Earticle

The Barind tract is threatened by desertification and undergoing rapid change. In view of this fact it is very much essential to manage this barind tract under proper land-use plan. The present study evaluates the effectiveness of high-resolution satellite data and computer aided GIS techniques in assessing land-use change detection for the period 1990 to 2007 within the study area, which is very much essential to manage this barind tract under proper land-use plan, and for proper land-use plan it is necessary to get reliable information. The present study found five major land-use such as current fallow, current agriculture, settlement, irrigation water and water bodies. From the result, it is found that current fallow and water bodies decrease while settlement and current agriculture increase. Study concludes that as Barind tract is threatened by desertification, decrease of water bodies is not a good sign for the study area.

항만, 호소, 하천의 준설 공사, 수저 골재 채취 공사, 준설물질이나 골재의 운송, 보관, 혹은 해양 투기 과정 등에서 준설물질이나 퇴적물이 수중으로 유출되어 수 환경을 오염시키는 경우가 생길 수 있다. . 수 환경에 유출된 부유물질을 관측하는 여러 기술들 중 위성자료를 활용한 관측기술의 국외 활용현황을 살펴보았다. 좁은 해역에서 발생하고 확산되는 준설토사의 흐름을 파악하기 위해서는 높은 해상도를 가진 위성센서가 주로 사용되며 시간적 변화를 감시하기 위해 재방문주기가 짧은 위성 센서의 자료가 유용하다. 따라서 NASA의 Terra와 Aqua 위성에 각각 설치된 Moderate Resolution Imaging Spectroradiometer (MODIS) 영상이 주로 사용되었다. MODIS는 250m*250m 해상도를 가지며 같은 지역을 매일 촬영하는 장점이 있다. Landsat 위성자료는 해상도가 높지만 (30m*30m) 재방문 주기가 16일로 부유물질 플룸의 시간에 따른 변이를 관측하기에 적절하지 않다. 물론 보조적인 자료로는 사용할 수 있다. 앞으로 우리나라에서는 하루에 최대 8회 관측하는 GOCI 위성을 많이 활용하게 될 것이다.

Sediment resuspension during dredging operation for maintenance, remediation of contaminated sediment or excavation of marine aggregate from the bottom sediment and subsequent movement of the resuspended sediment plume to the outside of the project area in river and sea are of great interest for the protection of aquatic environment. In order to monitor development and spreading of sediment plume, a potentially affected area should be observed simultaneously as often as possible. For this purpose, the remote sensing imagery has been successfully used in many countries as satellite sensors meet the above mentioned requirements. The most widely used remote sensing imagery is from Moderate Resolution Imaging Spectroradiometer (MODIS) band 1 (620-670nm) reflectance data with spatial resolution of 250nm. When monitoring dredging plumes, images from Landsat satellites having a repeat capture of every 16 days are used to assist the interpretation of MODIS images. A few Korean examples are shown here for future reference for Geostationary Ocean Color Imager (GOCI) uses from marine aggregate mining area.

High-resolution remote sensing image scene classification is a challenging visual task, and this study proposes a remote sensing image scene classification method based on Semantic Multi-Granularity Feature Learning Network (SMGFL-Net). The core idea is to learn global and multi-granularity local features from rearranged intermediate feature mappings, thus eliminating meaningless edges. These features are then fused into the final prediction. Through comparative studies, SMGFL-Net consistently outperforms other peer methods in terms of classification accuracy.

