All Issue

2025 Vol.58, Issue 12S-3 Preview Page

Research Article

Enhancing virtual sensors for riverine Total Nitrogen (T-N) and Total Phosphorus (T-P) using land-use characteristics

토지 이용 특성을 고려한 하천의 T-N, T-P 가상 센서 개선 방안 연구

Yumin Kanga
,
Su Han Namb
,
Sangung Leec
,
Tae Young Kimd
,
Jong Hyun Leee
,
Min Ju Kimf
,
Young Do Kimg*
강 유민a
,
남 수한b
,
이 상웅c
,
김 태영d
,
이 종현e
,
김 민주f
,
김 영도g*
aPh.D. Student, Department of Civil and Environmental Engineering, Myongji University, Yongin, Korea
bPh.D. Candidate, Department of Civil and Environmental Engineering, Myongji University, Yongin, Korea
cPh.D. Student, Department of Civil and Environmental Engineering, Myongji University, Yongin, Korea
dUndergraduate Student, Department of Civil and Environmental Engineering, Division of Smart Infrastructure Engineering, Myongji University, Yongin, Korea
eUndergraduate Student, Department of Civil and Environmental Engineering, Division of Smart Infrastructure Engineering, Myongji University, Yongin, Korea
fUndergraduate Student, Department of Civil and Environmental Engineering, Division of Smart Infrastructure Engineering, Myongji University, Yongin, Korea
gProfessor, Department of Civil and Environmental Engineering, Division of Smart Infrastructure Engineering, Myongji University, Yongin, Korea
a명지대학교 토목환경공학과 박사과정
b명지대학교 토목환경공학과 박사과정
c명지대학교 토목환경공학과 박사과정
d명지대학교 스마트시스템공과대학 스마트인프라공학부 건설환경공학전공 학사과정
e명지대학교 스마트시스템공과대학 스마트인프라공학부 건설환경공학전공 학사과정
f명지대학교 스마트시스템공과대학 스마트인프라공학부 건설환경공학전공 학사과정
g명지대학교 스마트시스템공과대학 스마트인프라공학부 건설환경공학전공 교수
*Corresponding Author
31 December 2025. pp. 1535-1549
PDF XML Citation
Abstract

Recent studies have highlighted that river water quality is strongly influenced by the spatial heterogeneity of land-use characteristics such as urbanization, agricultural intensity, and forest structure. However, most virtual-sensor-based water-quality prediction approaches still rely on uniform modeling frameworks that do not fully account for these land-use-driven differences in pollutant behavior. To address this gap, this study developed cluster-based virtual sensor models for estimating Total Nitrogen (T-N) and Total Phosphorus (T-P) concentrations across 37 watershed-outlet monitoring sites in the Han River Basin. We utilized water-quality variables including WT, DO, pH, EC, Chl-a, SS, NO3-N, NH3-N, and PO4-P and constructed prediction models using Multiple Linear Regression (MLR), Random Forest (RF), and Extreme Gradient Boosting (XGB). Spatial clustering was integrated to enhance model precision and interpretability by capturing land-use-based differences among sub-watersheds. Model evaluation showed that T-N predictions exhibited relatively consistent performance across algorithms, whereas T-P predictions demonstrated substantial variability among clusters and models, highlighting the strong nonlinearity and spatial heterogeneity associated with phosphorus dynamics. SHAP analysis revealed that NO3-N and NH3-N were the dominant predictors for T-N, while SS and PO4-P played leading roles in T-P estimation. Cluster-specific differences in predictor sensitivity and directional influence were also observed, particularly for SS and PO4-P, which exhibited sign reversals depending on land-use context and model type. These findings underscore the importance of incorporating land-use characteristics into machine-learning-based water-quality prediction and demonstrate the potential of cluster-informed virtual sensors to support more targeted and adaptive water-quality management strategies.

