2021 Vol. 12, No. 1

Research Paper
Deadly oasis: Recurrent annihilation of Cretaceous desert bryophyte colonies; the role of solar, climate and lithospheric forcing
Juan Pedro Rodríguez-López, Eduardo Barrón, Daniel Peyrot, Gary B. Hughes
2021, 12(1): 1-12. doi: 10.1016/j.gsf.2020.06.008
Abstract(263) HTML PDF(17)

Many oases (wet interdunes) are sedimentary systems characterized by high-frequency water-level oscillations, marked changes in salinity and intense biological activity at their margins. They are considered to be one of the most challenging environments on Earth for ecosystem development. These dynamic, depositional settings are usually unfavourable for fossilization and subsequent preservation of vegetal remains. This paper describes bryophyte (liverwort) assemblages occurring in three successive horizons interpreted to represent (i) recurrent early successional phases of biological soil crust colonization of wet interdune margins or (ii) exceptional preservation of floating or riparian liverworts in oasis pond waters associated with a progressive fall of the interdune water level. The record of in situ colonization surfaces characterized by delicate (e.g. lignin-free) three-dimensional structures represents an exceptional type of preservation herein associated with a rapid variation in phreatic interdune water level and the subsequent establishment of anoxic and reducing conditions. The occurrence of exceptionally preserved liverwort colonies coincides with the sedimentary record of, at least, three seismite levels in the oasis. Data gathered from the site suggests that the water table of the oasis was controlled by a combination of (i) a positive creation of accommodation space due to subsidence associated with movement on syn-sedimentary extensional faults, and (ii) the rise and fall of the oasis water table controlled by the oscillations of the groundwater system due to orbital changes which appear to drive the variability of the climate system. Rising groundwater levels flooded the oasis soil crusts and lead to the exceptional recurrent preservation of liverwort colonies at the oasis margins. Alternatively, considering the hypothesis of floating or riparian liverworts in the oasis pond waters, the fall in the level of the oasis water table placed the floating liverworts in contact with the oasis bottom sediments. This fall in the level of the oasis water table could indicate a cessation of accommodation space by syn-sedimentary extensional faults and/or a regional lowering of the groundwater system level associated with drought periods. Preliminary results indicate that oasis lamination between liverwort colonies records decadal and sub-decadal cyclicity, related with 11-year Schwabe sunspot and sub-decadal NAO cyclicities, conferring for every sedimentary cycle between liverwort colonies a duration of approximately 200 years, that otherwise matches the expected recurrence period for the De Vries cycle of solar activity.

Identification of the Early Jurassic mylonitic granitic pluton and tectonic implications in Namling area, southern Tibet
Yuanku Meng, Walter D. Mooney, Runlong Fan, Jinqing Liu, Youqing Wei
2021, 12(1): 13-28. doi: 10.1016/j.gsf.2020.07.010
Abstract(481) HTML PDF(25)

A number of studies revealed that the Gangdese magmatic belt of southern Tibet was closely related to the northward subduction of the Neo-Tethys oceanic lithosphere and Indo-Asian collision. However, pre-Cretaceous magmatism is still poorly constrained in the Gangdese magmatic belt, southern Tibet. Here, we conducted systematically geochronology and geochemistry studies on a newly-identified granitic pluton in the middle Gangdese magmatic belt (Namling area), southern Tibet. Zircon SHRIMP II U–Pb dating for one representative sample gives a weighted age of 184.2±1.8 Ma (MSWD = 1.11), corresponding to emplacement and crystallization age of the granitic pluton in the Early Jurassic (Pliensbachian). High SiO2 (68.9–72.1 wt.%) contents and intermediate Mg# values (35–38) together suggest that the newly-identified granitic pluton was probably formed by partial melting of crustal material with minor injection of mantle-derived magma, precluding an origin from melting of metasedimentary rocks that are characterized by low Mg# and high zircon δ18O values (>8‰). Geochemically, the newly-identified granitic pluton belongs to typical I-type granitic affinity, whereas this is inconsistent with aluminium saturation index (ASI = A/CNK ratios) and geochemical signatures. This suggests that zircon oxygen isotopes (4.30‰–5.28‰) and mineral features (lacking Al-rich minerals) are reliable indicators for discriminating granitic origin. Significantly depleted whole-rock Sr-Nd-Hf isotopic compositions and zircon εHf(t) values indicate that the granitic pluton was derived from partial melting of depleted arc-type lavas. In addition, the granitic pluton shows zircon δ18O values ranging from 4.30‰ to 5.28‰ (with a mean value of 4.77‰) that are consistent with mantle-derived zircon values (5.3‰ 0.6‰) within the uncertainties, indicating that the granitic pluton might have experienced weak short-living high-temperature hydrous fluid-rock interaction. Combined with the Sr-Nd-Hf-O isotopes and geochemical signatures, we propose that the newly-identified granitic pluton was originated from partial melting of depleted mafic lower crust, and experienced only negligible wall-rock contamination during ascent. Integrated with published data, we also propose that the initial subduction of the Neo-Tethys oceanic lithosphere occurred no later than the Pliensbachian of the Early Jurassic.

Quantification of submarine groundwater discharge (SGD) using radon, radium tracers and nutrient inputs in Punnakayal, south coast of India
S. Selvam, P. Muthukumar, Sruthy Sajeev, S. Venkatramanan, S. Y. Chung, K. Brindha, D. S. Suresh Babu, R. Murugan
2021, 12(1): 29-38. doi: 10.1016/j.gsf.2020.06.012
Abstract(238) HTML PDF(11)

The present study focused on the estimation of submarine groundwater discharge (SGD) and the effects of nutrient fluxes due to the SGD process. The parameters of SGD such as magnitude, character, and nutrient flux in Punnakayal region of South East coast of India were evaluated using multiple tracers of groundwater inputs in 2019. It was found that the elevated values for the tracers in the study area, displayed a gradational change in the values as move from estuarine part to the offshore. Simultaneous occurrence of fresh and saline SGD is observed on the study sites. Also, indicated that the SGD fluxes ranged from 0.04 to 0.12 m3 m-2 d-1 at the estuary and 0.03–0.15 m3 m-2 d-1 at the groundwater site. A substantially increased value for 222Rn activities is distinguished in the estuary to values over 312 dpm L-1. Nutrient embellishments were generally greatest at locations with substantial meteoric elements in groundwater; however, the recirculation of saltwater through the geological formation could provide a way of transferring terrestrially-derived nutrients to the coastal zone at many places.

Geochronological and geochemical constraints on the petrogenesis of late Mesoproterozoic mafic and granitic rocks in the southwestern Yangtze Block
Guichun Liu, Jing Li, Xin Qian, Qinglai Feng, Wei Wang, Guangyan Chen, Shaobin Hu
2021, 12(1): 39-52. doi: 10.1016/j.gsf.2020.07.005
Abstract(383) HTML PDF(24)

Late Mesoproterozoic igneous rocks in the SW Yangtze Block are important for understanding the role of it in reconstruction of the Rodinia supercontinent. In the present study, we report new geochronological, geochemical, and Nd–Hf isotopic data for the Cuoke plagioclase amphibolites and granites in the SW Yangtze Block. Geochronological results show that the plagioclase amphibolites and granites have similar late Mesoproterozoic zircon U–Pb ages of 1168–1162 Ma, constituting a bimodal igneous assemblage. The plagioclase amphibolites have high and variable TiO2 contents (1.15–4.30 wt.%) and Mg# (34–66) values, similar to the tholeiitic series. They are characterized by enrichment in LREEs and LILEs, and have OIB-like affinities with positive Nb and Ta anomalies. The plagioclase amphibolites have positive whole-rock εNd(t) (+3.2 to +4.3) and zircon εHf(t) (+4.3 to +10.7) values, indicating that they were derived from an OIB-like asthenospheric mantle source. The granites belong to the reduced peralkaline A-type series and have negative εNd(t) value of -6.0 and εHf(t) values of -5.8 to -13.8, indicating a derivation from the partial melting of ancient mafic lower crust. In combination with the ~1.05–1.02 Ga bimodal igneous assemblage in the SW Yangtze Block, we propose that the Cuoke 1168–1162 Ma igneous rocks were likely formed in a continental rift basin and argue against the existance of Grenvillian Orogen in the SW Yangtze Block during the late Mesoproterozoic.

