Earth Planets Space 69:114. https://doi.org/10.1186/s40623-017-0691-6
Li Q-Q, Yue T-X, Wang C-Q, Zhang W-J, Yu Y, Li B, Yang J, Bai G-C (2013) Spatially distributed modeling of soil organic matter across China: An application of artificial neural network approach. Catena 104:210–218. https://doi.org/10.1016/j.catena.2012.11.012
Lu L, Kashiwaya K, Koike K (2016) Geostatistics-based regional characterization of groundwater chemistry in a sedimentary rock area with faulted setting. Environ Earth Sci 75:829. https://doi.org/10.1007/s12665-016-5619-0
Maepa F, Smith RS, Tessema A (2021) Support vector machine and artificial neural network modelling of orogenic gold prospectivity mapping in the Swayze greenstone belt, Ontario, Canada. Ore Geol Rev 130:103968. https://doi.org/10.1016/j.oregeorev.2020.103968
Mosteller F, Tukey JW (1968) Data analysis, including statistics handbook of social psychology. Addison-Wesley
Mukhopadhyay A (1998) Spatial estimation of transmissivity using artificial neural network. Ground Water 37(3):458–464. https://doi.org/10.1111/j.1745-6584.1999.tb01125.x
Nair V, Hinton GE (2010) Rectified linear units improve restricted Boltzmann machines. In: Proc 27th Int Conf Mach Learn, pp 807–814
Naprstek T, Smith RS (2019) A new method for interpolating linear features in aeromagnetic data. Geophysics 84(3):15–24. https://doi.org/10.1190/GEO2018-0156.1
Nozaki T, Ishibashi J-I, Shimada K, Nagase T, Takaya Y, Kato Y, Takai K (2016) Rapid growth of mineral deposits at artificial seafloor hydrothermal vents. Sci Rep 6:22163. https://doi.org/10.1038/srep22163
Nozaki T, Nagase T, Takaya Y, Yamasaki T, Otake T, Yonezu K, Ikehata K, Totsuka S, Kitada K, Sanada Y, Yamada Y, Ishibashi J-I, Kumagai H (2021a) Subseafloor sulphide deposit formed by pumice replacement mineralisation. Sci Rep 11:8809. https://doi.org/10.1038/s41598-021-87050-z
Nozaki T, Nagase T, Ushikubo T, Shimizu K, Ishibashi J-I (2021b) Microbial sulfate reduction plays an important role at the initial stage of subseafloor sulfide mineralization. Geology 49(2):222–227. https://doi.org/10.1130/G47943.1
Ohmoto H (1996) Formation of volcanogenic massive sulfide deposits: the Kuroko perspective. Ore Geol Rev 10:135–177. https://doi.org/10.1016/0169-1368(95)00021-6
Panchal G, Ganatra A, Shah P, Panchal D (2011) Determination of over-learning and over-fitting problem in back propagation neural network. Int J Soft Comput 2(2):40–51. https://doi.org/10.5121/ijsc.2011.2204
Parra J, Emery X (2013) Geostatistics applied to cross-well reflection seismic for imaging carbonate aquifers. J Appl Geophys 92:68–75. https://doi.org/10.1016/j.jappgeo.2013.02.010
Pfeiffer H, Baumbach G, Sarachaga-Ruiz L, Kleanthous S, Poulida O, Beyaz E (2009) Neural modelling of the spatial distribution of air pollutants. Atmos Environ 43:3289–3297. https://doi.org/10.1016/j.atmosenv.2008.05.073
Ravindra K, Mor S (2019) Distribution and health risk assessment of arsenic and selected heavy metals in Groundwater of Chandigarh, India. Environ Pollut 250:820–830. https://doi.org/10.1016/j.envpol.2019.03.080
Robb L (2005) Introduction to ore-forming process. Blackwell Science
Samanta B, Bandopadhyay S, Ganguli R (2006) Comparative evaluation of neural network learning algorithms for ore grade estimation. Math Geol 38(2):175–197. https://doi.org/10.1007/s11004-005-9010-z
Sarma DD (2009) Geostatistics with applications in earth sciences. Springer, Dordrecht. https://doi.org/10.1007/978-1-4020-9380-7
Seo Y, Kim S, Singh VP (2015) Estimating spatial precipitation using regression kriging and artificial neural network residual kriging (RKNNRK) hybrid approach. Water Resour Manag 29:2189–2204. https://doi.org/10.1007/s11269-015-0935-9
Sergeev AP, Buevich AG, Baglaev EM, Shichkin AV (2019) Combining spatial autocorrelation with machine learning increases prediction accuracy of soil heavy metals. Catena 174:425–435. https://doi.org/10.1016/j.catena.2018.11.037
Song Y-Q, Yang L-A, Li B, Hu Y-M, Wang A-L, Zhou W, Cui X-S, Liu Y-L (2017) Spatial prediction of soil organic matter using a hybrid geostatistical model of an extreme learning machine and ordinary kriging. Sustainability 9:754. https://doi.org/10.3390/su9050754
Takaya Y, Yasukawa K, Kawasaki T et al. (2018) The tremendous potential of deep-sea mud as a source of rare-earth elements. Sci Rep 8:5763. https://doi.org/10.1038/s41598-018-23948-5
Tarasov DA, Buevich AG, Sergeev AP, Shichkin AV (2018) High variation topsoil pollution forecasting in the Russian Subarctic: Using artificial neural networks combined with residual kriging. Appl Geochem 88(B):188–197. https://doi.org/10.1016/j.apgeochem.2017.07.007
Tomita SA, Koike K, Goto T-N, Suzuki K (2020) Numerical simulation-based clarification of a fluid-flow system in a seafloor hydrothermal vent area in the middle Okinawa Trough. Geophys Res Lett 47:88681. https://doi.org/10.1029/2020GL088681
Tziachris P, Aschonitis V, Chatzistathis T, Papadopoulou M, Doukas ID (2020) Comparing machine learning models and hybrid geostatistical methods using environmental and soil covariates for soil pH prediction. ISPRS Int J Geo-Inf 9(4):276. https://doi.org/10.3390/ijgi9040276
Acknowledgements
This work was financially supported by the Cross-ministerial Strategic Innovation Promotion Program (SIP). Sincere thanks are extended to two anonymous reviewers for essential and constructive comments and suggestions that improved the clarity of this manuscript.
Author information
Authors and Affiliations
Contributions
All authors contributed to the study conception and design. Material preparation and data collection were performed by TN, YT and SK, and data analysis and spatial modeling were performed by KK, OY, VRS and SAT. The first draft of the manuscript was written by KK and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no competing interests.
Rights and permissions
About this article
Cite this article
Koike, K., Yono, O., de Sá, V.R. et al. Effectiveness of Neural Kriging for Three-Dimensional Modeling of Sparse and Strongly Biased Distribution of Geological Data with Application to Seafloor Hydrothermal Mineralization. Math Geosci 54, 1183–1206 (2022). https://doi.org/10.1007/s11004-022-10011-3
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11004-022-10011-3