지질조건에 따른 강우와 산사태의 특성분석

김경수 ( Kyeong Su Kim ),송영석 ( Young Suk Song ),조용찬 ( Yong Chan Cho ),김원영 ( Won Young Kim ),정교철 ( Gyo Cheol Jeong ) 대한지질공학회 2006 지질공학 Vol.16 No.2

이 연구는 1998년 여름철에 지질조건이 서로 다른 경기도 장흥, 경북 상주 및 포항지역에서 산사태가 많이 발생된 곳의 강우자료와 산사태들을 조사하여 강우조건에 따른 산사태의 특성을 분석하였다. 연구지역들에서는 2~4일간에 걸친 집중호우로 인해 많은 산사태들이 발생되었다. 강우자료는 지역별로 산사태밀집지역을 중심으로 반경 약 50 km 내에 위치한 관측소들에서의 기록을 기초로 하였으며, 산사태자료는 현장조사를 통하여 직접 확인된 산사태만을 대상으로 하였다. 산사태를 유발하는 강우조건은 크게 누적강우량과 강우강도로 구분하고 산사태는 지질조건을 고려한 지역별 발생빈도 등으로 구별하여 이들의 관계를 통계적으로 분석하였다. 산사태들은 집중호우에 의해 발생되었고 국지적으로 밀집되는 양상을 보이는 등 강우와 산사태는 매우 밀접한 연관성을 보였다. 강우를 지배하는 여러 요소 즉, 누적강우량, 강우강도 및 강우지속시간 등이 지역별로 다른 양상을 보였으며, 그에 따른 산사태의 양상이나 발생빈도 또한 각기 다른 것으로 분석되었다. 누적강우량이 많은 지역에서만 산사태가 집중적으로 발생하였고 그 지역을 벗어난 바로 인근지역에서는 산사태가 거의 발생되지 않음으로써 강우가 산사태를 유발하는 매우 중요한 외적 요인으로 간주되고, 누적강우량과 산사태빈도는 서로 상관관계를 갖는 것으로 분석되었다. To study the relationship between rainfall conditions and landslides according to a geological condition in landslides areas such asJangheung Kyounggi, Sangju and Pohang Kyoungbuk, the data of rainfall and landslides are investigated and analyzed. Many landslides occurred at these areas because of the heavy rainfall in two or four days of the summer 1998. The data of rainfall are collected in observatories within a 50km radius from landslides occurrence areas, and the data of landslides are investigated directly in landslides areas. The data of rainfall are the accumulative rainfall and the rainfall intensity, and the data of landslides are the occurrence frequency considering the geological condition. These data are analyzed statistically to know the relationship the rainfall and landslides. The landslides are concentrated in the heavy rainfall area from the analysis of these data. It knows that the landslides are triggered by the heavy rainfall. Meanwhile, the rainfall factors such as the accumulative rainfall, the rainfall intensity and the dropping time are different in each landslides area, and the shape and frequency of landslides are different respectively. The landslides have occurred in the area of high accumulative rainfall, while the landslides have not occurred around that area. Therefore, the rainfall is very important factor induced by the landslides, and the accumulative rainfall is really related to the frequency of landslides.