PetroGram: An excel-based petrology program for modeling of magmatic processes
Mesut Gündüz, Kürşad Asan
2021, 12(1): 81-92. doi: 10.1016/j.gsf.2020.06.010
Abstract(242) HTML PDF(17)

PetroGram is an Excel© based magmatic petrology program that generates numerical and graphical models. PetroGram can model the magmatic processes such as melting, crystallization, assimilation and magma mixing based on the trace element and isotopic data. The program can produce both inverse and forward geochemical models for melting processes (e.g. forward model for batch, fractional and dynamic melting, and inverse model for batch and dynamic melting). However, the program uses a forward modeling approach for magma differentiation processes such as crystallization (EC: Equilibruim Crystallization, FC: Fractional Crystallization, IFC: Imperfect Fractional Crystallization and In-situ Crystallization), assimilation (AFC: Assimilation Fractional Crystallization, Decoupled FC-A: Decoupled Fractional Crystallization and Assimillation, A-IFC: Assimilation and Imperfect Fractional Crystallization) and magma mixing. One of the most important advantages of the program is that the melt composition obtained from any partial melting model can be used as a starting composition of the crystallization, assimilation and magma mixing. In addition, PetroGram is able to carry out the classification, tectonic setting, multi-element (spider) and isotope correlation diagrams, and basic calculations including Mg#, Eu/Eu*, εSr and εNd widely used in magmatic petrology.

Landslide susceptibility modeling based on ANFIS with teaching-learning-based optimization and Satin bowerbird optimizer
Wei Chen, Xi Chen, Jianbing Peng, Mahdi Panahi, Saro Lee
2021, 12(1): 93-107. doi: 10.1016/j.gsf.2020.07.012
Abstract(133) HTML PDF(8)

As threats of landslide hazards have become gradually more severe in recent decades, studies on landslide prevention and mitigation have attracted widespread attention in relevant domains. A hot research topic has been the ability to predict landslide susceptibility, which can be used to design schemes of land exploitation and urban development in mountainous areas. In this study, the teaching-learning-based optimization (TLBO) and satin bowerbird optimizer (SBO) algorithms were applied to optimize the adaptive neuro-fuzzy inference system (ANFIS) model for landslide susceptibility mapping. In the study area, 152 landslides were identified and randomly divided into two groups as training (70%) and validation (30%) dataset. Additionally, a total of fifteen landslide influencing factors were selected. The relative importance and weights of various influencing factors were determined using the step-wise weight assessment ratio analysis (SWARA) method. Finally, the comprehensive performance of the two models was validated and compared using various indexes, such as the root mean square error (RMSE), processing time, convergence, and area under receiver operating characteristic curves (AUROC). The results demonstrated that the AUROC values of the ANFIS, ANFIS-TLBO and ANFIS-SBO models with the training data were 0.808, 0.785 and 0.755, respectively. In terms of the validation dataset, the ANFISSBO model exhibited a higher AUROC value of 0.781, while the AUROC value of the ANFIS-TLBO and ANFIS models were 0.749 and 0.681, respectively. Moreover, the ANFIS-SBO model showed lower RMSE values for the validation dataset, indicating that the SBO algorithm had a better optimization capability. Meanwhile, the processing time and convergence of the ANFIS-SBO model were far superior to those of the ANFIS-TLBO model. Therefore, both the ensemble models proposed in this paper can generate adequate results, and the ANFIS-SBO model is recommended as the more suitable model for landslide susceptibility assessment in the study area considered due to its excellent accuracy and efficiency.

Geochronological and geochemical evidence for a Late Ordovician to Silurian arc–back-arc system in the northern Great Xing’an Range, NE China
Bo Liu, Jia-Fu Chen, Bao-Fu Han, Jun-Lai Liu, Jia-Wei Li
2021, 12(1): 131-145. doi: 10.1016/j.gsf.2020.07.002
Abstract(55) HTML PDF(11)

The early Paleozoic tectonic evolution of the Xing'an–Mongolian Orogenic Belt is dominated by two oceanic basins on the northwestern and southeastern sides of the Xing'an Block, i.e., the Xinlin–Xiguitu Ocean and the Nenjiang Ocean. However, the early development of the Nenjiang Ocean remains unclear. Here, we present zircon U–Pb geochronology and whole-rock elemental and Sr–Nd isotopic data on the gabbros in the Xinglong area together with andesitic tuffs and basalts in the Duobaoshan area. LA-ICP-MS zircon U–Pb dating of gabbros and andesitic tuffs yielded crystallization ages of 443–436 Ma and 452–451 Ma, respectively. The Early Silurian Xinglong gabbros show calc-alkaline and E-MORB affinities but they are enriched in LILEs, and depleted in HFSEs, with relatively low U/Th ratios of 0.18–0.36 and εNd(t) values of -1.6 to +0.5. These geochemical features suggest that the gabbros might originate from a mantle wedge modified by pelagic sediment-derived melts, consistent with a back-arc basin setting. By contrast, the andesitic tuffs are characterized by high MgO (>5 wt.%), Cr (138–200 ppm), and Ni (65–110 ppm) contents, and can be termed as high-Mg andesites. Their low Sr/Y ratios of 15.98–17.15 and U/Th values of 0.24–0.25 and moderate (La/Sm)n values of 3.07–3.26 are similar to those from the Setouchi Volcanic Belt (SW Japan), and are thought to be derived from partial melting of subducted sediments, and subsequent melt-mantle interaction. The Duobaoshan basalts have high Nb (8.44–10.30 ppm) and TiO2 contents (1.17–1.60 wt.%), typical of Nb-enriched basalts. They are slightly younger than regional adakitic rocks and have positive εNd(t) values of +5.2 to +5.7 and are interpreted to be generated by partial melting of a depleted mantle source metasomatized by earlier adakitic melts. Synthesized with coeval arc-related igneous rocks from the southeastern Xing'an Block, we propose that the Duobaoshan high-Mg andesitic tuffs and Nbenriched basalts are parts of the Late Ordovician and Silurian Sonid Zuoqi–Duobaoshan arc belt, and they were formed by the northwestern subduction of the Nenjiang Ocean. Such a subduction beneath the integrated Xing'an–Erguna Block also gave rise to the East Ujimqin–Xinglong igneous belt in a continental back-arc basin setting. Our new data support an early Paleozoic arc–back-arc model in the northern Great Xing'an Range.

Adakite-like granitoids of Songkultau: A relic of juvenile Cambrian arc in Kyrgyz Tien Shan
D. Konopelko, R. Seltmann, A. Dolgopolova, I. Safonova, S. Glorie, J. De Grave, M. Sun
2021, 12(1): 147-160. doi: 10.1016/j.gsf.2020.08.006

The early Paleozoic Terskey Suture zone, located in the southern part of the Northern Tien Shan domain in Kyrgyzstan, comprises tectonic slivers of dismembered ophiolites and associated primitive volcanics and deepmarine sediments. In the Lake Songkul area, early-middle Cambrian pillow basalts are crosscut by the Songkultau intrusion of coarse-grained gneissose quartz diorites and tonalites with geochemical characteristics typical for high-SiO2 adakites (SiO2 > 56 wt.%, Al2O3 > 15 wt.%, Na2O > 3.5 wt.% and high Sr/Y and La/Yb ratios). The Songkultau granitoids have positive initial εNd (+3.8 to +6.4) and εHf (+12.3 to +13.5) values indicating derivation from sources with MORB-like isotopic signature. Volcanic formations, surrounding the Songkultau intrusion, have geochemical affinities varying from ocean floor to island arc series. This rock assemblage is interpreted as a relic of an early–middle Cambrian primitive arc where the adakite-like granitoids were derived from partial melting of young and hot subducted oceanic crust. An age of 505 Ma, obtained for the Songkultau intrusion, shows that hot subduction under the Northern Tien Shan continued until middle Cambrian. The primitive arc complexes were obducted onto the Northern Tien Shan domain, where the Andean type continental magmatic arc developed in Cambrian and Ordovician. Formation of the Andean type arc was accompanied by uplift, erosion and deposition of coarse clastic sediments. A depositional age of ca. 470 Ma, obtained for the gravellites in the Lake Songkul area, is in agreement with the timing of deposition for lower Ordovician conglomerates elsewhere in the Northern Tien Shan, and corresponds to the main phase of the Andean type magmatism. The Songkultau adakites in association with surrounding ocean floor and island arc formations constitute a relic of a primitive Cambrian arc and represent a juvenile domain of substantial size identified so far within the predominantly crustal-derived terranes of Tien Shan. On a regional scale this primitive arc can be compared with juvenile Cambrian arcs of Kazakhstan, Gorny Altai and Mongolia.