이 연구에서는 무인항공기에 탑재한 라이다(LiDAR) 센서를 활용해 공산성 일대의 정밀 지형 자료를 획득하고, 이를 지리정보시스템으로 분석해 지형과 수문학적 특성을 규명하였으며, 성벽의 위치와 길이 및 높이 등 정량적 정보를 원격으로 추출하였다. 이 결과, 공산성 성벽은 대부분 봉우리 나 능선과 같은 자연지형을 따라 축성하였으나, 일부 구간에서는 성벽을 관통하는 수계망이 있고 안쪽으로 집수역이 형성되는 등 배수 체계에 취약한 지점이 확인되었다. 이러한 장소는 과거 성벽의 붕괴가 발생했거나 구조적 변형으로 보수가 있었던 곳과 일치하는 것으로 보아 지형적 요인이 성벽 의 안정성에 영향을 미치는 요소로 작용함을 지시한다. 또한 사진과 음영기복도에서는 식별이 어려 운 성벽 위치를 곡률분석을 통해 명확히 추출하였으며, 성벽 하단부와 상단부의 고도 차이를 토대로 높이를 산출하였다. 특히 성벽의 높이가 높거나 낮게 나타난 지점에서 높은 상관도를 보였다. 한편 전체 성벽은 석성구간 1,936m와 토성구간 676m로 총 2,612m 길이를 가지나, 측정 간격에 따라 다소 왜곡이 발생하는 것으로 나타났다. 이는 문화유산의 제원 측정에 대한 기준 척도의 필요성을 시사하 는 것이다. 이 연구는 접근이 어려운 문화유산의 정량적 형상정보 구축에 드론라이다와 지리정보분 석 기법의 뛰어난 효용성을 보여주며, 이는 향후 지형적 위험 요소의 사전 평가와 보존관리계획 수립에 유용하게 활용될 수 있을 것이다.

This study acquired high-resolution topography using UAV-mounted LiDAR around the Gongsanseong Fortress and analyzed it in a GIS to characterize hydrological settings, and remotely extract quantitative attributes of the fortress wall, including its position, length and height. As a result, the Fortress walls were primarily constructed along natural topography, as peaks and ridges. However, certain sections, exhibited drainage vulnerabilities identified, such as water system penetrating the walls or the formation of water catchments within them. These locations tend to coincide with areas where the wall collapsed in the past or required repair due to structural deformation, indicating that topographic factors influence wall stability. Furthermore, curvature analysis allowed us to clearly identify locations of the Fortress that were difficult to discern in photographs and shaded relief maps. Wall heights were calculated based on the elevation difference between the upper and lower sections of the Fortress. A high correlation was observed, particularly at points where the Fortress appeared higher or lower. Meanwhile, the Fortress wall measures 2,612m long, with a stone section of 1,936m and an earthen section measuring 676m. However, some distortion occurred depending on the measurement interval. These suggest the need for a standardized scale in measuring the dimensions of cultural heritage sites. This study demonstrates the outstanding utility of UAV–LiDAR combined with GIS analysis techniques in establishing quantitative configuration information for cultural heritage sites that are difficult to access, and can be usefully utilized in the future for preliminary assessment of topographic risk factors and the development of conservation management schemes.

모니터링 기반 손상 진단은 체계적이고 효율적인 문화유산의 보존 관리를 위해 필수적 이다. 이를 위해 문화유산의 상시 모니터링이 가능하도록 석탑의 손상을 자동으로 탐지하고 시각화하는 딥러닝 시스템을 제안한다. 석탑 이미지에서 손상을 픽셀 단위로 탐지하고 시각화 하기 위해 Mask R-CNN을 활용하였으며, 우리나라 석탑에 특화된 데이터세트를 구축하여 딥러 닝 모델을 훈련하였다. 훈련된 모델을 이용하여 정림사지 오층 석탑을 대상으로 성능을 평가하 였다. 각 유형별 손상 탐지 재현율은 IoU 0.50 기준 0.62부터 0.86의 범위에 있었으며, 손상 영역 분할 재현율은 0.51부터 0.68의 범위로 나타났다. 이러한 결과는 인공지능을 문화유산 현장에 적용하여 새로운 안전관리 방법론을 제시하며 문화유산 보존에 있어 중요한 응용 가능성을 가진다.

Damage diagnosis through monitoring is essential for systematic and efficient conservation and management of cultural heritage. In this study, we developed a deep learning system that automatically detects and visualizes damage to stone pagodas to enable regular monitoring of cultural heritage. Mask R-CNN was used to detect and visualize damage in pixel units in stone pagoda images. A dataset specialized for stone pagodas in Korea was built and applied to train the model. The generalized performance of the trained model was evaluated on the five-story stone pagoda at Jeongnimsa Temple Site. The damage detection recall for each type was in the range of 0.86 to 0.62 based on IoU 0.50, and the damage area segmentation recall was in the range of 0.68 to 0.51. This study suggests a new safety management methodology by applying artificial intelligence to cultural heritage sites and has important applications in cultural heritage preservation.

