한국농공학회논문집 · 한국농공학회

Development of a Gate Release Guideline During Flood Season Using Proximal Policy Optimization Algorithm Based Agricultural Reservoir Operation Model
Proximal Policy Optimization 알고리즘 기반 농업용 저수지 운영 모의를 통한 홍수기 수문 방류 운영 기법 개발

Jeonghoon Ryu, Jihye Kwak, Jihye Kim, Sinae Kim, Seongjae Park, Moon Seong Kang
류정훈, 곽지혜, 김지혜, 김시내, 박성재, 강문성

The operation of agricultural reservoirs during flood seasons requires timely pre-release strategies to comply with restricted water levels and mitigate downstream risks. … + READ MORE

The operation of agricultural reservoirs during flood seasons requires timely pre-release strategies to comply with restricted water levels and mitigate downstream risks. Effective decision-making depends on accurate inflow, storage, and release data; however, such data are often limited. Reinforcement Learning (RL) offers an advantage in such uncertain environments by learning optimal policies without predefined answers. In this study, we developed a reservoir operation simulation technique using the Proximal Policy Optimization (PPO) algorithm and trained the model. The trained agent demonstrated stable convergence, with reward values concentrated in high-performance ranges and minimal policy fluctuations in later episodes. Simulations applying a maximum allowable release of 2,475 m³/s-based on observed historical operations-resulted in a 1.2% (623 m³) lower total release and a 10% (247.5 m³/s) lower peak release compared to the actual case. Final water levels remained within flood-season restricted water level. In a more conservative scenario reflecting Yedang Reservoir’s guideline (1,645 m³/s max release), the RL model achieved a 2.7% (1,439 m³) lower total release and a 33.5% (830 m³/s) reduction in peak release compared to the real operation. Although peak release duration increased from 7 to 15 hours, minimum and maximum water levels matched the observed case, and the final level (21.63 EL.m) was close to the target flood limit (21.5 EL.m). These results suggest that flood response could have been successfully managed even with stricter release constraints. - COLLAPSE

PDF XML

References Information

2026
Prediction of Reservoir Inflow Using the FPHM Linked with Reservoir Water Level Observations
FPHM과 저수지 수위 관측 자료를 연계한 저수지 유입량 예측

Seokhun Kang, Jung-Hun Song
강석훈, 송정헌

Effective management of agricultural reservoirs relies on accurate knowledge of inflow and outflow. However, direct measurements are often unavailable for small and … + READ MORE

Effective management of agricultural reservoirs relies on accurate knowledge of inflow and outflow. However, direct measurements are often unavailable for small and medium reservoirs operated by the Korea Rural Community Corporation (KRC). Water-level observations are more readily collected and can be used to estimate storage changes. By combining water-level observations with expert understanding of reservoir operations, indirect methods have been developed to reconstruct continuous inflow and outflow records. One such approach integrates hydrological models with water-level–based estimates. The Five-Parameter Hyperbolic Model (FPHM) is a recently developed lumped hydrological model with a parsimonious structure, but its application to agricultural reservoirs remains limited. This study evaluates the effectiveness of a framework that links FPHM with water-level–based inflow estimation. The framework was applied to four reservoirs, including Idong, Gopung, Gosam, and Cheongcheon, and the results were compared with those from a modified three-tank model. The results indicate that the FPHM reproduces reservoir inflow dynamics with comparable accuracy to the modified three-tank model, while showing reduced parameter-induced uncertainty under data-limited conditions. Through this comparison, we assess the applicability of FPHM for reconstructing reservoir inflow and supporting reservoir operation under data-scarce conditions. - COLLAPSE

PDF XML

References Information

2026
Loading Rate Based Pollution Characteristics of Fine Particulate Matter and Gaseous Precursors in Agricultural Areas
부하속도 기반 농업지역 초미세먼지 및 가스상전구물질 오염 특성

Jeong-Deok Baek, Sung-hyun Bae, Yun-sik Shin, Min-wook Kim, Jin-ho Kim, Il-hwan Seo, Ki-yeon Kim, Hung-soo Joo
백정덕, 배성현, 신윤식, 김민욱, 김진호, 서일환, 김기연, 주흥수

Fine particulate matter (PM_2.5) is a critical air pollutant affecting human health, ecosystems, and agricultural productivity. In agricultural areas, PM … + READ MORE

