• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.246-246
• /
• 2021

본 연구는 강우 발생시 유량을 추정하는 것에 목적이 있다. 이를 위해 본 연구는 선행연구의 모형 개발방법론에서 벗어나 딥러닝 알고리즘 중 하나인 합성곱 신경망 (convolution neural network)과 수문학적 이미지 (hydrological image)를 이용하여 강우 발생시 유량을 추정하였다. 합성곱 신경망은 일반적으로 분류 문제 (classification)을 해결하기 위한 목적으로 개발되었기 때문에 불특정 연속변수인 유량을 모의하기에는 적합하지 않다. 이를 위해 본 연구에서는 합성곱 신경망의 완전 연결층 (Fully connected layer)를 개선하여 연속변수를 모의할 수 있도록 개선하였다. 대부분 합성곱 신경망은 RGB (red, green, blue) 사진 (photograph)을 이용하여 해당 사진이 나타내는 것을 예측하는 목적으로 사용하지만, 본 연구의 경우 일반 RGB 사진을 이용하여 유출량을 예측하는 것은 경험적 모형의 전제(독립변수와 종속변수의 관계)를 무너뜨리는 결과를 초래할 수 있다. 이를 위해 본 연구에서는 임의의 유역에 대해 2차원 공간에서 무차원의 수문학적 속성을 갖는 grid의 집합으로 정의되는 수문학적 이미지는 입력자료로 활용했다. 합성곱 신경망의 구조는 Convolution Layer와 Pulling Layer가 5회 반복하는 구조로 설정하고, 이후 Flatten Layer, 2개의 Dense Layer, 1개의 Batch Normalization Layer를 배열하고, 다시 1개의 Dense Layer가 이어지는 구조로 설계하였다. 마지막 Dense Layer의 활성화 함수는 분류모형에 이용되는 softmax 또는 sigmoid 함수를 대신하여 회귀모형에서 자주 사용되는 Linear 함수로 설정하였다. 이와 함께 각 층의 활성화 함수는 정규화 선형함수 (ReLu)를 이용하였으며, 모형의 학습 평가 및 검정을 판단하기 위해 MSE 및 MAE를 사용했다. 또한, 모형평가는 NSE와 RMSE를 이용하였다. 그 결과, 모형의 학습 평가에 대한 MSE는 11.629.8 m3/s에서 118.6 m3/s로, MAE는 25.4 m3/s에서 4.7 m3/s로 감소하였으며, 모형의 검정에 대한 MSE는 1,997.9 m3/s에서 527.9 m3/s로, MAE는 21.5 m3/s에서 9.4 m3/s로 감소한 것으로 나타났다. 또한, 모형평가를 위한 NSE는 0.7, RMSE는 27.0 m3/s로 나타나, 본 연구의 모형은 양호(moderate)한 것으로 판단하였다. 이에, 본 연구를 통해 제시된 방법론에 기반을 두어 CNN 모형 구조의 확장과 수문학적 이미지의 개선 또는 새로운 이미지 개발 등을 추진할 경우 모형의 예측 성능이 향상될 수 있는 여지가 있으며, 원격탐사 분야나, 위성 영상을 이용한 전 지구적 또는 광역 단위의 실시간 유량 모의 분야 등으로의 응용이 가능할 것으로 기대된다.

• Journal of Wetlands Research
• /
• v.6 no.3
• /
• pp.71-81
• /
• 2004

In recent, the heavy rainfall is frequently occurred and the damage tends to be increased. So, more careful hydrologic analysis is required for the designs of the hydraulic or disaster prevention structures. The time distribution of a rainfall is one of the important factors for the estimation of peak flow in hydrologic and hydraulic designs. This study is to suggest a methodology for the estimation of a rainfall time distribution which can reflect the meteorologic and topographical characteristics of Daejeon area. We collect the 34 years' rainfall data recorded in the range of 1969 to 2002 for Daejeon area and we performed the rainfall analysis with the data in between May and October of each year. According to the Huff method, the collected data corresponds to the first quartile which the rainfall is concentrated in the primary stage but the suggested method shows the different rainfall distribution with the Huff method in time. The reason is that the Huff method determines the quartile in each storm event while the suggested one determines it by estimating the dimensionless distribution of rainfall in duration after the accumulation of rainfall in time. The rainfall distributions estimated by two methodologies were applied to the Gabcheon basin in Daejeon area for the estimation of flood flow. Here we use the SCS method for the effective rainfall and unit hydrograph for the flood discharge. As the results, the peak flow for 24-hour of 100-year frequency was estimated as a $3421.20m^3/sec$ by the Huff method and $3493.38m^3/sec$ by the suggested one. We can see the difference of $72.18m^3/sec$ in between two methods and thus we may carefully determine the rainfall time distribution and compute the effective rainfall for the estimation of the peak flow.