Assessing landslide hazard and risk: what we do and what we should do

( Fausto Guzzetti ) 대한지질공학회 2019 대한지질공학회 학술발표회논문집 Vol.2019 No.2

Landslides are present in all continents and play an important role in the evolution of landscapes. In many areas they represent a serious threat to people, properties, society and the environment. It is therefore not surprising that landslide hazard and risk assessment, a topic of scientific, technological and practical relevance, is becoming increasingly popular among scientists, practitioners, decision makers and concerned citizens. Despite numerous attempts and unquestionable progresses, the general assumptions and the most popular methodologies used to assess landslide hazard and for risk evaluation have not changed significantly in the last decades. Today, some of these assumptions show theoretical weakness, and the adopted methodologies reveal practical and operational limitations. In the talk, I deal with populations of landslides i.e., numerous landslides triggered in areas from tens to millions of square kilometers by a trigger (e.g., a rainstorm, earthquake, snowmelt event), or by multiple triggers in a period (from weeks to seasons) or in a long period (years to millennia) of time. Following an introduction on what we need to predict to assess landslide hazard and risk in a useful way, I will introduce the strategies and main methods used to detect and map landslides, and to predict populations of landslides in space (“where” landslide occur) and time (“when”, or how frequently they will occur), and the numerosity (magnitude) and size characteristics (length, width, area, volume) of the landslides. For landslide detection and mapping, I will consider traditional methods based on the visual interpretation of aerial photographs, and modern approaches that exploit the visual, semi-automatic or automatic analysis of a variety of remote sensing images. For landslide spatial prediction, I will discuss the results of a global review of classification-based methods for landslide susceptibility assessment, outlining some of the limitations of the current approaches. For the temporal prediction, leveraging on a global analysis of geographical landslide forecasting and early warning systems, I will discuss short term (from hours to days) forecast capabilities and the limitations. Next, I will discuss long term (from years to centuries) landslide projections considering the impact of the current and the expected climate variations on landslide projections. For landslide numerosity and size characteristics, I will present existing statistics of landslide area and volume obtained from large populations of event-triggered landslides. This will be followed by an analysis of the landslide consequences, with an emphasis on the consequences to the population of Italy, including the presentation of a spatial-temporal model of societal landslide risk for Italy. I will end the presentation by offering recommendations on what I think we need to do to make significant progresses in our collective capacity to predict the hazard posed by populations of landslides, and to anticipate and mitigate their risk.

Mechanical Behaviors of Pile-reinforced Landslides under Reservoir Impoundment Using Physical Model Testing

( Xinli Hu ) 대한지질공학회 2019 대한지질공학회 학술발표회논문집 Vol.2019 No.2

Piles have been widely used to stabilize landslides in reservoir area. The interaction between stabilizing piles and landslide body is complicated and is vital for evaluating the long-term safety of landslides and reservoirs. Taking the Majiagou landslide in Hubei Province, China, as a prototype, two physical models with and without piles were constructed with a geometrical similarity ratio of 1:40. The water level in the models was raised and maintained to an elevation of 0.89 m. The deformation, earth pressure, and pore-water pressure of the landslide and the strain of one representative pile were recorded during the tests. The comparison between the model with piles and that without piles shows that: (1) Piles significantly improved the deformation resistance of landslides. Landslides with piles can tolerate a larger deformation before failure and the displacement rate of landslides after failure was reduced; (2) The implantation of piles changed the soil pressure distribution of landslides. The soil pressures at the upslope and around the piles increased while those at the downslope decreased. The distribution of soil pressures along the depth at the uphill and downhill side of the piles changed from triangular to invert triangular and parabolic, respectively; (3) Piles arrested the anticipant slip surface that occurred in an unreinforced landslide, and at least two sliding bands traversed the landslide. (4) Comparison of the pile-reinforced landslide model with those in literature with piles with a much higher stiffness show that the stiffness of piles significantly affects the mechanical interaction between piles and landslides.

지질조건에 따른 자연사면 산사태의 기하학적 특성

김경수 ( Kyeong Su Kim ),송영석 ( Young Suk Song ),채병곤 ( Byung Gon Chae ),조용찬 ( Yong Chan Cho ),이춘오 ( Choon Oh Lee ) 대한지질공학회 2007 지질공학 Vol.17 No.1

