• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.97-97
• /
• 2023

최근 재해분야에서 인공신경망(ANN), 기계학습(ML), 딥러닝(DL) 등 AI 기술이 활용성이 점차 증가하고 있으며, 센싱정보와 연계한 시설물 안전관리, 원격탐사와 연계한 재해감시(녹조, 산사태, 산불 등), 수문시계열(수위, 유량 등) 예측, 레이더·위성강수 자료의 보정과 예측, 상하수도 관망누수예측 등 다양한 분야에서 AI 기술이 적용되고 그 활용성이 검증된 바 있다. 본 연구에서는 ML, DL, 물리기반신경망(Pysics-informed Neural Networks, PINNs)을 이용한 다양한 재해분석 사례를 소개하고, 그 활용성과 한계에 대해서 논의하고자 한다. 주요사례로는 (1) SAR영상과 기계학습을 이용한 재해피해지역(울진 산불) 감지, (2) 국가 디지털 정보를 이용한 산사태 위험지역 판별(인제 산사태) (3) 기계학습 및 딥러닝 기법을 이용한 위성강수 자료의 보정·예측 및 유출해석, (4) 수리해석을 위한 수치해석분야에서의 PINNs의 적용성(1차원 Saint-Venant 식 해석) 평가 연구결과를 공유한다. 특히, 자료의 입·출력 자료만으로 학습된 인공신경망 모형 대신 지배방정식(물리방정식)을 만족하도록 강제한 PINNs의 경우, 인공신경망 모형보다 우수한 모의능력을 보여주었으며, 향후 복잡한 수리모델링 등 수치해석분야에서 그 활용가능성이 매우 높을 것으로 판단된다.

• Journal of Korean Society for Geospatial Information Science
• /
• v.2 no.2 s.4
• /
• pp.43-53
• /
• 1994

GSIS(geo-spatial information system), particularly when utilized in hazard management decision, is one of hazard analysis tool. Data of GSIS input from digitizing or scanning of map or aerial photos. This paper focuses upon the hazard prediction in GSIS and RS analysis to assess map, aerialphotos, satellite imagery and soil map. This study found computation of hazard area analysis. the results is formed as raster data model of quadtree. Authors knew more accurate results of overlay. This paper shows building up integrated data base as well as search of hazard area in aerial photographs.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.42 no.3
• /
• pp.359-369
• /
• 2022

In slope stability analysis for landslides, mountains have various slopes and geographical features, and hence it is necessary to estimate stability using rigorous analysis methods. In this study, after the analysis of infiltration behavior through unsaturated layers due to rainfall, the stability of landslide was estimated to account for the variation of pore water pressures. In the analysis of slope stability, a three-dimensional slope analysis was compared with an infinite slope analysis in a case study of terrain in which an actual landslide occurred. In the three-dimensional slope stability analysis, it was found that the location of the failure and the failure area were predicted accurately based on the detailed geological information despite the variation of geographical features.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.22 no.3
• /
• pp.152-163
• /
• 2020

The present study introduces a procedure to prepare and manage a high-resolution rainfall and soil moisture (SM) database in the LAMP prediction system, especially for landslide researchers. The procedure also includes converting the data into spatial resolution suitable for their interest regions following proper map projection methods. The LAMP model precipitation and SM data are quantitatively and qualitatively evaluated to identify the model prediction characteristics using the ERA5 reanalysis precipitation and observed 10m depth SM data. A detailed process of converting LAMP Weather Research and Forecasting (WRF) output data for 10m horizontal resolution is described in a step-wise manner, providing technical convenience for users to easily convert NetCDF data from the WRF model into TIF data in ArcGIS. The converted data can be viewed and downloaded via the LAMP website (http://df.ncam.kr/lamp/index.do) of the National Center for AgroMeteorology. The constructed database will contribute to monitoring and prediction of landslide risk prior to landslide response steps and should be data quality controlled by more observation data.

• Korean Journal of Remote Sensing
• /
• v.26 no.1
• /
• pp.47-57
• /
• 2010

The aim of this study is to detect landslide using digital aerial photography and apply the landslide to landslide susceptibility mapping by artificial neural network (ANN) and geographic information system (GIS) at Jinbu area where many landslides have occurred in 2006 by typhoon Ewiniar, Bilis and Kaemi. Landslide locations were identified by visual interpretation of aerial photography taken before and after landslide occurrence, and checked in field. For landslide susceptibility mapping, maps of the topography, geology, soil, forest, lineament, and landuse were constructed from the spatial data sets. Using the factors and landslide location and artificial neural network, the relative weight for the each factors was determinated by back-propagation algorithm. As the result, the aspect and slope factor showed higher weight in 1.2-1.5 times than other factors. Then, landslide susceptibility map was drawn using the weights and finally, the map was validated by comparing with landslide locations that were not used directly in the analysis. As the validation result, the prediction accuracy showed 81.44%.