• Journal of Korea Water Resources Association
• /
• v.48 no.12
• /
• pp.1065-1075
• /
• 2015

In the estuary where the structure such as river-mouth weir has been installed, the flow is developed very complicatedly due to river water from upstream, tide of the sea and floodgate operation. Especially, if basin outlets more than one exists in one estuary, the boundary conditions will be significantly more complex form. Saemangeum(SMG) project area in Korea is the most typical example. There are Mankyung river and Dongjin river in upstream. The water of them inflows into SMG project area. In the downstream, river flow was drained from inland to sea over the SMG sea dike through the sluice. The connecting channel was located between Mankyung and Dongjin basins. It functions not only as transportation by ship in ordinary period but also as flood sharing by sending flood flow to each other in flood period. Therefore, in order to secure the safety against flood, it is very important to understand the flood sharing capacity for connecting channel. In this study, the flood control effect was analyzed using numerical simulation. Delft3D was used to numerical simulation and simulated period was set up with neap tide, in which the maximum flood stage occurred due to poor drainage. Actually, three connecting channels were designed in land use plan of the SMG Master Plan, but they were simplified to a single channel for conciseness of analysis in this study. According to the results of numerical analysis, the water level difference between two basins was increased and the maximum flood stage at dike sluice was also upraised depending on decrease of conveyance. And the velocity induced by same water level difference was decreased when the conveyance became smaller. In certain conveyance above, there was almost no flood control effect. Therefore, if the results of this study are considered for design of connecting channel, it will be expected to draw the optimal conveyance for minimizing dredging construction cost while maximizing the flood control effect.

• Journal of Korea Water Resources Association
• /
• v.54 no.spc1
• /
• pp.1095-1105
• /
• 2021

The advancement of dam operation is further required due to the upcoming rainy season, typhoons, or torrential rains. Besides, physical models based on specific rules may sometimes have limitations in controlling the release discharge of dam due to inherent uncertainty and complex factors. This study aims to forecast the water level of the nearest station to the dam multi-timestep-ahead and evaluate the availability when it makes a decision for a release discharge of dam based on LSTM (Long Short-Term Memory) of deep learning. The LSTM model was trained and tested on eight data sets with a 1-hour temporal resolution, including primary data used in the dam operation and downstream water level station data about 13 years (2009~2021). The trained model forecasted the water level time series divided by the six lead times: 1, 3, 6, 9, 12, 18-hours, and compared and analyzed with the observed data. As a result, the prediction results of the 1-hour ahead exhibited the best performance for all cases with an average accuracy of MAE of 0.01m, RMSE of 0.015 m, and NSE of 0.99, respectively. In addition, as the lead time increases, the predictive performance of the model tends to decrease slightly. The model may similarly estimate and reliably predicts the temporal pattern of the observed water level. Thus, it is judged that the LSTM model could produce predictive data by extracting the characteristics of complex hydrological non-linear data and can be used to determine the amount of release discharge from the dam when simulating the operation of the dam.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.39 no.4
• /
• pp.83-97
• /
• 2011

In this study, space allocation with an assessment system for improving the function of the ecological use and function of each space was undertaken to suggest appropriate goals and directions. Ogeum Park, used as the study site, is a stronghold green zone located in the fan-shaped area of the Songpa-gu green zone on a 22ha area site. The assessment system for this study consisted of a total of 3 steps: Step 1 includes the division of the spatial block, Step 2 assesses the proper functioning of each spatial block, while Step 3 includes space allocation and presents improvements for function by space. This study performed a basin analysis with the consideration that Ogeum Park is a forest area and divided the site into a total of 8 areas according to how the sections of land are used. The function of wildlife habitat included an analysis of plant ecology(vegetation type, vegetation layer, potential vegetation), animal ecology(wild birds), and waterways. The function of leisure and use included an analysis of gradients, noise, paths, status of use, and status of facilities. The evaluation of the function of habitat sorted items into native vegetation, vegetation diversification, vegetation potential, animal diversification and potential of animal habitats. The results of grading the evaluation scores by space in the function of habitat showed that the Areas IV and VII, which were 90% of the maximum point, were Grade A, Areas II and V were Grade B at 70% and Area I and VI were Grade C at 50%. Areas III and VIII, which were not found to be beyond the standard of 50%, were excluded. The evaluation of the functions of leisure and use classified items into use potential, use preference, use concentration, use diversification and use convenience. The results of the graded evaluation scores by space in use function, showed areas V and VI as Grade A, or 90% of the maximum score. Grade B, 70% of the maximum score, was given to Areas I and VII. Grade C, 50% of the maximum score, included Areas II, IV and VIII. Area III, graded lower the standard of 50%, was excluded. The study evaluated areas according to a common standard, classified spaces by proper functions into ecological spaces, environmentally-friendly use spaces and use spaces according to the standard of spatial distribution on the basis of the above results through a synthesis of grades of habitat function, leisure function and use. This offers ideas for the improvement of wildlife habitat and environment-friendly use functions by space.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.43 no.6
• /
• pp.16-24
• /
• 2015

