• Geophysics and Geophysical Exploration
• /
• v.23 no.3
• /
• pp.157-167
• /
• 2020

With the recent advancement of computer hardware and the contribution of open source libraries to facilitate access to artificial intelligence technology, the use of machine learning (ML) and deep learning (DL) technologies in various fields of exploration geophysics has increased. In addition, ML researchers have developed complex algorithms to improve the inference accuracy of various tasks such as image, video, voice, and natural language processing, and now they are expanding their interests into the field of automatic machine learning (AutoML). AutoML can be divided into three areas: feature engineering, architecture search, and hyperparameter search. Among them, this paper focuses on hyperparamter search with Bayesian optimization, and applies it to the problem of facies classification using seismic data and well logs. The effectiveness of the Bayesian optimization technique has been demonstrated using Vincent field data by comparing with the results of the random search technique.

• 한국지구과학회:학술대회논문집
• /
• 2010.04a
• /
• pp.20-20
• /
• 2010

강원도 태백 지역에 위치한 조선누층군의 세송층을 대상으로 암상변화, 해수면 변화, 안정탄소동위원소 층서를 조사하였다. 세송층은 중기에서 후기 캄브리아기에 속하는 지층으로서 직동지역과 사군다리지역에서 조사하였다. 직동지역의 세송층은 수평층리 이회암상, 리본 석회암상, 단괴 석회암상, 수평층리 사암상, 석회역암상으로 구성되어 있으며, 이들은 내부완사면, 중부완사면, 외부완사면의 세개의 상조합으로 분류되었다. 세송층의 하부로부터 약 30m 까지는 리본석회암상과 석회역암상, 일부 단괴석회암상이 나타나는 구간으로 중부완사면 환경에서 퇴적된 것으로 해석되며, 약 30m부터 42m 까지는 점차 이회암상의 비율이 증가하여 상향 세립화의 경향성을 보이는 구간으로 외부완사면 환경에서 퇴적된 것으로 해석된다. 사군다리 지역은 세송층의 상부가 나타나는 곳으로 평행엽층리사암상과 단괴석회암상이 관찰되며 중부완사면 환경에서 퇴적된 것으로 해석된다. 세송층의 석회암은 하부 이암, 중부 와케스톤, 상부 와케스톤과 팩스톤으로 변화하는 것으로 관찰되었다. 세송층에서의 안정탄소동위원소 잘량 분석결과 양의 값(2.5-3.0‰)을 보이는 Peak 구간이 하부로부터 25m 높이 부근에서 존재하는 것으로 나타났다. 이러한 분석결과는 Steptoean 시기(약 496 Ma)에 전 지구적으로 발생했던 안정탄소동위원소 값의 양(positive)의 변동을 반영한 SPICE(Steptoean positive carbon isotope excursion) 구간에 해당되는 것으로 해석된다. SPICE 구간은 세송층의 중상부에 해당하는 곳으로 와케스톤, 팩스톤의 석회암을 포함하는 석회역암상, 단괴 석회암상으로 구성되어 있으며 이는 상대적으로 해수면이 낮았던 시기를 지시한다. 이러한 SPICE는 큰 규모의 전 지구적 탄소순환의 변동을 대표하며 대륙 간의 대비를 용이하게 하고 퇴적당시의 환경적 변화를 해석하는데 도움을 준다. 북중국과 북미 Laurentia의 안정탄소동위원소 값을 비교해보면 두 지역 모두 세송층과 유사한 값을 보인다. 다만 북중국 지역은 Chuangia Zone에서 SPICE가 나타나는 반면 세송층은 Prochuangia Zone에서 SPICE가 나타난다는 차이가 있다.

• 한국해양학회지
• /
• v.22 no.1
• /
• pp.1-8
• /
• 1987

The Sinyangri Formation crops out in the vicinity of the Seongsan Peninsula, Cheju Island. Based on sedimentary structures, texture and composition, the lithologic sequence has been classified, in ascending stratigraphic order, into three lithofacies: parallel laminated sandstone facies (Facies I): conglomerate facies(Facies Il); and cross stratified sandstone facies (F acies Ill). Wedge-to-parallel, seaward-inclined in low angle less than 10$^{\circ}$lamina -sets with alternations of coarseand fine-grained sediments in the Facies I are the characteristic sedimentary structures in the foreshore depositional environment. Grains of this faciesare well sorted with good roundness compared with the other two facies, partly showing inverse graded bedding. Facies II,largely composed of claset-supported,very poorly-sorted conglomerates,does not pinch out but occurs continuously along the Sinyangri beach.Interstitial spaces between the clasts are mostly infilled with volcanic-ash and small amounts of well-rounded shell fragments.Maximum bed thickness as well as the size of imbedded basaltic clasts decreases to the south(toward Sinyangri). Large clasts with parallel lamination originated from the underlying Facies i,are generally elongated parallel to the bedding plane and display no systematic horizontal variations in size indicative of in-situ clasts.In view of the facts above it seems that large gravels from the basaltic rocks are transgressive lag conglomerates which are partly affected by the combination of longshore currents and propagating wave.Local occurrence of cross-strata dipping toward the south in the upper part of Facies IIreinforces the evidence of the action of longshore currents. Facies IIIis characterized by bidirectional trough cross-starifiction and wave ripples associated with the upper shoreface(surfzone) environments.In summary,the Sinyangri Formation represents the depositional environments of foreshore to upper shoreface truncated by disconformity between Facies Iand II.

