https://doi.org/10.15681/KSWE.2026.42.4.301

최용환(Yonghwan Choe) ; 정관호(Kwanho Jeong)

Accurate prediction of groundwater levels (GWL) is essential for the sustainable management of coastal aquifers, which are particularly susceptible to saltwater intrusion and climate change. This study developed a comprehensive prediction framework for a coastal aquifer in Busan, South Korea, featuring a complex ria coastline. We utilized six tree-based machine learning models: Decision Tree, Random Forest, Gradient Boosted Trees, XGBoost, LightGBM, and AdaBoost. A hybrid feature selection approach, combining Pearson correlation analysis and recursive feature elimination, effectively identified five optimal predictors from twelve hydrometeorological and water quality variables. After hyperparameter tuning through Bayesian optimization, AdaBoost achieved the best predictive performance (RMSE=0.040, NSE=0.996, R²=0.998). Taylor diagram analysis showed that the optimized ensemble models accurately captured the complex variability of coastal GWL, achieving high correlation (R>0.99) and low error (RMSD<0.06). Additionally, SHAP (SHapley Additive exPlanations) analysis revealed a statistical masking effect, where the strong temporal persistence of lagged GWL (GWL at t-1) obscured the short-term influences of meteorological variables. These results underscore the effectiveness of tree-based ensemble models as real-time monitoring tools for predicting near-future GWL dynamics. To improve proactive disaster management, including drought and inundation forecasting, future research should focus on developing exogenous models that exclude lagged GWL, allowing for a clearer understanding of the independent effects of meteorological variables and extending the forecasting horizons.

https://doi.org/10.15681/KSWE.2026.42.4.315

윤서린(Seorin Yoon) ; 박찬기(Chan-Gi Park) ; 박윤식(Youn Shik Park)

In South Korea, various studies have been conducted on water quality management in relation to soil loss management. The Ministry of Environment has adopted the Universal Soil Loss Equation (USLE) to estimate potential soil loss and has enacted the "Notice on the Current Status of Topsoil Erosion." However, the rainfall erosion capacity factor presented in the notice is based on historical data from 1973 to 1996, which does not reflect current rainfall conditions due to climate change. Similarly, the soil erodibility factor in the notice relies on outdated data, despite current soil data being updated to include 405 soil series. Consequently, the soil loss estimation in the notice was compared to one derived from the conventional USLE approach using recent rainfall and soil data. Additionally, a simple and quick method for estimating soil loss in South Korea, the Korean Soil Loss Equation (KORSLE), was also evaluated. The first finding of the study was that the rainfall erosivity factors in the notice and the conventional approach using recent rainfall data exhibited similar spatial trends, although individual values differed significantly. The second finding indicated that the soil erodibility factor had a greater impact on soil loss estimations than the rainfall erosivity factor. The third finding revealed that KORSLE has uncertainty regarding the spatial distribution characteristics of rainfall. Therefore, it is concluded that the notice needs to be revised to update the soil erodibility factor, and KORSLE should be updated to accurately represent spatial distribution characteristics.

https://doi.org/10.15681/KSWE.2026.42.4.325

강덕준(Dejun Jiang) ; 권혁구(Hyuk-Ku Kwon)