Keywords
T-N (Total Nitrogen)
T-P (Total Phosphorus)
Land Cover/Land Use (LCLU)
Clustering
Eutrophication
Virtual sensor

최근 하천 수질은 기후 변화와 더불어 도시화·농업화·산림 구조 등 토지이용 특성의 공간적 이질성에 따라 크게 달라지며, 동일한 수질 구성 인자라도 유역의 토지피복 배경에 따라 서로 다른 반응과 민감도를 보인다. 그럼에도 기존의 가상 센서 기반 수질 예측 연구는 이러한 토지이용도 기반 공간적 차이를 충분히 반영하지 못한 한계가 존재한다. 본 연구는 한강권역 37개 단위유역 말단 지점을 대상으로, WT, DO, pH, EC, Chl-a, SS, NO3-N, NH3-N, PO4-P 등의 수질 자료를 활용해 가상 센서 기반 T-N 및 T-P 농도 산정 모델을 구축하였다. 다중선형회귀(MLR), 랜덤 포레스트(RF), 그래디언트 부스팅(XGB)을 적용하고, 공간적 clustering 기법을 도입해 모델의 정밀도와 해석 가능성을 향상시켰다. 평가 결과, T-N 산정에서는 모델 간 성능 차이가 크지 않았으나, T-P 산정에서는 cluster와 모델 간 성능 편차가 두드러져 P 계열 항목의 지역 이질성과 비선형성이 중요한 영향을 미침을 확인했다. SHAP 분석 결과, T-N은 NO3-N, NH3-N이, T-P는 SS, PO4-P가 주요 기여 변수로 나타났으며, cluster별 민감도와 영향 방향성의 차이도 확인되었다. 특히 SS와 PO4-P는 모델 간 부호 차이를 보였다. 본 연구는 머신러닝 기반 수질 산정의 해석 가능성과 지역 맞춤형 수질 관리 전략의 기초자료 제공 가능성을 제시한다.

키워드
총질소(T-N)
총인(T-P)
토지 피복
군집화
부영양화
가상 센서
References
1

Andrade, P., Silva, I., Silva, M., Flores, T., Cassiano, J., and Costa, D.G. (2022). “A TinyML soft-sensor approach for low-cost detection and monitoring of vehicular emissions.” Sensors, Vol. 22, No. 10, pp. 1-18. doi: 10.3390/s22103838.

10.3390/s2210383835632247PMC9143421
2

Azedou, A., Amine, A., Kisekka, I., and Lahssini, S. (2025). “Genetic algorithm optimization of ensemble learning approach for improved land cover and land use mapping: Application to Talassemtane National Park.” Ecological Indicators, Vol. 177, 113776. doi: 10.1016/j.ecolind.2025.113776.

10.1016/j.ecolind.2025.113776
3

Bai, Y., Wong, C.P., Jiang, B., Hughes, A.C., Wang, M., and Wang, Q. (2018). “Developing China’s ecological redline policy using ecosystem services assessments for land use planning.” Nature Communications, Vol. 9, No. 1, pp. 1-13. doi: 10.1038/s41467-018-05306-1.

10.1038/s41467-018-05306-130072771PMC6072749
4

Belmiro, C., Silveira Neto, R. M., Barros, A., and Ospina, R. (2023). “Understanding the land use intensity of residential buildings in Brazil: An ensemble machine learning approach.” Habitat International, Vol. 139, 102896. doi: 10.1016/j.habitatint.2023.102896.

10.1016/j.habitatint.2023.102896
5

Breiman, L. (2001). “Random forests.” Machine Learning, Vol. 45, No. 1, pp. 5-32.

10.1023/A:1010933404324
6

Chaves, A.J., Martín, C., Llopis Torres, L., Díaz, M., Fernández-Ortega, J., Barberá, J.A., and Andreo, B. (2025). “A soft sensor open-source methodology for inexpensive monitoring of water quality: A case study of NO3-concentrations.” Journal of Computational Science, Vol. 85, 102522. doi: 10.1016/j.jocs.2024.102522.

10.1016/j.jocs.2024.102522
7

Chen, Y., Yang, J., Yu, W., Ren, J., Xiao, X., and Xia, J. C. (2023). “Relationship between urban spatial form and seasonal land surface temperature under different grid scales.” Sustainable Cities and Society, Vol. 89, 104374. doi: 10.1016/j.scs.2022.104374.

10.1016/j.scs.2022.104374
8

Chong, B. (2021). “K-means clustering algorithm: a brief review.” Academic Journal of Computing & Information Science, Vol. 4, No. 5, pp. 37-40. doi: 10.25236/AJCIS.2021.040506.