This study quantified storage and annual uptake of carbon for riparian greenspaces established in watersheds of four major rivers in South Korea and explored desirable strategies to improve carbon reduction effects of riparian greenspaces. Greenspace structure and planting technique in the 40 study sites sampled were represented by single-layered planting of small trees in low density, with stem diameter at breast height of $6.9{\pm}0.2cm$ and planting density of $10.4{\pm}0.8trees/100m^2$ on average. Storage and annual uptake of carbon per unit area by planted trees averaged $8.2{\pm}0.5t/ha$ and $1.7{\pm}0.1t/ha/yr$, respectively, increasing as planting density got higher. Mean organic matter and carbon storage in soils were $1.4{\pm}0.1%$ and $26.4{\pm}1.5t/ha$, respectively. Planted trees and soils per ha stored the amount of carbon emitted from gasoline consumption of about 61 kL, and the trees per ha annually offset carbon emissions from gasoline use of about 3 kL. These carbon reduction effects are associated with tree growth over five years to fewer than 10 years after planting, and predicted to become much greater as the planted trees grow. This study simulated changes in annual carbon uptake by tree growth over future 30 years for typical planting models selected as different from the planting technique in the study sites. The simulation revealed that cumulative annual carbon uptake for a multilayered and grouped ecological planting model with both larger tree size and higher planting density was approximately 1.9 times greater 10 years after planting and 1.5 times greater 30 years after than that in the study sites. Strategies to improve carbon reduction effects of riparian greenspaces suggest multilayered and grouped planting mixed with relatively large trees, middle/high density planting of native species mixed with fast-growing trees, and securing the soil environment favorable for normal growth of planting tree species. The research findings are expected to be useful as practical guidelines to improve the role of a carbon uptake source, in addition to water quality conservation and wildlife inhabitation, in implementing riparian greenspace projects under the beginning stage.

• Journal of Korea Water Resources Association
• /
• v.54 no.12
• /
• pp.1339-1348
• /
• 2021

In accordance with the River Law, the basic river maintenance plan is established every 5-10 years with a considerable national budget for domestic rivers, and various river surveys such as the river section required for HEC-RAS simulation for flood level calculation are being conducted. However, river survey data are provided only in the form of a pdf report to the River Management Geographic Information System (RIMGIS), and the original data are distributedly owned by designers who performed the river maintenance plan in CAD format. It is a situation that the usability for other purposes is considerably lowered. In addition, when using surveyed CAD-type cross-sectional data for HEC-RAS, tools such as 'Dream' are used, but the reality is that time and cost are almost as close as manual work. In this study, RAUT (River Information Auto Upload Tool), a tool that can solve these problems, was developed. First, the RAUT tool attempted to automate the complicated steps of manually inputting CAD survey data and simulating the input data of the HEC-RAS one-dimensional model used in establishing the basic river plan in practice. Second, it is possible to directly read CAD survey data, which is river spatial information, and automatically upload it to the river spatial information DB based on the standard data model (ArcRiver), enabling the management of river survey data in the river maintenance plan at the national level. In other words, if RIMGIS uses a tool such as RAUT, it will be able to systematically manage national river survey data such as river section. The developed RAUT reads the river spatial information CAD data of the river maintenance master plan targeting the Jeju-do agar basin, builds it into a mySQL-based spatial DB, and automatically generates topographic data for HEC-RAS one-dimensional simulation from the built DB. A pilot process was implemented.

• Journal of the Korean Society for Environmental Technology
• /
• v.19 no.6
• /
• pp.503-513
• /
• 2018

In order to continuous watershed management and improve water quality at Yeong-san river system, we analyzed and evaluated data on the monitoring of small stream in city and county boundaries within the watershed. In-period monitoring is estimated to be more frequent in the second quarter than the first quarter, so it should be considered when evaluating the target water quality by setting the target water quality. A small stream in the Yeong-san river system has higher concentration in the downstream area than the upstream area. As a result of calculating the load of the measuring point, Y.b B3(Pungyeongjeongcheon) and Y.b E1(Sampocheon) were high. The result of correlation analysis by monitoring point in order to evaluate the correlation between BOD and T-P items, BOD was highly correlated with COD and TOC, and was affected by emission of pollutants related to organic matter. T-P was highly correlated with SS and COD, and was affected by rainfall. This study will provide basic data and direction for designing efficient and scientific method for water quality management by analyzing accumulated water quality data by conducting long-term monitoring.