자연사면에서의 산사태를 평가함에 있어 발생양상과 기하특성을 파악하는 것은 매우 중요하다. 이 연구는 지질을 달리하는 3개지역(장흥, 편마암류; 상주, 화강암류; 포항, 제3기퇴적암류)에서 발생한 산사태를 대상으로 지질조건에 따른 산사태의 기하양상을 분석하였다. 3개지역에서는 1998년 여름에 150~588 mm의 집중호우로 인해 총 1,582개의 산사태가 발생하였다. 이들 산사태를 조사하여 지질조건별 기하특성을 분석하였다. 산사태는 90% 이상의 사면고도에서 발생빈도가 가장 높고 70% 미만의 고도에서는 매우 낮다. 사면경사는 26~30o 범위에서 가장 많이 발생하였고 20o 미만에서는 발생빈도가 매우 낮으며 이러한 양상은 지질별로 거의 유사한 경향성을 보였다. 규모는 부피로 보아 수십부터 수백를 넘는 경우도 있으며, 그 길이는 5~300 m, 폭은 3~50 m로서 좁고 긴 형태이고 깊이는 대부분 1 m 이내로서 매우 얕은 특징을 보였다. 그리고 유형은 기반암의 표면을 파괴면으로 하여 그 상부에 위치한 토층이 미끄러지는 전이형슬라이드가주를 이루며, 토석류로 이뤄진 사태물질이 하부사면과 계곡부로 이동되면서 유동성산사태로 변화되는 양상을 보였다. The recognitions of geometrical characteristics and occurrence conditions are very important to evaluate the landslides in natural terrains. In this paper, the geometrical characteristics of landslides are analyzed according to a geological condition in three landslides areas. The three landslides areas are classified to the geological condition. The three landslides areas are Jangheung, Sangju and Pohang. The geology of Jangheung area, Sangju area and Pohang area is gneiss, granite, and the tertiary sedimentary rock, respectively. During a heavy rainfall of 150~588 mm in these areas, 1,582 landslides have occurred in 1998. The geometrical characteristics according to the geological condition analyzed from the investigation of these landslides. The frequency of landslide is high exceedingly above 90% of a slope attitude, while the frequency is very low below 70%. The frequency of landslide is high exceedingly between 26o and 30o of slope angle, while the frequency is very low below 20o. The size of the landslides is ranged from several tens to several hundreds . The length is ranged from 5 m to 300 m, and the width is ranged from 3 m to 50 m. Also, the depth is less than 1m. Therefore, the landslides in these areas have small width, long length and shallow depth. The type of the landslides is changed from transitional slide at the scarp to debris flow at the low part of slope.

Deep Neural Network와 Convolutional Neural Network 모델을 이용한 산사태 취약성 매핑

공성현,백원경,정형섭,Gong, Sung-Hyun,Baek, Won-Kyung,Jung, Hyung-Sup 대한원격탐사학회 2022 大韓遠隔探査學會誌 Vol.38 No.6

Landslides are one of the most prevalent natural disasters, threating both humans and property. Also landslides can cause damage at the national level, so effective prediction and prevention are essential. Research to produce a landslide susceptibility map with high accuracy is steadily being conducted, and various models have been applied to landslide susceptibility analysis. Pixel-based machine learning models such as frequency ratio models, logistic regression models, ensembles models, and Artificial Neural Networks have been mainly applied. Recent studies have shown that the kernel-based convolutional neural network (CNN) technique is effective and that the spatial characteristics of input data have a significant effect on the accuracy of landslide susceptibility mapping. For this reason, the purpose of this study is to analyze landslide vulnerability using a pixel-based deep neural network model and a patch-based convolutional neural network model. The research area was set up in Gangwon-do, including Inje, Gangneung, and Pyeongchang, where landslides occurred frequently and damaged. Landslide-related factors include slope, curvature, stream power index (SPI), topographic wetness index (TWI), topographic position index (TPI), timber diameter, timber age, lithology, land use, soil depth, soil parent material, lineament density, fault density, normalized difference vegetation index (NDVI) and normalized difference water index (NDWI) were used. Landslide-related factors were built into a spatial database through data preprocessing, and landslide susceptibility map was predicted using deep neural network (DNN) and CNN models. The model and landslide susceptibility map were verified through average precision (AP) and root mean square errors (RMSE), and as a result of the verification, the patch-based CNN model showed 3.4% improved performance compared to the pixel-based DNN model. The results of this study can be used to predict landslides and are expected to serve as a scientific basis for establishing land use policies and landslide management policies.