10.25236/AJCIS.2021.040506
9

Fontanini, A.D., and Abreu, J. (2018). “A data-driven BIRCH clustering method for extracting typical load profiles for big data.” IEEE Power and Energy Society General Meeting, Portland, OR, U.S. doi: 10.1109/PESGM.2018.8586542.

10.1109/PESGM.2018.8586542
10

Han River Basin Environmental Office (HRBEO) (2025). Full-scale implementation of 2025 Paldang Lake algae countermeasures for securing safe drinking water for residents, Press Release, accessed May 3, 2025, <https://www.mcee.go.kr/hg/web/board/read.do?pagerOffset=70&maxPageItems=10&maxIndexPages=10&searchKey=&searchValue=&menuId=7001&orgCd=&boardId=1737590&boardMasterId=142&boardCategoryId=&decorator=>.

11

Harrison, J.W., Lucius, M.A., Farrell, J.L., Eichler, L.W., and Relyea, R.A. (2021). “Prediction of stream nitrogen and phosphorus concentrations from high-frequency sensors using random forests regression.” Science of the Total Environment, Vol. 763, 143005. doi: 10.1016/j.scitotenv.2020.143005.

10.1016/j.scitotenv.2020.143005
12

Hossain, M. (2025). “A simple and effective approach to investigate the dominant contaminant sources and accuracy in water quality estimation through Monte Carlo simulation, Gaussian Mixture Models (GMMs), and GIS machine learning methods.” Ecological Indicators, Vol. 171, 113188. doi: 10.1016/j.ecolind.2025.113188.

10.1016/j.ecolind.2025.113188
13

Ikotun, A.M., Ezugwu, A.E., Abualigah, L., Abuhaija, B., and Heming, J. (2023). “K-means clustering algorithms: A comprehensive review, variants analysis, and advances in the era of big data.” Information Sciences, Vol. 622, pp. 178-210. doi: 10.1016/j.ins.2022.11.139.

10.1016/j.ins.2022.11.139
14

Jain, A.K., Murty, M.N., and Flynn, P.J. (1999). “Data clustering: A review.” ACM Computing Surveys, Vol. 31, No. 3, pp. 264-323. doi: 10.1145/331499.331504.

10.1145/331499.331504
15

Kaufman, L., and Rousseeuw, P.J. (1990). Finding groups in data: An Introduction to cluster analysis. John Wiley & Sons, New York, NY, U.S.

10.1002/9780470316801
16

Kim, C., Do, T.N., and Kim, J. (2024). “Spatially explicit supply chain for nationwide CO2-to-fuel infrastructure: Data-driven optimization with Gaussian mixture model -based region screening and clustering.” Journal of Cleaner Production, Vol. 471, 143390. doi: 10.1016/j.jclepro.2024.143390.

10.1016/j.jclepro.2024.143390
17

Kim, H.K., Kim, K.H., Yun, S.T., Oh, J., Kim, H.R., Park, S.H., Kim, M.S., and Kim, T.S. (2019). “Probabilistic assessment of potential leachate leakage from livestock mortality burial pits: A supervised classification approach using a Gaussian mixture model (GMM) fitted to a groundwater quality monitoring dataset.” Process Safety and Environmental Protection, Vol. 129, pp. 326-338. doi: 10.1016/j.psep.2019.07.015.

10.1016/j.psep.2019.07.015
18

Kim, S.H., Lee, D.H., Kim, M.S., Rhee, H.P., Hur, J., and Shin, K.H. (2023). “Systematic tracing of nitrate sources in a complex river catchment: An integrated approach using stable isotopes and hydrological models.” Water Research, Vol. 235, 119755. doi: 10.1016/j.watres.2023.119755.

10.1016/j.watres.2023.119755
19

Korea Environment Corporation (K-eco) (2022). Annual report on water pollution incidents and responses (2022).

20

Li, D., Yang, C., and Li, Y. (2024). “A multi-subsystem collaborative Bi-LSTM-based adaptive soft sensor for global prediction of ammonia-nitrogen concentration in wastewater treatment processes.” Water Research, Vol. 254, 121347. doi: 10.1016/j.watres.2024.121347.