Switching from advancing to retreating subduction in the Neoproterozoic Tarim Craton, NW China: Implications for Rodinia breakup
Guanghui Wu, Shuai Yang, Wei Liu, R. Damian Nance, Xin Chen, Zecheng Wang, Yang Xiao
2021, 12(1): 161-171. doi: 10.1016/j.gsf.2020.03.013

Geodynamic drivers for the supercontinent cycle are generally attributed to either top-down (subduction-related) or bottom-up (mantle-related) processes. Compiled geochemical data and U–Pb ages and Hf isotopic signatures for magmatic and detrital zircons from the Tarim Craton reveal a distinct change in subduction style during the Neoproterozoic. The subduction cycle is recorded in increasing and decreasing intensity of subduction-related magmatic rocks and time-equivalent sedimentary successions, and converse trends of εHf(t) values and corresponding changes in crustal incubation time. These trends are consistent with a switch from advancing to retreating subduction. The switch likely occurred at ca. 760 Ma when zircon εHf(t) values increase and crustal incubation times decrease following a transitional shift between 800 Ma and 760 Ma. A switch at this time is consistent with Rodinia breakup and may have resulted in the late Neoproterozoic Tarim rift basin. The long-lived (ca. 500 Ma) subduction recorded in the Tarim Craton suggests the predominance of a top-down process for Rodinia breakup on this part of its margin.

Characteristics of carbonatites from the northern part of the Korean Peninsula: A perspective from distribution, geology and geochemistry
Jinkon Ju, Yunsong Kim, Mansik Gang
2021, 12(1): 173-181. doi: 10.1016/j.gsf.2020.02.005

The present study introduces the carbonatite in the northern part of the Korean Peninsula for the first time. Recent exploration and development of the phosphorus-bearing carbonate rocks in the area have accumulated new geological data which gave us an opportunity to study origin of the carbonate rocks. We conducted geological survey, geochemical analyses of trace elements and rare earth elements, and carbon and oxygen isotope analyses for the carbonatites from Ssangryong, Pungnyon, Yongyu and Puhung districts of the northern part of the Korean Peninsula. This research confirms that the phosphorus-bearing carbonate rocks are carbonatite originating from the mantle. The studied carbonatites are distributed at the junctions of ring and linear structures or around their margins and contain a greater amount of REEs, Y, and Sr than carbonate rocks. The carbonatites in Yongyu and Puhung area show evidence that they were formed from mantle plume generated at the lower mantle and display similar fractionation characteristics to carbonatites in Barrado Itapirapua in Brazil and Kalkfeld and Ondurakorume in Namibia. REE patterns of the carbonatites are typical of carbonatites and the carbon and oxygen isotope analyses demonstrate that the carbonatites were originated from mantle. The carbonatites from the northern part of the Korean Peninsula have a great potential for sources of REE, Y, PGE (platinum group elements), copper, and gold.

New sedimentological and palynological data from the Yarkand-Fergana Basin (Kyrgyz Tian Shan): Insights on its Mesozoic paleogeographic and tectonic evolution
Julien Morin, Marc Jolivet, Dave Shaw, Sylvie Bourquin, Elena Bataleva
2021, 12(1): 183-202. doi: 10.1016/j.gsf.2020.04.010

The Talas Fergana/Karatau Fault, is a major tectonic boundary separating the Kazakh-Turan domain to the west from the Tian Shan domain to the east. During the Jurassic, movements along the fault led to the opening of several basins. Still, the Mesozoic kinematics of the fault and the geodynamic mechanism that led to the opening of these basins are largely unconstrained. Located at its southwestern termination, the Yarkand-Fergana Basin is certainly the best exposed and however still poorly understood. In this study, we provide new sedimentological description of the Jurassic series from the northern part of the Yarkand-Fergana Basin as well as new palynological data. Following a Middle–Late Triassic period dominated by regional erosion, the onset of sedimentation in the Yarkand-Fergana Basin occurred during the Sinemurian(?)–Pliensbachian. The basin opened as a half graben controlled by the Talas Fergana/Karatau Fault and separated from the Fergana Basin by basement highs. Extension persisted during the late Pliensbachian–Middle Jurassic, leading to a general widening of the YarkandFergana Basin. Finally, Late Jurassic–Early Cretaceous renewed tectonic activity in the area led to the inversion of the north Yarkand-Fergana Basin. The Early to Middle Jurassic timing of development of the Yarkand-Fergana Basin suggests that the coeval movements along the Talas Fergana/Karatau Fault are not associated to the collision of the Qiangtang block along the southern margin of Eurasia. We favor the hypothesis of an opening controlled by transtension related to far field effects of back-arc extension along the Neo-Tethys subduction zone to the west.

A large epeiric methanogenic Bambuí sea in the core of Gondwana supercontinent?
Sergio Caetano-Filho, Pierre Sansjofre, Magali Ader, Gustavo M. Paula-Santos, Cristian Guacaneme, Marly Babinski, Carolina Bedoya-Rueda, Matheus Kuchenbecker, Humberto L. S. Reis, Ricardo I. F. Trindade
2021, 12(1): 203-218. doi: 10.1016/j.gsf.2020.04.005

Carbon isotope compositions of both sedimentary carbonate and organic matter can be used as key proxies of the global carbon cycle and of its evolution through time, as long as they are acquired from waters where the dissolved inorganic carbon (DIC) is in isotope equilibrium with the atmospheric CO2. However, in shallow water platforms and epeiric settings, the influence of local to regional parameters on carbon cycling may lead to DIC isotope variations unrelated to the global carbon cycle. This may be especially true for the terminal Neoproterozoic, when Gondwana assembly isolated waters masses from the global ocean, and extreme positive and negative carbon isotope excursions are recorded, potentially decoupled from global signals. To improve our understanding on the type of information recorded by these excursions, we investigate the paired δ13Ccarb and δ13C org evolution for an increasingly restricted late Ediacaran-Cambrian foreland system in the West Gondwana interior: the basal Bambuí Group. This succession represents a 1st-order sedimentary sequence and records two major δ13Ccarb excursions in its two lowermost lower-rank sequences. The basal cap carbonate interval at the base of the first sequence, deposited when the basin was connected to the ocean, hosts antithetical negative and positive excursions for δ13Ccarb and δ13Corg, respectively, resulting in Δ13C values lower than 25‰. From the top of the basal sequence upwards, an extremely positive δ13Ccarb excursion is coupled to δ13Corg, reaching values of +14‰ and -14‰, respectively. This positive excursion represents a remarkable basin-wide carbon isotope feature of the Bambuí Group that occurs with only minor changes in Δ13C values, suggesting change in the DIC isotope composition. We argue that this regional isotopic excursion is related to a disconnection between the intrabasinal and the global carbon cycles. This extreme carbon isotope excursion may have been a product of a disequilibria between the basin DIC and atmospheric CO2 induced by an active methanogenesis, favored by the basin restriction. The drawdown of sulfate reservoir by microbial sulfate reduction in a poorly ventilated and dominantly anoxic basin would have triggered methanogenesis and ultimately methane escape to the atmosphere, resulting in a13C-enriched DIC influenced by methanogenic CO2. Isolated basins in the interior of the Gondwana supercontinent may have represented a significant source of methane inputs to the atmosphere, potentially affecting both the global carbon cycle and the climate.