Accurate prediction of chlorophyll-a (Chl-a) in spatially heterogeneous river networks is consistently hindered by data scarcity at newly established monitoring sites. Limited historical records prevent deep learning models from capturing the dynamics preceding acute bloom events. This study proposes a spatial transfer learning framework to address these limitations in the Geum River basin, South Korea. A cost-sensitive Bidirectional GRU (Bi-GRU) source model, which captures bidirectional temporal dependencies in hydrological data, was pre-trained on a decade-long dataset from the Gapcheon River. A sigmoid-weighted loss function was incorporated to prioritize rare high-concentration bloom events. The generalized hydro-chemical representations encoded in the pre-trained Bi-GRU layers were then transferred to two data-scarce target sites, Daecheong (upstream) and Buyeo (downstream), by adjusting the final model output to fit local conditions. Compared to conventional models trained solely on limited local data, the transfer framework demonstrated superior recovery in the critical high-concentration range (top 25%). The locally trained model achieved an overall R² of 0.83 at Daecheong but collapsed entirely under peak bloom conditions (HC R² = -0.01). In contrast, the Transfer model maintained meaningful high-concentration accuracy, achieving HC R² of 0.87 at the downstream Buyeo site. Feature importance analysis revealed a spatial drift in dominant drivers, with predictive reliance shifting from broad thermal controls at upstream sites to localized biogeochemical factors, including pH, TOC, and EC, at the downstream site. This framework offers a scalable early warning pathway for monitoring-limited rivers, presenting a compelling alternative for data-scarce modeling.

https://doi.org/10.15681/KSWE.2026.42.4.338

최용호(Yong-Ho Choi) ; 정인균(In-Kyun Jeong) ; 주진철(Jin-Cheol Joo) ; 정광욱(Kwang-Wook Jeong)

Pungyeongjeong Stream, located in the Yeongsan River basin, demonstrates distinct land-use characteristics: agricultural areas are found upstream, while industrial and urban development dominate downstream. This study aimed to quantify baseflow contributions for effective stream management by applying a watershed model and assessing a model-based baseflow separation technique. We developed a watershed model (HSPF) to reproduce streamflow and performed baseflow separation, comparing the results with conventional empirical baseflow separation methods. The baseflow index (BFI) revealed that baseflow contributions were relatively higher in the upstream region, whereas surface runoff prevailed downstream due to a greater proportion of impervious surfaces. Increased urbanization led to reduced infiltration and groundwater recharge, resulting in diminished baseflow contributions. The BFI also exhibited notable temporal variability, with lower values in summer and higher values in winter, reflecting Korea's monsoon climate characterized by concentrated summer rainfall. When comparing the results with empirical methods, the model-based approach yielded results most similar to those derived using the Eckhardt method while more effectively capturing temporal variability and high-flow conditions. In contrast, empirical digital filter methods often misrepresented high-flow periods, likely due to their inclusion of slowly draining interflow in the baseflow component. Overall, the model-based baseflow separation approach outperformed conventional methods across various time scales. These findings underscore the importance of accounting for land-use diversity and baseflow dynamics in watershed-scale water management and provide a scientific foundation for enhancing stream water quality management amidst increasing urbanization and climate variability.

https://doi.org/10.15681/KSWE.2026.42.4.356

김지연(Ji-Yeon Kim) ; 맹지성(Ji-Sung Mang) ; 김나영(Na-Young Kim) ; 김동섭(Dong-Seop Kim) ; 박병주(Byung-Joo Park) ; 정문철(Moon-Chul Jung) ; 방종윤(Jong-Yun Bang) ; 채규정(Kyu-Jung Chae) ; 이의종(Eui-Jong Lee) ; 이상엽(Sang-Youp Lee) ; 황문현(Moon-Hyun Hwang)

This study presents a comprehensive evaluation comparing reverse osmosis (RO) and thermal vapor recompression (TVR) processes for treating mixed industrial wastewater at a contract treatment facility. The influent wastewater comprised a complex mixture of effluents from various industries, including chemical, steel, plating, automotive parts, electronics, research, and general manufacturing. This mixture exhibited high salinity and organic loading, posing significant challenges for conventional treatment methods. RO consistently achieved high removal efficiencies for major parameters, including TDS (99.9%), BOD (95.4%), TOC (96.3%), TN (96.4%), and TP (98.0%), outperforming TVR in organic and nutrient removal while maintaining comparable salt rejection. A nine-month monitoring campaign demonstrated stable flux performance under the tested operating conditions, with periodic clean-in-place (CIP) ensuring effective recovery. Regression analysis suggested that CIP intervals could theoretically be extended to 4?5 months; however, a conservative 3-month cycle was adopted for operational reliability. While RO permeate generally met discharge standards for nitrogen and phosphorus, additional post-treatment was necessary for BOD and TOC. By integrating biological processes (anoxic-oxic (AO)?membrane bioreactor (MBR)) with optional adsorption as post-treatment, residual organics were further removed, ensuring compliance with effluent regulations. Nonetheless, some trace contaminants and eco-toxicity indicators were less effectively eliminated when RO permeate was co-treated with TVR condensate, highlighting the need for optimized influent blending or advanced pretreatment. Overall, the results demonstrate the practical applicability of the RO-TVR hybrid treatment approach in contract industrial wastewater facilities and provide technical evidence supporting its implementation in the field.