10.1016/j.watres.2024.121347
21

Li, K., Zhao, J., Li, Y., and Lin, Y. (2025). “Identifying trade-offs and synergies among land use functions using an XGBoost-SHAP model: A case study of Kunming, China.” Ecological Indicators, Vol. 172, 113330. doi: 10.1016/j.ecolind.2025.113330.

10.1016/j.ecolind.2025.113330
22

Li, W., Li, H., and Luo, Y. (2023). “Dynamic and static enhanced BIRCH for functional data clustering.” IEEE Access, Vol. 11, pp. 111448-111465. doi: 10.1109/ACCESS.2023.3322929.

10.1109/ACCESS.2023.3322929
23

Lin, X., Wu, X., Chao, J., Ge, X., Tan, L., Liu, W., Sun, Z., and Hou, J. (2024). “Effects of combined ecological restoration measures on water quality and underwater light environment of Qingshan Lake, an urban eutrophic lake in China.” Ecological Indicators, Vol. 163, 112107. doi: 10.1016/j.ecolind.2024.112107.

10.1016/j.ecolind.2024.112107
24

Lundberg, S.M., and Lee, S.-I. (2017). “A unified approach to interpreting model predictions.” In Proceedings of the 31st Conference on Neural Information Processing Systems (NeurIPS 2017), Long Beach, CA, U.S., pp. 4765-4774.

25

Ma, Y., Wang, J., Xiong, J., Sun, M., and Wang, J. (2024). “Risk assessment for cropland abandonment in mountainous area based on AHP and PCA—Take Yunnan Province in China as an example.” Ecological Indicators, Vol. 158, 111287. doi: 10.1016/j.ecolind.2023.111287.

10.1016/j.ecolind.2023.111287
26

Macnaughton-Smith, P., Williams, W.T., Dale, M.B., Mockett, L.G. (1964). “Dissimilarity analysis: A new technique of hierarchical sub-division.” Nature, Vol. 202, No. 4936, pp. 1034-1035.

10.1038/2021034a0
27

Martin, D., Kühl, N., and Satzger, G. (2021). “Virtual Sensors.” Business and Information Systems Engineering, Vol. 63, No. 3, pp. 315-323. doi: 10.1007/s12599-021-00689-w.

10.1007/s12599-021-00689-w
28

Murphy, K., Heery, B., Sullivan, T., Zhang, D., Paludetti, L., Lau, K. T., Diamond, D., Costa, E., O’Connor, N., and Regan, F. (2015). “A low-cost autonomous optical sensor for water quality monitoring.” Talanta, Vol. 132, pp. 520-527. doi: 10.1016/j.talanta.2014.09.045.

10.1016/j.talanta.2014.09.045
29

Nam, S.H., Kwon, S., and Kim, Y.D. (2024). “Development of a basin-scale total nitrogen prediction model by integrating clustering and regression methods.” Science of the Total Environment, Vol. 920, 170765. doi: 10.1016/j.scitotenv.2024.170765.

10.1016/j.scitotenv.2024.170765
30

Randriamihamison, N., Vialaneix, N., and Neuvial, P. (2021). “Applicability and interpretability of ward’s hierarchical agglomerative clustering with or without contiguity constraints.” Journal of Classification, Vol. 38, No. 2, pp. 363-389. doi: 10.1007/s00357-020-09377-y.

10.1007/s00357-020-09377-y
31

Rashidov, A., Akhatov, A., and Mardonov, D. (2024). “The distribution algorithm of data flows based on the BIRCH clustering in the internal distribution mechanism.” 2024 International Russian Smart Industry Conference (SmartIndustryCon), Sochi, Russia, pp. 923-927. doi: 10.1109/SmartIndustryCon61328.2024.10516193.

10.1109/SmartIndustryCon61328.2024.10516193
32

Singh, P., Haritwal, D.K., Seth, S., Ramana, G.V., and Datta, M. (2024). “Electrical conductivity profiling for rapid contamination assessment in unsaturated zones: A case study of an MSW landfill.” Science of the Total Environment, Vol. 951, 175773. doi: 10.1016/j.scitotenv.2024.175773.

10.1016/j.scitotenv.2024.175773
33

Tokuda, E.K., Comin, C.H., and Costa, L. da F. (2022). “Revisiting agglomerative clustering.” Physica A: Statistical Mechanics and Its Applications, Vol. 585, 126433. doi: 10.1016/j.physa.2021.126433.