• Korean Journal of Remote Sensing
• /
• v.33 no.6_2
• /
• pp.1047-1060
• /
• 2017

In recent years, global warming has been continuing and abnormal weather phenomena are occurring frequently. Especially in the 21st century, the intensity and frequency of hydrological disasters are increasing due to the regional trend of water. Since the damage caused by disasters in urban areas is likely to be extreme, it is necessary to prepare a landslide susceptibility maps to predict and prepare the future damage. Therefore, in this study, we analyzed the landslide vulnerability using the logistic model and assessed the management plan after the landslide through the field survey. The landslide area was extracted from aerial photographs and interpretation of the field survey data at the time of the landslides by local government. Landslide-related factors were extracted topographical maps generated from aerial photographs and forest map. Logistic regression (LR) model has been used to identify areas where landslides are likely to occur in geographic information systems (GIS). A landslide susceptibility map was constructed by applying a LR model to a spatial database constructed through a total of 13 factors affecting landslides. The validation accuracy of 77.79% was derived by using the receiver operating characteristic (ROC) curve for the logistic model. In addition, a field investigation was performed to validate how landslides were managed after the landslide. The results of this study can provide a scientific basis for urban governments for policy recommendations on urban landslide management.

• Geotechnical Engineering
• /
• v.15 no.10
• /
• pp.16-26
• /
• 1999

• Journal of the Korean Geotechnical Society
• /
• v.35 no.10
• /
• pp.17-31
• /
• 2019

This study describes a prediction method for rainfall-induced landslides and subsequently debris flows in a regional scale areas. Special attention is given to the calculation of the propagation of debris flows by considering rainfall infiltration into soil slopes and soil entrainments by debris flows. The proposed method was verified by comparing the analytical results and the measured ones reported by the previous research. As a result, predictions and observations were quite similar in terms of the front position, the velocity, volume and momentum of debris flows. Even when applied to natural mountain slope with complicated terrain, numerical results and observations were similar. At last, the combined analysis of landslides and debris flows were conducted. The landslides prediction showed a predictive rate of about 83%, and the result of the final volume of debris flow showed an error rate of 3%. As a result, the proposed combined method for landslides and debris flows overcomes the problem of separating the landslides analysis and the debris flows simulation. Especially, the proposed method can analyze the effects of rainfall on entrainments by debris flows as well as rainfall-induced landslides and the behavior of debris flows.

• Proceedings of the Korean Association of Geographic Inforamtion Studies Conference
• /
• 2010.06a
• /
• pp.261-266
• /
• 2010

그 동안 산사태 연구는 강우에 의한 예측 연구나 결정론적 유한요소법 및 한계평형 연구가 많이 이루어 졌지만 자연의 요소 이외의 인간 활동적 요소에 대한 연구는 외국의 사례에 비해 선행연구가 많이 없는 실정이다. 이번 연구 논문에서는 산림청의 사면평가표를 GIS을 이용해 강우 조건과 사회적 인자(도로, 빌딩, 주택 등)에 따른 산사태 위험지역들을 GIS Map상에서 어떤 차이가 나는지 알아보았다. 또한 사면평가표의 항목에서 사회적 인자 항목이 산사태 발생에 있어 통계적으로 유의한지를 다변량 통계분석을 통해 알아보았다.

• Proceedings of the Korean Society of Disaster Information Conference
• /
• 2016.11a
• /
• pp.428-432
• /
• 2016

자연재해는 전 세계적으로 몇 조원에 이르는 피해를 주고 있으며 근래에는 기후변화로 인한 이상기후현상으로 집중호우의 빈도가 증가하고 있으며 홍수, 산사태 토석류 등 발생빈도가 늘어나고 있는 추세이다. 최근 2011년 서울에서 발생한 우면산 토석류의 피해와 강원도 일부 지역의 피해로 인해 많은 피해가 일어났으며, 국가에서는 산사태 피해를 최소화하기 위해서는 장기적으로는 위험한 지역에 구조물을 설치하는 등 예방사방을 실시하고, 단기적으로는 위험요소의 제거와 위험지역의 예측을 통한 경계피난체재를 철저히 할 필요성이 강조되고 있는 실정이다. 특히, 강원도는 산지로 이루어져 있으며, 매년 집중호우나 태풍이 지나간 다음, 많은 양의 강우량으로 인해 산길을 지나는 도로에 토석류로 인해 길이 막혀 이동이 불가능 하게 되거나 산지 아래에 있는 건물이 산에서 쓸려서 내려오는 토사와 나무에 의해 파괴되어 물질적인 피해를 입는 영상이 뉴스를 통해 많이 접하고 있다. 본 연구에서는 토석류에 대한 전반적인 특성을 설명하고 체계적인 관리가 필요하다. 연구 지역은 춘천지역을 대상으로 SINMAP 모델 기법의 조도계수와 특성을 고려한 위험지도를 작성하였고 산사태 일부 지역에 대한 위험도 평가를 하였다.