집중호우시 발생된 삼척지역의 산사태 특성

송영석 ( Young Suk Song ),장윤호 ( Yoon Ho Jang ),김진석 ( Jin Seok Kim ) 대한지질공학회 2005 지질공학 Vol.15 No.4

본 연구에서는 2002년 태풍 "루사(Rusa)"와 2003년 태풍 "매미(Meami)"에 의한 집중호우시 삼척지역에 발생된 산사태를 조사하여, 대상지역의 강우와 산사태의 관계를 규명하였다. 삼척지역의 연평균강우량은 1,200~1,300㎜이며, 2002년 및 2003년에는 태풍 "루사" 및 "매미"로 인한 집중호우로 인하여 연평균강우량이 2,000㎜이상으로 증가되었음을 알 수 있다. 최대시간강우강도와 2일간 누적강우량이 상대적으로 큰 지역에서 산사태 발생개소 및 붕괴면적이 크게 발생되고 있으므로, 산사태는 강우강도와 누적강우량에 직접적인 영향을 받고 있음을 확인할 수 있다. 삼척지역에서 발생된 산사태는 태풍에 의한 집중호우에 의한 것이므로 산사태 발생당일 최대시간강우강도에 크게 영향을 미치는 것으로 나타났다. 합리적인 산사태 발생규모를 예측하기 위하여 기존의 방법을 종합적으로 적용하여 최대시간강우강도와 산사태 붕괴면적을 모두 고려할 수 있는 새로운 방법을 제안하였다. 기존의 방법 및 새로 제안된 방법을 적용한 결과 삼척지역의 산사태 발생규모는 대규모 산사태임을 알 수 있다. Landslides induced by heavy rainfall from typhoon "Rusa" in 2002 and typhoon "Meami" in 2003 were investigated at Samcheok area, and the relationship between landslides and rainfall on that area was analyzed. The average annual rainfall at Samcheok area is generally 1,200㎜~1,300㎜. However, the average annual rainfall at Samcheok for 2002 and 2003 was increased more than 2,000㎜ because of typhoon "Rusa" and typhoon "Meami". The number of landslides and the landslides area are largely occurred in a area of the relatively high maximum hourly rainfall and 2days cumulative rainfall. Therefore, it confirmed that landslides are directly depended on the hourly rainfall and the cumulative rainfall. The landslides at Samcheok area induced by heavy rainfall due to typhoon are more influenced by the maximum hourly rainfall at the landslide occurrence day. In order to predict a rational landslide size, a new method included the maximum hourly rainfall and the landslide area in a traditional way was proposed. As the result of applying the new proposed method, the landslide size at Samcheok area is involved in the large scale landslide.

Mechanism Peculiarities of Moscow landslides

( Andrey Kazeev ),( German Postoev ),( Ksenia Fedotova ) 대한지질공학회 2019 대한지질공학회 학술발표회논문집 Vol.2019 No.2

There are 15 sites with large deep landslides on Moscow territory. They have a complex mechanism of development, have a huge mass, and cause big problems in implementation of protection measures. The catastrophic phase (the moment of separation of a new landslide block from the native ground mass) in the cycle of development of deep compression-extrusion landslides was investigated in detail. During this phase, the main features of the mechanism of landslides of considered type are observed. The engineering geological conditions were analyzed of the territory with deep landslides. The scale and dynamics of deformations were investigated. The slip surface depths were instrumentally determined on sites. The reasons of two catastrophic activations of deep landslides occurred in 2006 and 2009 in Moscow were investigated. The dependences between different geomorphological parameters were received, which represent very informative characteristics of the landslide process. The landslide hazard assessment and reliable prediction of possible landslide activation, as well as the implementation protection measures, are impossible without clear understanding of the landslide mechanism, conditions and mechanics of the stress-strain state changes in a landslide mass prior to reaching their limit values. Neither they are possible without taking into account the dynamics of the landslide processes and the criteria that mark transition of the landslide development into a catastrophic phase.