Boron isotopic variations in tourmaline from metacarbonates and associated calc-silicate rocks from the Bohemian Massif: Constraints on boron recycling in the Variscan orogen
Lukáš Krmíček, Milan Novák, Robert B. Trumbull, Jan Cempírek, Stanislav Houzar
2021, 12(1): 219-230. doi: 10.1016/j.gsf.2020.03.009

Various metacarbonate and associated calc-silicate rocks form minor but genetically significant components of the lithological units in the Bohemian Massif of the Variscan orogen in Central Europe. These rocks vary in terms of their lithostratigraphy, chemical composition and mineral assemblage (dolomite/calcite ratio, silicate abundance). Tourmaline is present in five paragenetic settings within the metacarbonate and calc-silicate units. Type I comprises individual, euhedral, prismatic grains and grain aggregates in a carbonate-dominant (calcite±dolomite) matrix poor in silicates. Type II is characterized by euhedral to subhedral grains and coarse- to fine-grained aggregates in silicate-rich layers/nests within metacarbonate bodies whereas type III occurs as prismatic grains and aggregates at the contact zones between carbonate and associated silicate host rocks. Type IV is in veins crosscutting metacarbonate bodies, and type V tourmaline occurs at the exocontacts of elbaite-subtype granitic pegmatite. Tourmaline from the different settings shows distinctive compositional features. Typical for type I are Mg-rich compositions, with fluor-uvite > dravite >> magnesio-lucchesiite. Tourmalines from type II silicate-rich layers/nests are highly variable, corresponding to oxy-schorl, magnesio-foitite, Al-rich dravite and fluor-uvite. Typical for type III tourmalines are Ca,Ti-bearing oxy-dravite compositions. The type IV veins feature dravite and fluor-uvite tourmaline compositions whereas type V tourmaline is Li,F-rich dravite. Tourmaline is the only Bbearing phase in paragenetic types I–IV, where it is characterised by two principal ranges of B-isotope composition (δ11B =-13‰ to -9‰ and -18‰ to -14‰). These ranges correspond to regionally different units of the Moldanubian Zone. Thus, the Svratka Unit (Moldanubian Zone s.l.) contains only isotopically lighter tourmaline (δ11B = -18‰ to -14‰), whereas metacarbonates in the Polička unit (Teplá–Barrandian Zone) and Olešnice unit (Moravicum of the Moravo-Silesian Zone) has exclusively isotopically heavier tourmaline (δ11B = -9‰ to -13‰). Tourmalines from metacarbonates in the Variegated Unit cover both ranges of isotope composition. The isotopically light end of the B isotope range may indicate the presence of continental evaporites within individual investigated areas. On the other hand, variations in the range of ~8 δ-units is consistent with the reported shift in B isotopic composition of metasedimentary rocks of the Bohemian Massif due to the prograde metamorphism from very-low grade to eclogite facies. In contrast to the metacarbonate-hosted settings, tourmaline of paragenetic type V from the exocontact of granitic pegmatites displays a significantly heavier range of δ11B (as low as –7.7‰ to –0.6‰), which is attributed to partitioning of 10B to cogenetic axinite and/or different B-signature of the source pegmatite containing tourmaline with heavy δ11B signature.

Backward automatic calibration for three-dimensional landslide models
Giacomo Titti, Giulia Bossi, Gordon G. D. Zhou, Gianluca Marcato, Alessandro Pasuto
2021, 12(1): 231-241. doi: 10.1016/j.gsf.2020.03.011
Abstract(112) HTML PDF(2)

Back-analysis is broadly used for approaching geotechnical problems when monitoring data are available and information about the soils properties is of poor quality. For landslide stability assessment back-analysis calibration is usually carried out by time consuming trial-and-error procedure. This paper presents a new automatic Decision Support System that supports the selection of the soil parameters for three-dimensional models of landslides based on monitoring data. The method considering a pool of possible solutions, generated through permutation of soil parameters, selects the best ten configurations that are more congruent with the measured displacements. This reduces the operator biases while on the other hand allows the operator to control each step of the computation. The final selection of the preferred solution among the ten best-fitting solutions is carried out by an operator. The operator control is necessary as he may include in the final decision process all the qualitative elements that cannot be included in a qualitative analysis but nevertheless characterize a landslide dynamic as a whole epistemological subject, for example on the base of geomorphological evidence. A landslide located in Northeast Italy has been selected as example for showing the system potentiality. The proposed method is straightforward, scalable and robust and could be useful for researchers and practitioners.

A comparative study of geometric and geostatistical methods for qualitative reserve estimation of limestone deposit
Thomas Busuyi Afeni, Victor Oluwatosin Akeju, Adeyemi Emman Aladejare
2021, 12(1): 243-253. doi: 10.1016/j.gsf.2020.02.019
Abstract(1159) HTML PDF(24)

Mining projects especially relating to limestone deposits require an accurate knowledge of tonnage and grade, for both short and long-term planning. This is often difficult to establish as detailed exploration operations, which are required to get the accurate description of the deposit, are costly and time consuming. Geologists and mining engineers usually make use of geometric and geostatistical methods, for estimating the tonnage and grade of ore reserves. However, explicit assessments into the differences between these methods have not been reported in literature. To bridge this research gap, a comparative study is carried out to compare the qualitative reserve of Oyo-Iwa limestone deposit located in Nigeria, using geometric and geostatistical methods. The geometric method computes the reserve of the limestone deposit as 74,536,820 t (mean calcite, CaO grade = 52.15) and 99,674,793 t (mean calcite, CaO grade = 52.32), for the Northern and Southern zones of the deposit, respectively. On the other hand, the geostatistical method calculates the reserve as 81,626,729.65 t (mean calcite, CaO grade = 53.36) and -100,098,697.46 t (mean calcite, CaO grade = 52.96), for the two zones, respectively. The small relative difference in tonnage estimation between the two methods (i.e., 9.51% and 0.43%), proves that the geometric method is effective for tonnage estimation. In contrast, the relative difference in grade estimation between the two methods (i.e., 2.32% and -1.26%) is not negligible, and could be crucial in maintaining the profitability of the project. The geostatistical method is, therefore, more suitable, reliable and preferable for grade estimation, since it involves the use of spatial modelling and cross-validated interpolation. In addition, the geostatistical method is used to produce quality maps and three-dimensional (3-D) perspective view of the limestone deposit. The quality maps and 3-D view of the limestone deposit reveal the variability of the limestone grade within the deposit, and it is useful for operational management of the limestone raw materials. The qualitative mapping of the limestone deposit is key to effective production scheduling and accurate projection of raw materials for cement production.

Sicilian serpentinite xenoliths containing abiotic organics with nanodiamond clusters as key model for prebiotic processes
Sergei K. Simakov, Vittorio Scribano, Nikolai N. Mel'Nik, Germana Barone
2021, 12(1): 255-263. doi: 10.1016/j.gsf.2020.04.008

Rock fragments from the deepest parts of a buried hydrothermal system belonging to the Mesozoic Tethys Ocean entered as xenoliths in a Miocenic diatreme, hence brought to the surface, in the Hyblean Plateau (Sicily). Some xenoliths consist of strongly serpentinized ultramafic rocks bearing blebs of abiotic organic matter, where clusters of amorphous carbon nanoparticles, including nanodiamonds, are immersed. Such an occurrence conjures up established hypotheses that diamond surfaces are suitable catalytic platforms stimulating the assemblage of complex bio-organic molecules relevant to the emergence of life on Earth. The appearance of bio-organic molecules under primitive Earth conditions is one of the major unsolved questions on the origin of life. Here we report new micro-Raman spectra on blebs of abiotic organic matter from a selected xenolith. Diamond bands were related to hydrogenated nanocrystalline diamonds, with size of nearly 1–1.6 nm, formed from organics at low pressures and temperatures. In particular, diamond surfaces can give rise to crystalline interfacial water layers that may have played a fundamental role in the early biosphere evolution as a good medium for rapidly transporting positive charges in the form of hydrated protons. Nowadays, proton gradients in alkaline hydrothermal vents along oceanic ridges are generally viewed as key pre-biotic factors. In general, serpentinites span the entire geological record, including prebiotic times. These hydrous ultramafic rocks often display evidence of abiotic carbon species, both organic and inorganic, including nanodiamonds, being also capable to give rise to chemiosmotic processes and proton gradients necessary to the organisms, such as the “Last Universal Common Ancestor” (LUCA), in the prebiotic Earth.