• Economic and Environmental Geology
• /
• v.57 no.2
• /
• pp.243-251
• /
• 2024

In this study, a surface geological survey was first conducted to investigate aggregate resources in the Yeongam area of Jeollanam-do, and a drilling survey was conducted in the lower part of the surface, which was difficult to identify through a surface geological survey, to determine the spatial distribution of aggregates. Drilling sites were selected considering the topographical development and Quaternary alluvium characteristics of the study area, and river aggregate drilling surveys were conducted at a total of 5 points and land aggregate drilling surveys were conducted at a total of 28 points. Borehole core sediments were classified into seven sedimentary units to determine whether they could be used as aggregates, and optically stimulated luminescence dating was performed on representative boreholes to measure the depositional period for each sedimentary unit. As a result of the study, most of the Yeongam area had a very wide river basin, so it was estimated that there would be a large amount of aggregate, but the amount of aggregate was evaluated to be very small compared to other cities and counties. Most of the unconsolidated sedimentary layers in the Yeongam area are composed of blue-grey marine clay with a vertical thickness of more than 10 m. The sand-gravel layer corresponding to the aggregate section is distributed in the lower part of the marine clay, thinly covering the bedrock weathering zone. This is because the amount of aggregates themselves is small and most of the aggregates are distributed at a depth of 10 m below the surface, which is currently difficult to develop, so the possibility of developing aggregates is evaluated to be very low. As a result of dating, it can be seen that the blue-grey marine clay layer is an intertidal sedimentary layer formed as the sea level rose rapidly about 10,000 years ago. The deposition process continued from 10,000 years ago to the present, and as a result, a very thick clay layer was deposited. This clay layer was formed very dominantly for about 6,000 to 8,000 years, and the sand-gravel layer in the section where aggregates deposited in the Pleistocene period can exist was measured to have been deposited at about 13.0 to 19.0 ka, and about 50 ka, showing that it was deposited as paleo-fluvial deposits before the marine transgression process.

• Korean Journal of Ecology and Environment
• /
• v.35 no.1 s.97
• /
• pp.52-61
• /
• 2002

This research was conducted to evaluate the effect of forest management practices on pH and electrical conductivity to get fundamental information on water purification capacity after forest operation. Rainfall, throughfall and stemflow　were　sampled at the study sites which consist of Abies holophylla and Pinus koraiensis　in Gwangreung Experimental Forest for S months from May to November 1999. Mean pH of the throughfall of the beginning of the event was higher in management (thinning and pruning) sites of Abies holophylla and Pinus koraiensis stands than nonmanagement site of Abies holophylla and Pinus koraiensis stands. In addition, pH of the throughfall of the total amount of the event showed similar trends which are higher pH in the management sites compared with the non- management sites. This result indicates that managements such as thinning and pruning improve tree butler capacity of rainfall pH. According to the linear regression results, pH of the throughfall of the total amount of the event in non-management sites = 0.735${\times}$pH of the throughfall of the beginning of the event in non-management sites+1.849 ($R^2\;=\;0.82$) and pH of the throughfall of the total amount of the event in management sites= 0.863${\times}$pH of the throughfall of the beginning of the event in management sites +1.0242 ($R^2\;=\;0.87$). In case of stemflow pH, pH of the sternflow of the total amount of　the event in non-management sites = 0.53${\times}$pH of the stemflow of the beginning of　the event in non- management sites+2.7709 ($R^2\;=\;0.64$) and pH of the stemflow of the total amount of the event in management sites = 0.5854${\times}$pH of the stemflow of the beginning of the event in management sites+2.7045 ($R^2\;=\;0.65$). Electrical conductivity (EC) of the throughfall of the beginning and total amount of the event was highest in non- management site in Abies holophylla, followed by management sites in fsies Abies holophylla, non-management site in Pinus koraiensis, and management sites in Pinus koraiensis stands, respectively. According to the linear regression results, EC of the throughfall of the total amount of the event in non-managementsites = 0.4045${\times}$EC of the throughfall of the beginning of the event in non-management sites+26.766 ($R^2\;=\;0.69$) and EC of the throughfall of the total amount of the event in management sites = 0.6002${\times}$EC of the throughfall of the beginning of the event in management sites+8.0184 ($R^2\;=\;0.54$). In case of stemflow EC, EC of thestemflow of the total amount of the event in non-management sites = 0.6298${\times}$EC of the stemflow of the beginning of the event in non-management sites+11.582 ($R^2\;=\;0.72$) and pH of the stemflow of the total amount of the event in management sites =0.602${\times}$pH of the stemflow of the beginning of the event in management sites+20.783($R^2\;=\;0.49$).