일반적으로 산성은 사찰, 가마터, 주거지와 달리 유구가 드러나 있어 상대적으로 조사는 쉽다. 다만 규모가 있고 산림 속에 있어 지상 조사만으로는 전체적인 형태를 확인하기는 어렵다. 해외에서는 LiDAR를 활용하여 밀림 속의 정착지나 도로와 같은 유적을 확인한 사례는 많이있다. 그러나 국내에서 지금까지 관련 사례는 거의 없는 실정이다. 항공 고고학 분석 방법이고지형분석이란 이름으로 도입해서 충적지 및 훼손 성벽을 추정 · 복원한 사례가 있으나 LiDAR를활용한 원격 탐사 연구는 현재 거의 없는 실정이다. 이에 본 연구에서는 예산군 소재 예산산성을 대상으로 LiDAR를 활용한 시각화를 시도하였다. 객관적인 고고학적 정보를 제공하기 위해 이미지 시각화 및 GIS 기술을 활용하여 LiDAR 데이터를가공하였다. 결과를 근거로 예산산성의 전체적인 경관을 조망하고 기존 발굴 보고서와의 비교를통해 LiDAR 활용방안에 대한 연구방법을 제시하였다. LiDAR를 이용한 연구는 접근도가 떨어지는 원격지 혹은 삼림 등에 분포하는 유적의 형태를파악할 수 있게 한다. 특히 다양한 종류의 자연적, 인공적 장애물에도 불구하고 지표면의 건축물을발견하고 특정 대상을 관찰할 수 있는 LiDAR의 속성은 관련 연구자에게 기술적 객관성을 제공해줄 수 있다. 이는 국토 대부분이 산지로 이루어진 한국의 지형적 특성에 효과적으로 활용될 수 있을 것이다.

In general, mountain fortresses are relatively easy to investigate, unlike temples, kilns, and dwellings, as the remains are exposed. However, it is difficult to confirm the overall shape only by ground survey because it is large and is located in the forest. Overseas, there are many examples of using LiDAR to identify ruins such as settlements or roads in the jungle. However, there are few related cases in Korea so far. Although there are cases in which aerial archeology analysis method was introduced under the name of paleomorphological analysis and estimated and restored alluvial sites and damaged walls, there are currently few remote sensing studies using LiDAR. In this paper, we attempted to restore the historical landscape using LiDAR for Yesan Mountain Fortress located in Yesan-gun. LiDAR data was processed using image visualization and GIS technology to provide objective archaeological information. Based on the results, the overall landscape of Yesan Mountain Fortress was viewed and a research method for LiDAR utilization was presented through comparison with existing excavation reports. Research using LiDAR makes it possible to grasp the shape of ruins distributed in remote areas or forests with poor access. In particular, the property of LiDAR, which can detect buildings on the ground surface and observe specific objects despite various types of natural and artificial obstacles, can provide technical objectivity to related researchers. This can be effectively utilized for the topographical characteristics of Korea, where most of the land is mountainous.

The purpose of this study was to estimate the forest biomass using satellite imagery and machine learning techniques. In this study, Random Forest, XGBoost, SVM, Multiple Linear Regression were used for forest biomass estimation. Research Forest management plan(8th) data and Sentinel-2 imagery information were used to analysis. As the dependent variables, forest biomass was calculated using volume information, and the biomass expansion factor. The 10 bands of Sentinel-2 were used as independent variable. The optimal forest biomass estimation model was selected by comparing the calculated value based on the Research Forest management plan data and the estimate based on the machine learning techniques. MAE, RMSE, and R2 were calculated for comparison of estimated biomass statistics. As a result, the XGBoost model showed the highest RMSE(61.63ton/ha), MAE(44.16ton/ha), and the highest R2(0.48) value, and was evaluated as the optimal biomass estimation model. The average amount of biomass for sub compartments estimated using the XGBoost model was 225.3tons/ha, which was underestimated by 3.4 tons/ha compared to the average amount of biomass calculated using the Research Forest management plan.