Geology, geochronology and geochemistry of large Duobaoshan Cu–Mo–Au orefield in NE China: Magma genesis and regional tectonic implications
Wen-yan Cai, Ke-yong Wang, Jian Li, Li-juan Fu, Chun-kit Lai, Han-lun Liu
2021, 12(1): 265-292. doi: 10.1016/j.gsf.2020.04.013
Abstract(93) HTML PDF(11)

Duobaoshan is the largest porphyry-related Cu–Mo–Au orefield in northeastern (NE) Asia, and hosts a number of large-medium porphyry Cu (PCDs), epithermal Au and Fe–Cu skarn deposits. Formation ages of these deposits, from the oldest (Ordovician) to youngest (Jurassic), have spanned across over 300 Ma. No similar orefields of such size and geological complexity are found in NE Asia, which reflects its metallogenic uniqueness in forming and preserving porphyry-related deposits. In this study, we explore the actual number and timing of magmatic/ mineralization phases, their respective magma genesis, fertility, and regional tectonic connection, together with the preservation of PCDs. We present new data on the magmatic/mineralization ages (LA–ICP–MS zircon U–Pb, pyrite and molybdenite Re–Os dating), whole-rock geochemistry, and zircon trace element compositions on four representative deposits in the Duobaoshan orefield, i.e., Duobaoshan PCD, Tongshan PCD, Sankuanggou Fe–Cu skarn, and Zhengguang epithermal Au deposits, and compiled published ones from these and other mineral occurrences in the orefield.
In terms of geochronology, we have newly summarized seven magmatic phases in the orefield: (1) Middle–Late Cambrian (506–491 Ma), (2) Early and Middle Ordovician (485–471 Ma and ~462 Ma), (3) Late Ordovician (450–447 Ma), (4) Early Carboniferous and Late-Carboniferous to Early Permian (351–345 and 323–291 Ma), (5) Middle–Late Triassic (244–223 Ma), (6) Early–Middle and Late Jurassic (178–168 Ma and ~150 Ma), and (7) Early Cretaceous (~112 Ma). Three of these seven major magmatic phases were coeval with ore formation, including (1) Early Ordovician (485–473 Ma) porphyry-type Cu–Mo-(Au), (2) Early–Middle Triassic (246–229 Ma) porphyry-related epithermal Au-(Cu–Mo), and (3) Early Jurassic (177–173 Ma) Fe–Cu skarn mineralization. Some deposits in the orefield, notably Tongshan and Zhengguang, were likely formed by more than one mineralization events.
In terms of geochemistry, ore-causative granitoids in the orefield exhibit adakite-like or adakite-normal arc transitional signatures, but those forming the porphyry-/epithermal-type Cu–Mo–Au mineralization are largely confined to the former. The varying but high Sr/Y, Sm/Yb and La/Yb ratios suggest that the ore-forming magmas were mainly crustal sourced and formed at different depths (clinopyroxene-/amphibole-/garnet-stability fields). The adakite-like suites may have formed by partial melting of the thickened lower crust at 35–40 km (for the Early Ordovician arc) and >40 km (for the Middle–Late Triassic arc) depths. The Early Jurassic Fe–Cu skarn orecausative granitoids show an adakitic-normal arc transitional geochemical affinity. These granitoids were likely formed by partial melting of the juvenile lower crust (35–40 km depth), and subsequently modified by assimilation and fractional crystallization (AFC) processes.
In light of the geological, geochronological and geochemical information, we proposed the following tectonometallogenic model for the Duobaoshan orefield. The Ordovician Duobaoshan may have been in a continental arc setting during the subduction of the Paleo-Asian Ocean, and formed the porphyry-related deposits at Duobaoshan, Tongshan and Zhengguang. Subduction may have ceased in the latest Ordovician, and the regional tectonics passed into long subsidence and extension till the latest Carboniferous. This extensional tectonic regime and the Silurian terrestrial-shallow marine sedimentation had likely buried and preserved the Ordovician Duobaoshan magmatic-hydrothermal system. The south-dipping Mongol-Okhotsk Ocean subduction from north of the orefield had generated the Middle–Late Triassic continental arc magmatism and the associated Tongshan PCD and Zhengguang epithermal Au mineralization (which superimposed on the Ordovician PCD system). The Middle Jurassic closure of Mongol-Okhotsk Ocean in the northwestern Amuria block (Erguna terrane), and the accompanying Siberia-Amuria collision, may have placed the Paleo-Pacific subduction system in NE China (including the orefield) under compression, and formed the granodiorite-tonalite and Fe–Cu skarn deposits at Sankuanggou and Xiaoduobaoshan. From the Middle Jurassic, the consecutive accretion of Paleo-Pacific arc terranes (e.g., Sikhote-Alin and Nadanhada) onto the NE Asian continental margin may have gradually distant the Duobaoshan orefield from the subduction front, and consequently arc-type magmatism and the related mineralization faded. The minor Late Jurassic and Cretaceous unmineralized magmatism in the orefield may have triggered mainly by the far-field extension led by the post-collisional (Siberia-Amuria) gravitational collapse and/or Paleo-Pacific backarc-basin opening.

Phase equilibrium modelling of the amphibolite facies metamorphism in the Yelapa-Chimo Metamorphic Complex, Mexico
Fabián Gutiérrez-Aguilar, Peter Schaaf, Gabriela Solís-Pichardo, Gerardo F. Arrieta-García, Teodoro Hernández-Treviño, Carlos Linares-López
2021, 12(1): 293-312. doi: 10.1016/j.gsf.2020.05.001

The Yelapa-Chimo Metamorphic Complex forms part of the Jalisco Block in western Mexico and exposes a wide range of Early Cretaceous metamorphic rocks; such as paragneiss, orthogneiss, amphibolites, and migmatites. However, the pressure–temperature (P–T) conditions of metamorphism and partial melting remain poorly studied in the region. To elucidate metamorphic P–T conditions, phase equilibrium modelling was applied to two sillimanite–garnet paragneisses, one amphibole–orthogneiss, and one amphibolite. Sillimanite–garnet paragneisses exhibit a lepidoblastic texture with a biotite + sillimanite + kyanite + garnet + quartz + plagioclase + K-feldspar mineral assemblage. Amphibole–orthogneiss and amphibolite display a nematoblastic texture with an amphibole + (1) plagioclase + quartz + (1) titanite assemblage and an amphibole + (2) plagioclase + (2) titanite + ilmenite retrograde mineral assemblage. Pseudosections calculated for the two sillimanite–garnet paragneiss samples show P–T peak conditions at ~6–7.5 kbar and ~725–740 ℃. The results for amphibole–orthogneiss and the amphibolite yield P–T peak conditions at ~8.5–10 kbar and ~690–710 ℃. The mode models imply that metasedimentary and metaigneous units can produce up to ~20 vol% and ~10 vol% of melt, respectively. Modelling within a closed system during isobaric heating suggests that melt compositions of metasedimentary and metaigneous units are likely to have direct implications for the petrogenesis of the Puerto Vallarta Batholith. Our new data indicate that the Yelapa-Chimo Metamorphic Complex evolved through a metamorphic gradient between ~23–33 ℃ km-1 and the metamorphic rocks formed at depths between ~22 km and ~30 km with a burial rate of ~2.0 km Ma-1. Finally, the P–T data for both metasedimentary and metaigneous rocks provide new constraints on an accretionary framework, which is responsible for generating metamorphism and partial melting in the YelapaChimo Metamorphic Complex during the Early Cretaceous.

Water in coesite: Incorporation mechanism and operation condition, solubility and P-T dependence, and contribution to water transport and coesite preservation
Wei Yan, Yanyao Zhang, Yunlu Ma, Mingyue He, Lifei Zhang, Weidong Sun, Christina Yan Wang, Xi Liu
2021, 12(1): 313-326. doi: 10.1016/j.gsf.2020.05.007
Abstract(106) HTML PDF(7)

A series of coesite, coexisting with or without a liquid phase, was synthesized in the nominal system SiO2–H2O at 800–1450 ℃ and 5 GPa. Micro-Raman spectroscopy was used to identify the crystalline phase, electron microprobe and LA-ICP-MS were employed to quantify some major and trace elements, and unpolarized FTIR spectroscopy was applied to probe the different types of hydrogen defects, explore water-incorporation mechanisms and quantify water contents. Trace amounts of Al and B were detected in the coesite. Combining our results with the results in the literatures, we have found no positive correlation between the Al contents and the “Al”-based hydrogen concentrations, suggesting that previously proposed hydrogen-incorporation mechanism H+ + Al3+ ↔ Si4+ does not function in coesite. In contrast, we have confirmed the positive correlation between the B contents and the B-based hydrogen concentrations. The hydrogen-incorporation mechanism H+ + B3+ ↔ Si4+ readily takes place in coesite at different P-T conditions, and significantly increases the water content at both liquid-saturated and liquid-undersaturated conditions. For the SiO2–H2O system, we have found that type-I hydrogarnet substitution plays a dictating role in incorporating water into coesite at liquid-saturated condition, type-II hydrogarnet substitution contributes significantly at nearly dry condition, and both operate at conditions in between. The water solubility of coesite, as dictated by the type-I hydrogarnet substitution, positively correlates with both P and T, cH2O = -105(30) + 5.2(32)×P + 0.112(26)×T, with cH2O in wt ppm, P in GPa and T in ℃. Due to its low water solubility and small fraction in subducted slabs, coesite may contribute insignificantly to the vertical water transport in subduction zones. Furthermore, the water solubility of any coesite in exhuming ultra-high pressure metamorphic rocks should be virtually zero as coesite becomes metastable. With an adequately fast waterdiffusion rate, this metastable coesite should be completely dry, which may have been the key factor to the partial preservation of most natural Coe. As a byproduct, a new IR experimental protocol for accurate water determination in optically anisotropic nominally anhydrous minerals has been found. Aided with the empirical method of Paterson (1982) it employs multiple unpolarized IR spectra, collected from randomly-orientated mineral grains, to approximate both total integrated absorbance and total integrated molar absorption coefficient. Its success relies on a high-level orientation randomness in the IR analyses.