• The Journal of Engineering Geology
• /
• v.8 no.2
• /
• pp.163-173
• /
• 1998

Suite of log analysis techniques consisting of geophysical well log, geological core log, and physical core log have been made to understand the well log responses and to determine the lithology of a test borehole-PABH1 located in Pungam sedimentary basin, Sosok, Hongchon-gun, Kangwon Province. Geological core logging has been precisely made over the cores taken between 64 and 124 meters, and 11 groups of rock types were deduced. Using the core samples divided by 11 groups, geophysical property measurements consisting of resistivity, natural gamma and density were made. Each rock group in the area is shown to have its characteristic physical response from geophysical well log and geophysical core logs. The outstanding physical responses particularly shown from siltstone, coarse sandstone to conglomerate, and granitic gneiss in the area were effectively used as keybeds in correlating the geophysical well logs to the result of geological core logs.

• Proceedings of the Mineralogical Society of Korea Conference
• /
• 2002.05a
• /
• pp.115-117
• /
• 2002

• The Journal of the Petrological Society of Korea
• /
• v.22 no.4
• /
• pp.263-272
• /
• 2013

We investigated the lithology and petrography of granites and metasedimentary rocks in Deokjeok Island at the western margin of the Gyeonggi massif. The major lithology comprises the biotite granite that intrudes all other types of rocks. A minor amount of mylonitized porphyritic granite crops out along the southeastern coast. Metasedimentary rocks in the north are further divided into: (1) sheared quartzite-schist to the northeast; and (2) relatively less-deformed, low-grade metasedimentary rocks to the northwest. The former contains quartz grains showing undulatory extinction and subgrain aggregates as well as minor amount of primary chlorite and biotite in the muscovite-rich matrix. Metamorphic condition belongs to the greenschist facies or the biotite zone. On the other hand, the latter unit consists of meta-conglomerate, meta-sandstone, meta-pelite, and black slate. Regardless of the lithology, the intensity of deformation apparently increases eastward to develop the flow banding of quartz in the shear zone.

• Journal of the Mineralogical Society of Korea
• /
• v.18 no.3 s.45
• /
• pp.165-181
• /
• 2005

The granitic rocks distributed in the southern part of the Yangsan Fault are classified into five distinct rock facies based on the field relation, petrography and geochemical characteristics. These five different rock facies can be grouped into two considering their origins. Group I, which reveals various evidences of magma mixing, includes three rock facies of granodiorite, enclave-rich porphyritic granite, and enclave-poor porphyritic granite. Group H intruding Croup I includes equigranular granite and micrographic granite with no evidence of magma mixing. It is suggested that the distinctively different trace element and isotopic chemistries between group I and II, support evolution from the different parental magma. It is suggested that the three rock facies in group I were generated by different degrees of magma mixing in addition to fractionation of plagioclase. MMEs experienced fractionation of biotite. The two facies in group H seem to have been generated from different parent magma from group I and evolved by fractionation of K-feldspar. The Rb-Sr whole-rock ages of the group I rocks yield $59.2\~58.9Ma$, and those of the group II rocks give 53. $3\~51.7Ma$, regardless of their distribution whether they occur in the eastern or western parts of the Yangsan Fault. Based on Sm-Nd isotope compositions, depleted mantle model ages $(T_2DM)$ of the group I range $0.8\~0.9Ga$, while those of the group II$0.6\~0.7Ga$.

• Geophysics and Geophysical Exploration
• /
• v.23 no.3
• /
• pp.131-140
• /
• 2020

In training a classification model using machine learning, the acquisition of training data is a very important stage, because the amount and quality of the training data greatly influence the model performance. However, when the cost of obtaining data is so high that it is difficult to build ideal training data, the number of samples for each class may be acquired very differently, and a serious data-imbalance problem can occur. If such a problem occurs in the training data, all classes are not trained equally, and classes containing relatively few data will have significantly lower recall values. Additionally, the reliability of evaluation indices such as accuracy and precision will be reduced. Therefore, this study sought to overcome the problem of data imbalance in two stages. First, we introduced weighted accuracy and weighted precision as new evaluation indices that can take into account a data-imbalance ratio by modifying conventional measures of accuracy and precision. Next, oversampling was performed to balance weighted precision and recall among classes. We verified the algorithm by applying it to the problem of facies classification. As a result, the imbalance between majority and minority classes was greatly mitigated, and the boundaries between classes could be more clearly identified.

• Geophysics and Geophysical Exploration
• /
• v.16 no.4
• /
• pp.268-274
• /
• 2013

Continuous seismic observation of volcanic earthquakes related to migration of magma or gas, is one of the most important methods used to monitor active volcanoes. In this review paper, we introduce the characteristics of volcanic earthquakes and their classification based on locations of foci, waveforms, frequencies, lithology, and source mechanisms. We also present some examples of successful warnings of volcanic eruptions and the present status of some observations of volcanic earthquakes in Japan, the United States, Italy, and New Zealand.

• 한국지구물리탐사학회:학술대회논문집
• /
• 1999.08a
• /
• pp.156-175
• /
• 1999

Recently the application of geophysical well logging to geotechnical site investigation is increasing, because the merit that geophysical logs provide the high resolution and in-situ physical properties in volumes of rock surrounding the borehole. Geophysical well logs are used to identify lithologic boundaries and fracture, to determine the physical properties of rock(i.e., density, velocity etc.), and to detect permeable fracture zones that could be conduits for ground water movement through the rocks. The principle of heat-pulse meter, the calibration of gamma-gamma logging, and principles and data processing of full waveform sonic logging are briefly reviewed, and the case studies of geophysical logs are discussed. Correlation between velocity by sonic logging and rock mass classification such as RMR(Rock Mass Rating) value is considered.