The precision level of forest inventory assessment is a key factor in the forest industry to optimize timber value and forest management planning. Light Detection and Ranging (LiDAR) application in forestry has recently introduced an effective tool to estimate and monitor the precision level of forest inventory. This research investigates the opportunities and limitations of integrated MLS and ALS LiDAR application in estimating individual tree DBH and tree height. The three circular fitting algorithms(integrated RANSAC and circle fitting , minimum enclosing circle, least squares Ellipse fitting) were used to estimate the DBH. In addition, the height was calculated by the Crown Height Model (CHM). The result of DBH estimation was the most accurate in Ellipse fitting algorithm. To compare previous circular fitting applications, integrated MLS and ALS had better performance estimating tree DBH. Lastly, the estimated tree height using CHM was compared with the ground truth, and R2, RMSE of height are 0.60, 2.0 m. Also, the trends of less accuracy in small trees were identified in PCDs extraction at the detect of tree top from the overlapped PCDs among the surrounded tall trees PCDs. This result indicated that small trees are limited to segment top of tree PCDs due to overlaps by tall trees in high dense forest structures. Also, taller trees have limited to capturing ground point due to the high dense of forest crown.

증발산은 토양 표면에서 일어나는 증발 과정과 식물의 광합성으로 인해 발생하는 증산 작용을 포함한 수문 기상인자로 수문 순환과정에서 중요한 역할을 차지한다. 현재 국내외에서는 증발산을 산정하고 공간적인 거동을 파악하기 위한 연구가 활발히 진행되고 있지만 특정 지역에서의 토지 피복의 차이나 식생으로 인해 거동을 이해하는데 많은 제약이 따른다. 본 연구에서는 고해상도의 영상을 제공하는 Landsat 위성이 기반이 되는 원격탐사 기반 에너지 수지 모형인 Mapping EvapoTRanspiration with Internalized Calibration (METRIC) 모형과 MODerate resolution Imaging Spectroradiometer (MODIS) 위성 기반의 Penman-Monteith 알고리즘으로 산정된 증발산의 공간 분포를 비교하였다. 토지 피복별로 분류한 후 두 공간 분포를 비교하여 침엽수림과 활엽수림에서 가장 높은 상관관계를 갖는 것을 확인하였고 두 모형에 대한 적용성이 높음을 알 수 있다. 본 연구를 바탕으로 원격탐사 기반 고해상도 증발산 지도를 제작하여 시공간적 변동성과 계절 변화에 따른 거동을 파악할 수 있을 것이다.

Evapotranspiration (ET) including evaporation from a land surface and transpiration from photosynthesis of vegetation is a hydrological factor that has an important role in water cycle. However, there is a limitation to understand it due to heterogeneity of land cover and vegetation. In this study, Mapping EvapoTRanspiration with Internalized Calibration (METRIC) model, one of the energy balance models, and MODerate resolution Imaging Spectroradiometer (MODIS) satellite based well-known Penman-Monteith algorithm were compared. Two ET maps were categorized and compared by land cover classification. The results represented overall applicability of the two models with the highest correlation coefficients in needleleaf and broadleaf forests. This study will be useful to estimate remote sensing based ET maps with high resolution and to figure out spatio-temporal variability and seasonal changes.

Dramatic price increases of fossil fuels and the economic development of emerging nations accelerates the transformation of forest lands into monocultures, e.g. for biofuel production. On this account, cost efficient methods to enable the monitoring of land resources has become a vital ambition. The application of remote sensing techniques has become an integral part of forest attribute estimation and mapping. The aim of this study was to evaluate the potentials of the kNN method by combining terrestrial with remotely sensed data for the development of a pixel-based monitoring system for the small scaled mosaic of different land use types of the off-reserve forests of the Goaso forest district in Ghana, West Africa. For this reason, occurrence and distribution of land use types like cocoa and non-timber forest resources, such as bamboo and raphia palms, were estimated, applying the kNN method to ASTER satellite data. Averaged overall accuracies, ranging from 79% for plantain, to 83% for oil palms, were found for single-attribute classifications, whereas a multi-attribute approach showed overall accuracies of up to 70%. Values of k between 3 and 6 seem appropriate for mapping bamboo. Optimisation of spectral bands improves results considerably.