Focus Paper
Rhyolites in continental mafic Large Igneous Provinces: Petrology, geochemistry and petrogenesis
Mahesh Halder, Debajyoti Paul, Sarajit Sensarma
2021, 12(1): 53-80. doi: 10.1016/j.gsf.2020.06.011

We present a detailed review of the petrological and geochemical aspects of rhyolite and associated silicic volcanic rocks (up to 20 vol% of all rocks) reported to date from twelve well known Phanerozoic continental mafic Large Igneous Provinces (LIPs). These typically spread over ≤104 km2 (rarely 105 km2 for Paraná-Etendeka) area and comprise ≤104 km3 of extrusive silicic rocks, erupted either during or after the main basaltic eruption within<5 Myr, with some eruption(s) continuing for ≤30 Myr. These rhyolites and associated silicic volcanic rocks (60 81 wt.% of SiO2) are mostly metaluminous to peraluminous and are formed via (i) fractional crystallization of parental mafic magma with negligible crustal contamination, and (ii) melting of continental crust or assimilation and fractional crystallization (AFC) of mafic magma with significant crustal contribution. Rhyolites formed by extensive fractional crystallization are characterized by the presence of clinopyroxene phenocrysts, exhibit steep negative slopes in bivariate major oxides plots and weak to no Nb-Ta anomaly; these typically have temperature >900 ℃. Rhyolites formed by significant crustal contribution are characterized by strong negative Nb-Ta anomalies, absence of clinopyroxene phenocrysts, and are likely to have a magma temperature <900 ℃. Geochemical signatures suggest rhyolite melt generation in the plagioclase stability field with a minor fraction originating from lower crustal depths. A large part of the compositional variability in rhyolites, particularly the SrNd-Pb-O isotope ratios, suggests a significant role of continental crust (upper crustal melting or AFC) in the evolution of these silicic rocks in the continental mafic LIPs.

Early Paleozoic accretionary orogens along the Western Gondwana margin
Sebastián Oriolo, Bernhard Schulz, Silvana Geuna, Pablo D. González, Juan E. Otamendi, Jiří Sláma, Elena Druguet, Siegfried Siegesmund
2021, 12(1): 109-130. doi: 10.1016/j.gsf.2020.07.001

Early Paleozoic accretionary orogens dominated the Western Gondwana margin and were characterized by nearly continuous subduction associated with crustal extension and back-arc basin development. The southwestern margin is represented by Famatinian and Pampean basement realms exposed in South America, both related to the protracted Paleozoic evolution of the Terra Australis Orogen, whereas the northwestern margin is mainly recorded in Cadomian domains of Europe and adjacent regions. However, no clear relationships between these regions were so far established. Based on a compilation and reevaluation of geological, paleomagnetic, petrological, geochronological and isotopic evidence, this contribution focuses on crustal-scale tectonic and geodynamic processes occurring in Western Gondwana accretionary orogens, aiming at disentangling their common Early Paleozoic evolution. Data show that accretionary orogens were dominated by high-temperature/lowpressure metamorphism and relatively high geothermal gradients, resulting from the development of extended/hyperextended margins and bulk transtensional deformation. In this sense, retreating-mode accretionary orogens characterized the Early Paleozoic Gondwana margin, though short-lived pulses of compression/ transpression also occurred. The existence of retreating subduction zones favoured mantle-derived magmatism and mixing with relatively young (meta)sedimentary sources in a thin continental crust. Crustal reworking of previous forearc sequences due to trenchward arc migration thus took place through assimilation and anatexis in the arc/back-arc regions. Therefore, retreating-mode accretionary orogens were the locus of Early Paleozoic crustal growth in Western Gondwana, intimately associated with major flare-up events, such as those related to the Cadomian and Famatian arcs. Slab roll back, probably resulting from decreasing convergence rates and plate velocities after Gondwana assembly, was a key factor for orogen-scale geodynamic processes. Coupled with synchronous oblique subduction and crustal-scale dextral deformation, slab roll back might trigger toroidal mantle flow, thus accounting for bulk dextral transtension, back-arc extension/transtension and a large-scale anticlockwise rotation of Gondwana mainland.

Thematic Section: Big Data and Machine Learning in Geoscience and Geoengineering—Editorial
Thematic Section: Big Data and Machine Learning in Geoscience and Geoengineering—Research Paper
Advanced prediction of tunnel boring machine performance based on big data
Jinhui Li, Pengxi Li, Dong Guo, Xu Li, Zuyu Chen
2021, 12(1): 331-338. doi: 10.1016/j.gsf.2020.02.011

Predicting the performance of a tunneling boring machine is vitally important to avoid any possible accidents during tunneling boring. The prediction is not straightforward due to the uncertain geological conditions and the complex rock-machine interactions. Based on the big data obtained from the 72.1 km long tunnel in the Yin-Song Diversion Project in China, this study developed a machine learning model to predict the TBM performance in a real-time manner. The total thrust and the cutterhead torque during a stable period in a boring cycle was predicted in advance by using the machine-returned parameters in the rising period. A long short-term memory model was developed and its accuracy was evaluated. The results show that the variation in the total thrust and cutterhead torque with various geological conditions can be well reflected by the proposed model. This real-time predication shows superior performance than the classical theoretical model in which only a single value can be obtained based on the single measurement of the rock properties. To improve the accuracy of the model a filtering process was proposed. Results indicate that filtering the unnecessary parameters can enhance both the accuracy and the computational efficiency. Finally, the data deficiency was discussed by assuming a parameter was missing. It is found that the missing of a key parameter can significantly reduce the accuracy of the model, while the supplement of a parameter that highly-correlated with the missing one can improve the prediction.

Non-parametric machine learning methods for interpolation of spatially varying non-stationary and non-Gaussian geotechnical properties
Chao Shi, Yu Wang
2021, 12(1): 339-350. doi: 10.1016/j.gsf.2020.01.011

Spatial interpolation has been frequently encountered in earth sciences and engineering. A reasonable appraisal of subsurface heterogeneity plays a significant role in planning, risk assessment and decision making for geotechnical practice. Geostatistics is commonly used to interpolate spatially varying properties at un-sampled locations from scatter measurements. However, successful application of classic geostatistical models requires prior characterization of spatial auto-correlation structures, which poses a great challenge for unexperienced engineers, particularly when only limited measurements are available. Data-driven machine learning methods, such as radial basis function network (RBFN), require minimal human intervention and provide effective alternatives for spatial interpolation of non-stationary and non-Gaussian data, particularly when measurements are sparse. Conventional RBFN, however, is direction independent (i.e. isotropic) and cannot quantify prediction uncertainty in spatial interpolation. In this study, an ensemble RBFN method is proposed that not only allows geotechnical anisotropy to be properly incorporated, but also quantifies uncertainty in spatial interpolation. The proposed method is illustrated using numerical examples of cone penetration test (CPT) data, which involve interpolation of a 2D CPT cross-section from limited continuous 1D CPT soundings in the vertical direction. In addition, a comparative study is performed to benchmark the proposed ensemble RBFN with two other non-parametric data-driven approaches, namely, Multiple Point Statistics (MPS) and Bayesian Compressive Sensing (BCS). The results reveal that the proposed ensemble RBFN provides a better estimation of spatial patterns and associated prediction uncertainty at un-sampled locations when a reasonable amount of data is available as input. Moreover, the prediction accuracy of all the three methods improves as the number of measurements increases, and vice versa. It is also found that BCS prediction is less sensitive to the number of measurement data and outperforms RBFN and MPS when only limited point observations are available.