Fine particulate matter (PM_2.5) is a critical air pollutant affecting human health, ecosystems, and agricultural productivity. In agricultural areas, PM_2.5 is emitted directly through farming activities and indirectly formed via gas–particle conversion involving sulfur dioxide (SO₂), nitrogen dioxide (NO₂), and ammonia (NH₃). However, concentration-based assessments are strongly influenced by meteorological dilution in open agricultural environments, limiting their ability to represent pollutant transport and exposure. To address this limitation, this study applied a loading-rate approach integrating pollutant concentrations with wind-driven transport. Long-term monitoring was conducted at eight agricultural sites across South Korea from January 2024 to June 2025, with PM_2.5, gaseous precursors, and meteorological parameters measured at 5-minute intervals. Loading rates were calculated using pollutant concentrations and air exchange rates, and February and August 2024 were selected for comparative analysis. Statistical analyses included principal component analysis, Pearson correlation, and multiple linear regression. Loading rates exhibited greater variability than concentrations and distinct diurnal patterns, with most pollutants peaking 1–2 hours after meteorological maxima. PM_2.5 loading rates consistently clustered with SO₂ and NO₂, suggesting close statistical associations with secondary inorganic aerosol formation. Regression analysis showed high explanatory power for PM_2.5 loading rates (R=0.97, p<0.001), identifying SO₂ as the largest positive coefficient, while NO₂ showed season-dependent significance and NH₃ exhibited earlier peaks and time-lagged behavior. Overall, loading-rate metrics more effectively capture pollutant transport and exposure characteristics in agricultural environments and highlight the possible importance of sulfur-driven secondary formation pathways in PM_2.5 management. - COLLAPSE

PDF XML

References Information

2026
Establishing Future Research Directions Based on the Assessment of Agricultural Infrastructure Technology Levels and Technology Gap Survey Results
농공기술 수준진단 및 기술격차 조사 결과에 따른 후속 연구 방향 정립

Seungheon Lee, Jongwon Do, Sunghee Lee
이승헌, 도종원, 이성희

This study assesses agricultural infrastructure technology levels and gaps relative to leading countries through an expert Delphi survey and proposes future R&D … + READ MORE

This study assesses agricultural infrastructure technology levels and gaps relative to leading countries through an expert Delphi survey and proposes future R&D directions. A two-round Delphi survey (Round 1: 153; Round 2: 137) evaluated 14 technologies across four fields: diversified farming, water use, water safety, and water environment. As of 2025, South Korea's overall technology level was estimated at 80.9% (U.S.=100%), with an average technology gap of 3.9 years. The overall level and gap were aggregated across the 14 technologies using the median of expert ratings (with consistent treatment of missing responses), and the 2030 values reflect the experts’ forward-looking estimates in Round 2. Korea shows strengths in conventional civil engineering domains (e.g., reclaimed farmland development) but lags in emerging data- and AI-driven domains (e.g., monitoring, analytics, and smart management). The findings indicate an urgent transition toward data- and AI-centric infrastructure, supported by expanded R&D investment, increased industry participation, and strengthened evaluation systems. - COLLAPSE

PDF XML

References Information

2026
Predicting Fish Occurrence Probability in the Nakdong River Basin Using Species Distribution Models
종분포모델을 활용한 낙동강 유역의 어류 서식 확률 예측

Soyoung Woo, Wonjin Kim, Yongwon Kim, Yonggwan Lee, Seongjoon Kim
우소영, 김원진, 김용원, 이용관, 김성준

In this study, Species Distribution Models (SDMs) were developed to predict the occurrence probability of six representative fish species in the Nakdong … + READ MORE

In this study, Species Distribution Models (SDMs) were developed to predict the occurrence probability of six representative fish species in the Nakdong River Basin. Species occurrence data were obtained from biomonitoring survey, and environmental variables were constructed using hydrological and water quality data generated by the Soil and Water Assessment Tool (SWAT), considering not only basin characteristics but also various physical and chemical riverine conditions such as flow rate and water quality. The occurrence probability of each fish species was predicted by SDM based on Random Forest algorithm. The SDMs showed high predictive accuracy with values exceeding 0.899, and differences in occurrence probability of each species were observed depending on stream order and habitat preference. Furthermore, response curve showed the influence of environmental variable on species occurrence. These results represent an integrated analysis of factors affecting fish distribution across multiple spatial scales, from the watershed to individual stream reaches. This study provides a scientific foundation for conservation and restoration strategies in the Nakdong River Basin by applying SDMs to freshwater fish and identifying key environmental determinants of habitat suitability. - COLLAPSE

PDF XML

References Information

2026
Characterizing Agricultural Drought Onset and Lag Using Gridded Climate and Hydrological Data
기후⋅수문 자료 기반 농업가뭄 발생 및 시차 특성 분석