Remote sensing image fusion plays a crucial role in enhancing image information. However, the limitations of existing fusion technologies in terms of computational resources and storage capacity make real-time processing difficult. Therefore, a lightweight fusion method based on knowledge distillation is proposed for vehicle remote sensing image fusion. The knowledge distillation technology is used to transfer the complex teacher model knowledge to the lightweight student model, which realizes the significant reduction of model complexity while maintaining high fusion accuracy. Experimental results show that the proposed method performs well on DroneVehicle dataset and the model weight is only 0.641M. 2025 The Korean Institute of Communications and Information Sciences. Publishing Services by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

We analyze a remote sensing system in the Internet of things, where uncoordinated nodes send status updates to a common receiver to achieve information freshness, quantified through age of information. We consider a finite horizon scheduling over a random multiple access channel, where colliding messages are lost. We show that nodes must adopt a further randomization to deviate from identical schedules and escape collision deadlocks. Moreover, we discuss the impact of feedback availability if, due to, e.g., energy expenditure, it decreases the number of transmission opportunities.

Hyperspectral unmixing (HU) is crucial for extracting material information from hyperspectral images (HSI) obtained through remote sensing. Although linear unmixing methods are widely used due to their simplicity, they only address linear mixing effects. Nonlinear mixing models, while more complex, often focus solely on the nonlinear aspects affecting individual pixels. However, in practice, light reflected from materials within a pixel experiences linear and nonlinear interactions, necessitating a hybrid mixing model (HMM) that leverages spatial and spectral information. This work proposes a novel deep learning-based autoencoder (AE) with dual-stream decoders to enhance spectral unmixing. Our approach employs multitask learning (MTL) to process spatial and spectral information concurrently. Specifically, one decoder stream performs linear unmixing from HSI patches, while the other stream utilizes fully connected layers to capture and model the nonlinear interactions within the data. By integrating linear and nonlinear information, our method improves the accuracy of unmixing the mixed spectrum. We validate the effectiveness of our architecture on three real-world HSI datasets and compare its performance against various baseline methods. Experimental results consistently demonstrate that our approach outperforms existing methods, as evidenced by superior spectral angle distance (SAD) and mean squared error (MSE) metrics.

Mechanized timber harvesting operations often cause soil disturbance, such as compaction and rutting. The extent of soil disturbance is highly dependent on environmental factors and operation methods. As soil disturbance has a long-term impact on forest productivity, it is critical to assess its effects. The objective of this study was to investigate soil surface deformation caused by forest machinery traffic in a steep slope clear-cut area and to compare manual methods with remote sensing. Following the timber harvesting operation design, we established the experimental treatments representing the number and direction of forest machinery passes (1D, 1-downward; 1U, 1-upward; 3R, 3-round-trip; 5R, 5-round-trip). Soil rut depth and cross-section were manually measured using pinboard and estimated by mobile LiDAR system (MLS) and unmanned aerial vehicle structure from motion algorithm (UAV SfM). There was a significant difference in soil rut depth based on the number of passes (p = 0.00), while no significant difference based on the direction of passes. Rut depth in 1D (22.2cm) was significantly higher than 3R (15.7cm), with no significant differences among 1D, 1U (20.0cm), and 5R (19.6cm) or among 3R, 1U, and 5R. These findings suggest that most soil disturbance occurs during the initial passes of forest machinery. The comparison of pinboard and MLS data revealed a significant relationship (R2 = 0.74, slope = 1.00, p = 0.00). Comparing pinboard and UAV SfM data, we found that MLS is more accurate than UAV SfM in assessing soil surface deformation (R2 = 0.60, slope = 0.81, p = 0.00). The results reveal the need to establish optimized driving routes for forest machinery to minimize soil disturbance and suggest the potential of using MLS and UAV SfM for future soil disturbance assessments.

통신해양기상위성은 다목적 정지궤도위성으로서 Ka대역 통신탑재체, 기상센서 및 해양센서를 하나의 위성플랫폼에 탑재한 복합위성이다. 본 논문에서는 한국정부의 자금으로 개발되는 첫번째 혁신적인 정지궤도 통신해양기상위성 프로그램에 대해서 소개하고자 한다. 위성플랫폼은 아스트리움의 UROSTAR 3000 통신위성을 기반으로 하고 있으며, 세 개의 다른 탑재체를 효과적으로 수용하기 위하여 화성탐사선 Express를 일부 활용하였다. 세개의 탑재체 중 통신탑재체는 스위칭 다중빔 기술을 검증하고 광대역 멀티미디어 통신서비스를 시험하는데 목적이 있다. 기상센서임무는 고해상도 멀티분광 센서로 지속적으로 한반도 기상데이타를 산출하는데 있으며, 세계 최초의 정지궤도 해양센서는 한반도의 어류자원정보및 장단기 해양정보의 모니터링을 목적으로 하고 있다. 통신임무와 원격탐사임무를 동시에 수행해야 하므로 위성체의 요구사항은 매우 복잡하여 이를 만족시키기 위한 설계 및 조립/시험의 난이도는 매 우 높다고 할 수 있겠다.