Failure Modes of Large-scale Landslides Induced by 2009 Typhoon Morakot in Southern Taiwan

( Keng-hao Kang ),( Hao-hsun Lin ),( Yii-wen Pan ),( Jyh-jong Liao ) 대한지질공학회 2019 대한지질공학회 학술발표회논문집 Vol.2019 No.2

After the catastrophic landslide at the Shiaolin village during the 2009 Typhoon Morakot, many large-scale landslide areas were delineated using LiDAR DEM. Among those, 356 large-scale landslides caused by the Typhoon Morakot are in southern Taiwan. In addition to the delineation task, the assessment of landslide activity is also essential for risk management. This study aims to classify the failure modes of large scale landslides in order to provide information for further assessment. For remote sensing interpretation, this study adopted google satellite images, multi-period aerial photos, 5x5m/1x1m DEM, and geologic maps. Arc GIS was used to calculate the geomorphic index of the landslides before and after 2009 Typhoon Morakot. We used a variety of parameters including drainage system, shape of bare areas, and characteristics of topography to identify the failure modes of landslides. As a result, this study classify the large-scale landslides induced by the Typhoon Morakot into six categories: colluvium slide, dip slope slide, weathered bedrock slide, rock avalanche, shallow debris slide, and complex type. Among those cases, ten sites are possibly related with gravitational slope deformation.

Assessment of geological hazards in landslide risk using the analysis process method

Peixi Guo,Seyyed Behnam Beheshti,Maryam Shokravi,Amir Behshad 국제구조공학회 2023 Steel and Composite Structures, An International J Vol.47 No.4

Landslides are one of the natural disasters that cause a lot of financial and human losses every year It will be all over the world. China, especially. The Mainland China can be divided into 12 zones, including 4 high susceptibility zones, 7 medium susceptibility zones and 1 low susceptibility zone, according to landslide proneness. Climate and physiography are always at risk of landslides. The purpose of this research is to prepare a landslide hazard map using the Hierarchical Analysis Process method. In the GIS environment, it is in a part of China watershed. In order to prepare a landslide hazard map, first with Field studies, a distribution map of landslides in the area and then a map of factors affecting landslides were prepared. In the next stage, the factors are prioritized using expert opinion and hierarchical analysis process and nine factors including height, slope, slope direction, geological units, land use, distance from Waterway, distance from the road, distance from the fault and rainfall map were selected as effective factors. Then Landslide risk zoning in the region was done using the hierarchical analysis process model. The results showed that the three factors of geological units, distance from the road and slope are the most important have had an effect on the occurrence of landslides in the region, while the two factors of fault and rainfall have the least effect The landslide occurred in the region.

Simulation of Landslides Impact on the Building

( Valentina Burova ) 대한지질공학회 2019 대한지질공학회 학술발표회논문집 Vol.2019 No.2

In Russian Federation more than 50% of the cities are located in the landslides-prone areas. The paper describes the results of landslides simulation on buildings with application of Automatic Design System PLAXIS 3D, which allows an additional subsidence of basements and foundations to be estimated when exposed to landslides. Two principle schemes have been analyzed. The first one, the landslide is localized above the object along the slope and acts on it from the outside. The second one, the landslide is formed within the compressible thickness of the object basement. On the whole, 17 variants of computations have been made taking into account different impact schemes and buildings’ basements. The results of simulation have been compared with maximum permissible additional deformations for buildings of different structural types and their states. Analysis of the simulation results showed the excess of permissible additional subsidence is typical for the second scheme when the landslide is located within the compressible thickness of the concrete frame building basement with column foundation. For this case, the curves of relationship between of additional subsidence of basements and landslides of various sizes were constructed. This study was supported by the Russian Science Foundation, project no. 16-17-00125