Landslide identification using machine learning
Haojie Wang, Limin Zhang, Kesheng Yin, Hongyu Luo, Jinhui Li
2021, 12(1): 351-364. doi: 10.1016/j.gsf.2020.02.012
Abstract(149) HTML PDF(14)

Landslide identification is critical for risk assessment and mitigation. This paper proposes a novel machinelearning and deep-learning method to identify natural-terrain landslides using integrated geodatabases. First, landslide-related data are compiled, including topographic data, geological data and rainfall-related data. Then, three integrated geodatabases are established; namely, Recent Landslide Database (RecLD), Relict Landslide Database (RelLD) and Joint Landslide Database (JLD). After that, five machine learning and deep learning algorithms, including logistic regression (LR), support vector machine (SVM), random forest (RF), boosting methods and convolutional neural network (CNN), are utilized and evaluated on each database. A case study in Lantau, Hong Kong, is conducted to demonstrate the application of the proposed method. From the results of the case study, CNN achieves an identification accuracy of 92.5% on RecLD, and outperforms other algorithms due to its strengths in feature extraction and multi dimensional data processing. Boosting methods come second in terms of accuracy, followed by RF, LR and SVM. By using machine learning and deep learning techniques, the proposed landslide identification method shows outstanding robustness and great potential in tackling the landslide identification problem.

Estimation of diaphragm wall deflections for deep braced excavation in anisotropic clays using ensemble learning
Runhong Zhang, Chongzhi Wu, Anthony T. C. Goh, Thomas Böhlke, Wengang Zhang
2021, 12(1): 365-373. doi: 10.1016/j.gsf.2020.03.003

This paper adopts the NGI-ADP soil model to carry out finite element analysis, based on which the effects of soft clay anisotropy on the diaphragm wall deflections in the braced excavation were evaluated. More than one thousand finite element cases were numerically analyzed, followed by extensive parametric studies. Surrogate models were developed via ensemble learning methods (ELMs), including the eXtreme Gradient Boosting (XGBoost), and Random Forest Regression (RFR) to predict the maximum lateral wall deformation (δhmax). Then the results of ELMs were compared with conventional soft computing methods such as Decision Tree Regression (DTR), Multilayer Perceptron Regression (MLPR), and Multivariate Adaptive Regression Splines (MARS). This study presents a cutting-edge application of ensemble learning in geotechnical engineering and a reasonable methodology that allows engineers to determine the wall deflection in a fast, alternative way.

Application of soft computing techniques for shallow foundation reliability in geotechnical engineering
Rahul Ray, Deepak Kumar, Pijush Samui, Lal Bahadur Roy, A. T. C. Goh, Wengang Zhang
2021, 12(1): 375-383. doi: 10.1016/j.gsf.2020.05.003

This research focuses on the application of three soft computing techniques including Minimax Probability Machine Regression (MPMR), Particle Swarm Optimization based Artificial Neural Network (ANN-PSO) and Particle Swarm Optimization based Adaptive Network Fuzzy Inference System (ANFIS-PSO) to study the shallow foundation reliability based on settlement criteria. Soil is a heterogeneous medium and the involvement of its attributes for geotechnical behaviour in soil-foundation system makes the prediction of settlement of shallow a complex engineering problem. This study explores the feasibility of soft computing techniques against the deterministic approach. The settlement of shallow foundation depends on the parameters γ (unit weight), e0 (void ratio) and CC (compression index). These soil parameters are taken as input variables while the settlement of shallow foundation as output. To assess the performance of models, different performance indices i.e. RMSE, VAF, R2, Bias Factor, MAPE, LMI, U95, RSR, NS, RPD, etc. were used. From the analysis of results, it was found that MPMR model outperformed PSO-ANFIS and PSO-ANN. Therefore, MPMR can be used as a reliable soft computing technique for non-linear problems for settlement of shallow foundations on soils.

Modelling of shallow landslides with machine learning algorithms
Zhongqiang Liu, Graham Gilbert, Jose Mauricio Cepeda, Asgeir Olaf Kydland Lysdahl, Luca Piciullo, Heidi Hefre, Suzanne Lacasse
2021, 12(1): 385-393. doi: 10.1016/j.gsf.2020.04.014

This paper introduces three machine learning (ML) algorithms, the ‘ensemble’ Random Forest (RF), the ‘ensemble’ Gradient Boosted Regression Tree (GBRT) and the MultiLayer Perceptron neural network (MLP) and applies them to the spatial modelling of shallow landslides near Kvam in Norway. In the development of the ML models, a total of 11 significant landslide controlling factors were selected. The controlling factors relate to the geomorphology, geology, geo-environment and anthropogenic effects: slope angle, aspect, plan curvature, profile curvature, flow accumulation, flow direction, distance to rivers, water content, saturation, rainfall and distance to roads. It is observed that slope angle was the most significant controlling factor in the ML analyses. The performance of the three ML models was evaluated quantitatively based on the Receiver Operating Characteristic (ROC) analysis. The results show that the ‘ensemble’ GBRT machine learning model yielded the most promising results for the spatial prediction of shallow landslides, with a 95% probability of landslide detection and 87% prediction efficiency.

Deep learning based classification of rock structure of tunnel face
Jiayao Chen, Tongjun Yang, Dongming Zhang, Hongwei Huang, Yu Tian
2021, 12(1): 395-404. doi: 10.1016/j.gsf.2020.04.003
Abstract(133) HTML PDF(11)

The automated interpretation of rock structure can improve the efficiency, accuracy, and consistency of the geological risk assessment of tunnel face. Because of the high uncertainties in the geological images as a result of different regional rock types, as well as in-situ conditions (e.g., temperature, humidity, and construction procedure), previous automated methods have limited performance in classification of rock structure of tunnel face during construction. This paper presents a framework for classifying multiple rock structures based on the geological images of tunnel face using convolutional neural networks (CNN), namely Inception-ResNet-V2 (IRV2). A prototype recognition system is implemented to classify 5 types of rock structures including mosaic, granular, layered, block, and fragmentation structures. The proposed IRV2 network is trained by over 35,000 out of 42,400 images extracted from over 150 sections of tunnel faces and tested by the remaining 7400 images. Furthermore, different hyperparameters of the CNN model are introduced to optimize the most efficient algorithm parameter. Among all the discussed models, i.e., ResNet-50, ResNet-101, and Inception-v4, Inception-ResNet-V2 exhibits the best performance in terms of various indicators, such as precision, recall, F-score, and testing time per image. Meanwhile, the model trained by a large database can obtain the object features more comprehensively, leading to higher accuracy. Compared with the original image classification method, the sub-image method is closer to the reality considering both the accuracy and the perspective of error divergence. The experimental results reveal that the proposed method is optimal and efficient for automated classification of rock structure using the geological images of the tunnel face.

Stochastic seismic slope stability assessment using polynomial chaos expansions combined with relevance vector machine
Qiu-Jing Pan, Yat-Fai Leung, Shu-Chien Hsu
2021, 12(1): 405-414. doi: 10.1016/j.gsf.2020.03.016

This paper presents probabilistic assessment of seismically-induced slope displacements considering uncertainties of seismic ground motions and soil properties. A stochastic ground motion model representing both the temporal and spectral non-stationarity of earthquake shakings and a three-dimensional rotational failure mechanism are integrated to assess Newmark-type slope displacements. A new probabilistic approach that incorporates machine learning in metamodeling technique is proposed, by combining relevance vector machine with polynomial chaos expansions (RVM-PCE). Compared with other PCE methods, the proposed RVM-PCE is shown to be more effective in estimating failure probabilities. The sensitivity and relative influence of each random input parameter to the slope displacements are discussed. Finally, the fragility curves for slope displacements are established for sitespecific soil conditions and earthquake hazard levels. The results indicate that the slope displacement is more sensitive to the intensities and strong shaking durations of seismic ground motions than the frequency contents, and a critical Arias intensity that leads to the maximum annual failure probabilities can be identified by the proposed approach.

Prediction of load-displacement performance of grouted anchors in weathered granites using FastICA-MARS as a novel model
Hao Shen, Jinhui Li, Sixin Wang, Zewei Xie
2021, 12(1): 415-423. doi: 10.1016/j.gsf.2020.05.004

With the rising needs of better prediction of the load-displacement performance of grouted anchors in an era of developing large-scale underground infrastructures, the existing methods in literature lack an accurate analytical model for the real-life projects or rigorous understanding of the parameters such as grouting pressures. This paper proposes FastICA-MARS as a novel data-driven approach for the prediction of the load-displacement performance of uplift-resisting grouted anchors. The hybrid and data-driven FastICA-MARS approach integrates the multivariate adaptive regression splines (MARS) technique with the FastICA algorithm which is for Independent Component Analysis (ICA). A database of 4315 observations for 479 different anchors from 7 different projects is established. The database is then used to train, validate and compare the FastICA-MARS approach with the classical MARS approach. The developed FastICA-MARS model can provide more accurate predictions than MARS. Moreover, the developed FastICA-MARS model is easy to interpret since the evaluation of the parameter importance of the independent components can be conducted along with the considerations of the correlations with the original variables. It is noteworthy to point out that the grouting pressures play a central role in the proposed model, which is considered of paramount importance in engineering practices but has not been properly taken into account in any prior analytical or empirical predictive models for the load-displacement relationships.