COMS satellite is a multipurpose satellite in the geostationary orbit, which accommodates multiple payloads of the Ka band Satellite Communication Payload, Meteorological Imager, and Geostationary Ocean Color Imager into a single spacecraft platform. In this paper, Korea’s first innovative geostationary Communication, Ocean and Meteorological Satellite (COMS) program is introduced which is fully funded by Korean Government. The satellite platform is based on the Astrium EUROSTAR 3000 communication satellite, but creatively combined with MARS Express satellite platform to accommodate three different payloads efficiently for COMS. The goals of the Ka band satellite communication mission are to in-orbit verify the performances of advanced communication technologies and to experiment wide-band multi-media communication service. The Meteorological Imager mission is to continuously extract meteorological products with high resolution and multi-spectral imager, to detect special weather such as storm, flood, yellow sand, and to extract data on long-term change of sea surface temperature and cloud. The Geostationary Ocean Color Imager mission aims at monitoring of marine environments around Korean peninsula, production of fishery information (Chlorophyll, etc.), and monitoring of long-term/short-term change of marine ecosystem. The system design difficulties are in the different kinds of payload mission requirements of communication and remote sensing purposes and how to combine them into one to meet the overall satellite requirements. In this paper, Ka band communication payload system is more highlighted.

History of remote sensing studies of the Moon and asteroids changed when lunar samples were returned by the Apollo 11 mission and many meteorites were discovered on Antarctica starting in 1969. Discovery of the isotopic similarity between lunar and terrestrial materials led us to the giantimpact model to form the Moon. In addition, the existence and nature of space weathering were also discovered in 1993 by analyzing the Apollo samples. Another change occurred in 2010 when the Hayabusa spacecraft returned particles of asteroid Itokawa that proved the identity between many S-type asteroids and ordinary chondrites and the existence of space weathering similar to the Moon. The second sample return from asteroids occurred in 2020 when the Hayabusa2 spacecraft returned samples of Ctype asteroid Ryugu. In spite of some expectations, it was a pristine CI1 chondrite material that was free from terrestrial contaminations suffered by known CI1 chondrite meteorites. Sample return missions drastically improved the accuracy of our knowledge on the raw materials of solar system planets and will surely keep revealing the secrets behind the birth of this special planet Earth. This part of history also teaches us that scientists should proclaim the truth against denial or persecution by others.

지난 10년간 우리나라에서 발생한 자연재해의 88%는 호우와 태풍에 의해 발생했으며 정량적 피해 를 산출하기 어려운 가뭄피해 또한 광역적으로 발생하고 있다. 본 연구에서는 한반도를 포함한 동아 시아 지역에 홍수, 가뭄 등의 수재해 정보를 모니터링하고 예측하기 위한 수재해 정보 플랫폼과 홍수/가뭄 모니터링 시스템을 개발하였다. 수재해 정보 플랫폼은 인공위성, 레이더, 지상관측 자료를 활용하여 홍수와 가뭄 등의 수재해를 모니터링 할 수 있는 시스템을 갖추고 있다. 본 플랫폼에는 피해규모가 상대적으로 큰 도시지역을 대상으로 고해상도 레이더 기반 홍수 모니터링 시스템이 운 영되고 있으며 위성기반의 임진강 접경지역의 홍수 모니터링 시스템과 한반도를 포함한 광역적 가 뭄 감시 및 예측 시스템이 운영되고 있다. 수재해 정보 플랫폼의 홍수/가뭄 모니터링 시스템을 통해 국민의 생명과 재산을 보호하고 홍수, 가뭄 등 수재해의 선제적 대응이 가능한 해결책이 되길 기대 한다.