10.1016/j.physa.2021.126433
34

Wang, M., Li, Y., Yuan, H., Zhou, S., Wang, Y., Ikram, R.M.A., and Li, J. (2023). “An XGBoost-SHAP approach to quantifying morphological impact on urban flooding susceptibility.” Ecological Indicators, Vol. 156, 111137. doi: 10.1016/j.ecolind.2023.111137.

10.1016/j.ecolind.2023.111137
35

Wang, Z., Sun, F., Guo, Q., Gooddy, D.C., and Wu, F. (2024). “Global scale identification of catchments phosphorus source shifts with urbanization: A phosphate oxygen isotope and Bayesian mixing model approach.” Water Research, Vol. 250, 121026. doi: 10.1016/j.watres.2023.121026.

10.1016/j.watres.2023.121026
36

Xu, Y., Zhang, D., Lin, J., Peng, Q., Lei, X., Jin, T., Wang, J., and Yuan, R. (2024). “Prediction of phytoplankton biomass and identification of key influencing factors using interpretable machine learning models.” Ecological Indicators, Vol. 158, 111320. doi: 10.1016/j.ecolind.2023.111320.

10.1016/j.ecolind.2023.111320
37

Yang, J., Zhang, D., Frangi, A.F., and Yang, J.Y. (2004). “Two-Dimensional PCA: A new approach to appearance-based face representation and recognition.” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 26, No. 1, pp. 131-137. doi: 10.1109/TPAMI.2004.1261097.

10.1109/TPAMI.2004.1261097
38

Yu, W., Liu, F., Jiao, X., Fan, P., Yang, H., Zhang, Y., Li, J., Chen, J., and Li, X. (2025). “Human-induced N-P imbalances will aggravate GHG emissions from lakes and reservoirs under persisting eutrophication.” Water Research, Vol. 276, 123240. doi: 10.1016/j.watres.2025.123240.

10.1016/j.watres.2025.123240
39

Zhang, H., Zhang, L., Wang, S., and Zhang, L.S. (2022). “Online water quality monitoring based on UV-Vis spectrometry and artificial neural networks in a river confluence near Sherfield-on-Loddon.” Environmental Monitoring and Assessment, Vol. 194, 630. doi: 10.1007/s10661-022-10118-4.

10.1007/s10661-022-10118-435920913PMC9349112
40

Zhang, W., Wang, W., Xu, S., Sun, Q., Shi, W., Man, J., Yu, S., Yang, Y., Wu, W., and Hu, X. et al. (2025). “Effectively mitigated eutrophication risk by strong biological carbon pump (BCP) effect in karst reservoirs.” Water Research, Vol. 278, 123395. doi: 10.1016/j.watres.2025.123395.

10.1016/j.watres.2025.123395
41

Zhang, Y., Li, W., Wen, W., Zhuang, F., Yu, T., Zhang, L., and Zhuang, Y. (2024). “Universal high-frequency monitoring methods of river water quality in China based on machine learning.” Science of the Total Environment, Vol. 947, 174641. doi: 10.1016/j.scitotenv.2024.174641.

10.1016/j.scitotenv.2024.174641
42

Zou, H., and Ren, B. (2023). “Analyzing topsoil heavy metal pollution sources and ecological risks around antimony mine waste sites by a joint methodology.” Ecological Indicators, Vol. 154, 110761. doi: 10.1016/j.ecolind.2023.110761.

10.1016/j.ecolind.2023.110761
Information
  • Publisher :KOREA WATER RESOURECES ASSOCIATION
  • Publisher(Ko) :한국수자원학회
  • Journal Title :Journal of Korea Water Resources Association
  • Journal Title(Ko) :한국수자원학회 논문집
  • Volume : 58
  • No :12
  • Pages :1535-1549
  • Received Date : 2025-09-22
  • Revised Date : 2025-11-17
  • Accepted Date : 2025-11-18
  • DOI :https://doi.org/10.3741/JKWRA.2025.58.S-3.1535
Journal Informaiton Journal of Korea Water Resources Association Journal of Korea Water Resources Association
Online Submission
Archives
  • scopus
  • KCI-수자원
  • NRF
  • KOFST
  • KISTI Current Status
  • KISTI Cited-by
  • crosscheck
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close