Probabilistic outlier detection for sparse multivariate geotechnical site investigation data using Bayesian learning
Shuo Zheng, Yu-Xin Zhu, Dian-Qing Li, Zi-Jun Cao, Qin-Xuan Deng, Kok-Kwang Phoon
2021, 12(1): 425-439. doi: 10.1016/j.gsf.2020.03.017

Various uncertainties arising during acquisition process of geoscience data may result in anomalous data instances (i.e., outliers) that do not conform with the expected pattern of regular data instances. With sparse multivariate data obtained from geotechnical site investigation, it is impossible to identify outliers with certainty due to the distortion of statistics of geotechnical parameters caused by outliers and their associated statistical uncertainty resulted from data sparsity. This paper develops a probabilistic outlier detection method for sparse multivariate data obtained from geotechnical site investigation. The proposed approach quantifies the outlying probability of each data instance based on Mahalanobis distance and determines outliers as those data instances with outlying probabilities greater than 0.5. It tackles the distortion issue of statistics estimated from the dataset with outliers by a re-sampling technique and accounts, rationally, for the statistical uncertainty by Bayesian machine learning. Moreover, the proposed approach also suggests an exclusive method to determine outlying components of each outlier. The proposed approach is illustrated and verified using simulated and real-life dataset. It showed that the proposed approach properly identifies outliers among sparse multivariate data and their corresponding outlying components in a probabilistic manner. It can significantly reduce the masking effect (i.e., missing some actual outliers due to the distortion of statistics by the outliers and statistical uncertainty). It also found that outliers among sparse multivariate data instances affect significantly the construction of multivariate distribution of geotechnical parameters for uncertainty quantification. This emphasizes the necessity of data cleaning process (e.g., outlier detection) for uncertainty quantification based on geoscience data.

Intelligent modelling of clay compressibility using hybrid meta-heuristic and machine learning algorithms
Pin Zhang, Zhen-Yu Yin, Yin-Fu Jin, Tommy H. T. Chan, Fu-Ping Gao
2021, 12(1): 441-452. doi: 10.1016/j.gsf.2020.02.014

Compression index Cc is an essential parameter in geotechnical design for which the effectiveness of correlation is still a challenge. This paper suggests a novel modelling approach using machine learning (ML) technique. The performance of five commonly used machine learning (ML) algorithms, i.e. back-propagation neural network (BPNN), extreme learning machine (ELM), support vector machine (SVM), random forest (RF) and evolutionary polynomial regression (EPR) in predicting Cc is comprehensively investigated. A database with a total number of 311 datasets including three input variables, i.e. initial void ratio e0, liquid limit water content wL, plasticity index Ip, and one output variable Cc is first established. Genetic algorithm (GA) is used to optimize the hyper-parameters in five ML algorithms, and the average prediction error for the 10-fold cross-validation (CV) sets is set as the fitness function in the GA for enhancing the robustness of ML models. The results indicate that ML models outperform empirical prediction formulations with lower prediction error. RF yields the lowest error followed by BPNN, ELM, EPR and SVM. If the ranges of input variables in the database are large enough, BPNN and RF models are recommended to predict Cc. Furthermore, if the distribution of input variables is continuous, RF model is the best one. Otherwise, EPR model is recommended if the ranges of input variables are small. The predicted correlations between input and output variables using five ML models show great agreement with the physical explanation.

Machine learning for pore-water pressure time-series prediction: Application of recurrent neural networks
Xin Wei, Lulu Zhang, Hao-Qing Yang, Limin Zhang, Yang-Ping Yao
2021, 12(1): 453-467. doi: 10.1016/j.gsf.2020.04.011

Knowledge of pore-water pressure (PWP) variation is fundamental for slope stability. A precise prediction of PWP is difficult due to complex physical mechanisms and in situ natural variability. To explore the applicability and advantages of recurrent neural networks (RNNs) on PWP prediction, three variants of RNNs, i.e., standard RNN, long short-term memory (LSTM) and gated recurrent unit (GRU) are adopted and compared with a traditional static artificial neural network (ANN), i.e., multi-layer perceptron (MLP). Measurements of rainfall and PWP of representative piezometers from a fully instrumented natural slope in Hong Kong are used to establish the prediction models. The coefficient of determination (R2) and root mean square error (RMSE) are used for model evaluations. The influence of input time series length on the model performance is investigated. The results reveal that MLP can provide acceptable performance but is not robust. The uncertainty bounds of RMSE of the MLP model range from 0.24 kPa to 1.12 kPa for the selected two piezometers. The standard RNN can perform better but the robustness is slightly affected when there are significant time lags between PWP changes and rainfall. The GRU and LSTM models can provide more precise and robust predictions than the standard RNN. The effects of the hidden layer structure and the dropout technique are investigated. The single-layer GRU is accurate enough for PWP prediction, whereas a double-layer GRU brings extra time cost with little accuracy improvement. The dropout technique is essential to overfitting prevention and improvement of accuracy.

Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimization
Wengang Zhang, Chongzhi Wu, Haiyi Zhong, Yongqin Li, Lin Wang
2021, 12(1): 469-477. doi: 10.1016/j.gsf.2020.03.007
Abstract(137) HTML PDF(9)

Accurate assessment of undrained shear strength (USS) for soft sensitive clays is a great concern in geotechnical engineering practice. This study applies novel data-driven extreme gradient boosting (XGBoost) and random forest (RF) ensemble learning methods for capturing the relationships between the USS and various basic soil parameters. Based on the soil data sets from TC304 database, a general approach is developed to predict the USS of soft clays using the two machine learning methods above, where five feature variables including the preconsolidation stress (PS), vertical effective stress (VES), liquid limit (LL), plastic limit (PL) and natural water content (W) are adopted. To reduce the dependence on the rule of thumb and inefficient brute-force search, the Bayesian optimization method is applied to determine the appropriate model hyper-parameters of both XGBoost and RF. The developed models are comprehensively compared with three comparison machine learning methods and two transformation models with respect to predictive accuracy and robustness under 5-fold cross-validation (CV). It is shown that XGBoost-based and RF-based methods outperform these approaches. Besides, the XGBoostbased model provides feature importance ranks, which makes it a promising tool in the prediction of geotechnical parameters and enhances the interpretability of model.

Bayesian geological and geophysical data fusion for the construction and uncertainty quantification of 3D geological models
Hugo K. H. Olierook, Richard Scalzo, David Kohn, Rohitash Chandra, Ehsan Farahbakhsh, Chris Clark, Steven M. Reddy, R. Dietmar Müller
2021, 12(1): 479-493. doi: 10.1016/j.gsf.2020.04.015

Traditional approaches to develop 3D geological models employ a mix of quantitative and qualitative scientific techniques, which do not fully provide quantification of uncertainty in the constructed models and fail to optimally weight geological field observations against constraints from geophysical data. Here, using the Bayesian Obsidian software package, we develop a methodology to fuse lithostratigraphic field observations with aeromagnetic and gravity data to build a 3D model in a small (13.5 km×13.5 km) region of the Gascoyne Province, Western Australia. Our approach is validated by comparing 3D model results to independently-constrained geological maps and cross-sections produced by the Geological Survey of Western Australia. By fusing geological field data with aeromagnetic and gravity surveys, we show that 89% of the modelled region has >95% certainty for a particular geological unit for the given model and data. The boundaries between geological units are characterized by narrow regions with<95% certainty, which are typically 400–1000 m wide at the Earth's surface and 500–2000 m wide at depth. Beyond ~4 km depth, the model requires geophysical survey data with longer wavelengths (e.g., active seismic) to constrain the deeper subsurface. Although Obsidian was originally built for sedimentary basin problems, there is reasonable applicability to deformed terranes such as the Gascoyne Province. Ultimately, modification of the Bayesian engine to incorporate structural data will aid in developing more robust 3D models. Nevertheless, our results show that surface geological observations fused with geophysical survey data can yield reasonable 3D geological models with narrow uncertainty regions at the surface and shallow subsurface, which will be especially valuable for mineral exploration and the development of 3D geological models under cover.