88% of the natural disasters in South Korea over the past ten years were caused by heavy rains and typhoons. In addition, drought damage, which is challenging to estimate quantitative damage, occurs in a wide area. In this study, a water hazard information platform was developed to monitor and predict water hazard information such as drought and flood in a wide area of East Asia, including the Korean Peninsula. The water hazard information platform is equipped with a system that can monitor water disasters such as drought and floods by using satellite, radar, automatic weather station data. This platform has an X-band radar system for predicting and monitoring urban floods with a significant damage scale. The Imjin River flood prediction system in the border area is operating for flood monitoring. The drought monitoring and forecasting system is operating using satellite-based in a wide area. Through the water hazard information platform, it is expected to protect the lives and property of the people and as a critical solution for a preemptive response to water hazards such as floods and droughts.

High-resolution land cover maps are essential in fields such as forest resource management, urban green space planning, and environmental protection. In recent years, Unmanned Aerial Vehicles (UAVs) have increasingly become influential in land cover mapping due to their flexibility, low cost, and fast data acquisition capability. However, accurately classifying high-resolution image data collected by UAVs remains a challenge due to the complexity of the data and the substantial computational resources required for processing. To address this problem, this study combines UAV remote sensing data with Object-Based Image Analysis (OBIA) to optimize feature selection to improve the accuracy of land cover classification and provide more reliable data support. In this study, combinations of four feature types were evaluated using a Decision Tree (DT) algorithm in eight scenarios. The results showed that a comparison with spectral features alone and the combination of other feature types can significantly improve the classification accuracy. Height features contribute the most to enhancing the classification results, followed by spectral and geometric features, while the contribution of texture features is relatively limited. In addition, the optimal feature combination selected by the Recursive Feature Elimination (RFE) method further validates its effectiveness in improving land cover classification results. Finally, the best feature combination achieved a classification accuracy of 72.00% and a Kappa coefficient of 0.6543, proving the effectiveness of the feature selection and optimization strategy.

원격광학측정장비(Solar Occultation Flux ; SOF)는 태양 적외선(IR) 스펙트럼을 기반으로 한다. 해당 장비는 오염물질 의 풍하 및 횡풍을 따라 수직 통합된 농도를 측정하여, 총 배출량을 제공한다. 플럭스는 풍상측의 유입을 제외한 오염물 질의 추정치로 계산된다. SOF 시스템 외, SO2 측정을 위해 UV/visible 시스템(SkyDOAS)를 사용했다. SkyDOAS는 천 정에서 산란된 빛을 측정하며, SOF와 다르게 직달광을 사용하지 않는다. 본 연구에서는 국내 일관제철소 중 한 개 사업 장을 대상으로 20년 12월부터 21년 09월까지 총 6회 황산화물(SO2) 배출량을 측정하고 사업장에서 제시한 배출량 자료 와 비교하였다. 사업장의 평균 SO2 배출량은 322 kg/hr 였으며, 장비 측정결과는 평균 638 kg/hr로 약 2배 높은 배출량을 나타냈다. 보다 정확한 측정을 위해서는 풍향·풍속의 정확도 향상을 위해 라이다를 이용하여 고도별 풍향·풍속 자료를 이 용할 예정이며, 사업장 내부 측정을 통해 외부에서의 배출량 간섭을 최소화하여 사업장에서 배출되는 오염물질만을 측정 하여 비교한다면 보다 정확한 사업장의 효과적인 관리가 가능할 것으로 판단된다.

Solar Occultation Flux (SOF) is based on the Solar Infrared (IR) spectrum. The equipment measures vertically integrated concentrations along the wind and crosswinds of pollutants and provides total emissions. Flux is calculated as an estimate of contaminants excluding inflow from the windward side. In addition to SOF systems, UV/visible systems (SkyDOAS) are used for SO2 measurements. SkyDOAS measures scattered light in the zenith and does not utilize direct solar light in contrast to the SOF instrument. In this study, the total amount of sulfur oxide (SO2) emissions from the integrated steel mill located in Chungcheongnam-do was measured six times from December 20 to September 21 and compared with the emission data presented by the workplace. The average emission at a workplace was 322 kg/hr, and the equipment measurement results showed that the emission was twice as high as 638 kg/hr on average. For more accurate measurement, LIDAR will be used to improve the accuracy of wind direction and wind speed, and if only pollutants emitted from the workplace are measured and compared by minimizing emission interference from the outside, it will be possible to effectively manage the workplace.