https://doi.org/10.15681/KSWE.2026.42.4.373

박상훈(Sanghun Park) ; 최아름(Ahreum Choi) ; 이상림(Sang Rim Lee) ; 장지이(Jiyi Jang)

Microbial contamination is increasingly recognized as a significant environmental issue that impacts water quality, air quality, ecosystem integrity, and public health. Traditionally, microorganisms have been studied in isolation within aquatic or atmospheric environments. However, recent evidence underscores the water?air interface as a crucial pathway for microbial exchange and the formation of complex pollution. Microorganisms from wastewater, agricultural runoff, and eutrophic waters can become aerosolized through physical processes such as bubble bursting and spray generation. These aerosols can be transported over long distances and redeposited into terrestrial and aquatic systems via wet and dry deposition, creating a bidirectional microbial circulation between water and air. This review synthesizes recent international studies and emerging domestic case studies from Korea on water?air microbial interactions, with a focus on coastal, urban, and marine environments. It summarizes the key mechanisms that control microbial aerosolization, selective transfer, and atmospheric survival. Additionally, we discuss how the exchange of microorganisms between water and air contributes to the spread of pathogens, toxins, and antibiotic resistance genes, linking environmental contamination to ecosystem integrity and potential health risks. By comparing global research trends with domestic studies in Korea, this review identifies critical knowledge gaps, particularly the need for integrated, long-term monitoring of water and air compartments. Finally, we outline future research directions and policy implications for the integrated management of water?air microbial pollution within a One Health framework.

https://doi.org/10.15681/KSWE.2026.42.4.386

이정희(Junghee Lee) ; 김민희(Minhee Kim)

N-(1,3-dimethylbutyl)-N'-phenyl-p-phenylenediamine quinone (6PPD-Q), a transformation product of the widely used tire antiozonant 6PPD, has become a contaminant of significant environmental concern due to its widespread occurrence, enhanced persistence, and pronounced toxicity. This review synthesizes current knowledge on the formation pathways, environmental fate, global occurrence, and toxicological implications of 6PPD-Q, with an emphasis on its relevance to environmental monitoring, risk assessment, and management strategies. 6PPD continuously enters the environment through tire and road wear particles and is transformed into 6PPD-Q via multiple abiotic and biotic pathways, including ozonation, photooxidation, soil-mediated oxidation, microbial degradation, and atmospheric reactions. Compared to its precursor, 6PPD-Q exhibits greater environmental stability and persistence, particularly in aquatic and soil systems, where it can partition into suspended particles, sediments, and soil organic matter. Global monitoring studies have identified 6PPD-Q in road runoff, stormwater, snowmelt, rivers, groundwater, sediments, soils, and influents and effluents of wastewater treatment plants, often showing pulse-like increases associated with rainfall and snowmelt events. Toxicity data indicate that 6PPD-Q poses severe hazards to sensitive aquatic organisms, especially salmonids, even at ng/L concentrations. Emerging evidence also suggests potential adverse effects on plants, mammals, and humans. Collectively, these findings establish 6PPD-Q as a critical tire-derived transformation product that requires integrated multi-media monitoring and refined risk evaluation. Future research should prioritize standardized monitoring frameworks, chronic and mixture toxicity characterization, and life-cycle-based mitigation strategies to support effective regulation and environmental management of 6PPD-Q.