Young-Sik Mun, Won-Ho Nam, Seokkyun Yu
문영식, 남원호, 유석균

PDF XML

References Information

2026
Analysis of Temperature-Salinity-Based Stratification Characteristics of Estuarine Reservoir Using HSPF-EFDC Coupled Model and PEA Index
HSPF-EFDC 연계모형과 PEA 지표를 활용한 하구 담수호의 수온⋅염도 기반 성층화 특성 분석

Jaeyoung An, Sinae Kim, Jihye Kwak, Seongjae Park, Suhwan Park, Seoju Oh, Seonjae Park, Jihye Kim, Moon-Seong Kang
안재영, 김시내, 곽지혜, 박성재, 박수환, 오서주, 박선재, 김지혜, 강문성

This study aimed to estimate watershed inflows to an estuarine reservoir using the Hydrological Simulation Program-Fortran (HSPF), simulate water level, temperature, and … + READ MORE

This study aimed to estimate watershed inflows to an estuarine reservoir using the Hydrological Simulation Program-Fortran (HSPF), simulate water level, temperature, and salinity dynamics using the Environmental Fluid Dynamics Code (EFDC), and quantify the spatiotemporal characteristics of stratification using the Potential Energy Anomaly (PEA) index. Previous studies on estuarine reservoirs have mainly focused on salinity or temperature distributions, while integrated quantification of stratification based on an energy perspective remains limited. To address this gap, a coupled watershed–reservoir modeling framework was developed and applied to the Ganwol estuarine reservoir. PEA was computed from EFDC outputs and summarized as volume-weighted means to reflect spatial variability. The results showed a seasonal pattern, with weak stratification in winter, rapid development in spring, peak conditions in summer, and decay in autumn. The mean PEA over the simulation period was 9.347 J/m³, indicating generally weak stratification, whereas strong stratification occurred mainly between May and September, with the highest frequency in June to August. Interannual variability in stratification intensity was also observed, suggesting the influence of thermal and salinity structures. Correlation analysis showed that PEA was strongly associated with surface–bottom temperature and salinity differences, while inflow and operational variables had relatively weak relationships. These findings indicate that stratification is primarily controlled by density differences associated with temperature and salinity rather than direct hydraulic effects. Periods and regions of strong stratification are likely to experience suppressed vertical mixing and increased risk of bottom water quality deterioration, suggesting the need for management strategies such as artificial mixing. - COLLAPSE

PDF XML

References Information

2026
Analysis of Water Quality Changes and Driving Factors Before and After Weir Construction in the Yeongsan River Basin: Focusing on Seungchon and Juksan Weirs
영산강 유역 하천보 건설 전⋅후 수질 변화와 영향 요인 분석: 승촌보⋅죽산보를 중심으로

Yongjin Park, Yonggwan Lee, Jinuk Kim, Yongwon Kim, Wonjun Jeon, Seongjoon Kim
박용진, 이용관, 김진욱, 김용원, 전원준, 김성준

This study analyzed long-term water quality variations before (2000~2008) and after (2016~2024) weir construction using data from monitoring stations representative of the … + READ MORE

This study analyzed long-term water quality variations before (2000~2008) and after (2016~2024) weir construction using data from monitoring stations representative of the basin. Statistical techniques, including the Mann-Kendall test, Welch’s T-test, and Load Duration Curve (LDC) analysis, were applied to eight water quality parameters: water temperature (WT), dissolved oxygen (DO), biochemical oxygen demand (BOD), chemical oxygen demand (COD), suspended solids (SS), total nitrogen (T-N), total phosphorus (T-P), and chlorophyll-a (Chl-a). Following weir construction, significant improvements were observed in nutrient and biodegradable organic concentrations due to enhanced watershed management: BOD, SS, T-N, and T-P decreased by 26.5% (6.6 → 4.8 mg/L), 17.8% (19.1 → 15.6 mg/L), 38.0% (8.2 → 5.2 mg/L), and 74.6% (0.6 → 0.1 mg/L), respectively. Conversely, indicators related to water stagnation and internal production deteriorated: COD and Chl-a increased by 29.0% (7.0 → 9.0 mg/L) and 42.1% (31.2 → 44.2 mg/m³), respectively, accompanied by increases in WT (4.0%) and DO (9.7%). These findings reveal a structural shift in pollution mechanisms where physical factors, such as increased residence time, dominate over chemical nutrient limitations (P-limitation), leading to a paradox of increased algal blooms despite nutrient reduction. - COLLAPSE

PDF XML

References Information

2026