All Issue

2024 Vol.57, Issue 12 Preview Page

Research Article

Estimating urban inundation using physics-informed deep learning: a case study of the Oncheon-cheon catchment

물리정보 기반 딥러닝을 이용한 도시침수 해석: 온천천 유역 사례

Hyuna Wooa
,
Hyeonjin Choib
,
Minyoung Kimc
,
Seong Jin Nohd*
우 현아a
,
최 현진b
,
김 민영c
,
노 성진d*
aMaster’s Course, Department of Civil Engineering, Kumoh National Institute of Technology, Gumi, Korea
bPh.D. Course, Department of Civil Engineering, Kumoh National Institute of Technology, Gumi, Korea
cMaster’s Course, Department of Civil Engineering, Kumoh National Institute of Technology, Gumi, Korea
dAssociate Professor, Department of Civil Engineering, Kumoh National Institute of Technology, Gumi, Korea
a금오공과대학교 토목공학과 석사과정
b금오공과대학교 토목공학과 박사과정
c금오공과대학교 토목공학과 석사과정
d금오공과대학교 토목공학과 부교수
*Corresponding Author
31 December 2024. pp. 989-1001
PDF XML Citation
Abstract

The increasing frequency of urban flooding, driven by extreme rainfall events due to climate change, necessitates the rapid production of flood information for prompt response. However, physically based models are limited in real-time applications due to the high computational cost of generating high-resolution flood information. This study developed and evaluated a Convolutional Neural Network (CNN)-based deep learning model that can rapidly and accurately reproduce the spatial distribution analysis results of a physically based model for urban flooding. For the Oncheon-cheon catchment, flood analysis data at a 10-meter spatial resolution were generated using a physically based model for various synthetic rainfall scenarios. This data was then used to train and construct the CNN model. The model was validated using quantitative metrics, including the Hit Rate (HR), False Alarm Ratio (FAR), and Critical Success Index (CSI), by comparing the maximum inundation depth analysis results of the 2014 and 2020 historical flood events to those of the physically based model. Furthermore, a sensitivity analysis was conducted to assess the impact of temporal variability in rainfall on the performance of the deep learning model. The maximum inundation depth simulation exhibited high reproducibility, with an HR value of 0.98, an FAR of 0.12, and a CSI of 0.85 for the 2014 event, and an HR value of 0.92, an FAR of 0.19, and a CSI of 0.76 for the 2020 event. The analysis of rainfall window size indicated that the CNN model achieved optimal performance with a 6-hour rainfall window for the 2014 event and a 2-hour window for the 2020 event. Additionally, the deep learning-based model reduced computation time by approximately 98% compared to the sequential calculation of the physically based model and by approximately 90% compared to the parallel calculation, enabling near real-time computation. The proposed deep learning-based model is a computationally efficient and accurate alternative to physically based models, making it a useful tool for real-time urban flood prediction and early warning systems.

Keywords
Artificial intelligence
Deep learning
Physical information
Urban inundation
Flood prediction

기후변화로 인한 집중호우의 증가로 도시침수의 발생 빈도가 높아짐에 따라, 침수 위험 지역에서의 신속한 대응을 위해 2차원 침수해석 정보가 필요하다. 하지만, 기존 물리과정 기반 모형은 고해상도 침수 정보 생산에 높은 계산 비용이 요구되어 실시간 활용에 한계가 있다. 본 연구에서는 물리과정 기반 모형의 도시침수 공간분포 해석 결과를 신속하고 정확하게 모사할 수 있는 CNN (Convolutional Neural Network) 기반 딥러닝 모형을 개발하고 그 적용성을 평가하였다. 부산 온천천 유역을 대상으로 물리과정 기반 모형으로 다양한 가상 강수 시나리오에 대해 10 m 공간해상도의 2차원 침수 정보를 생성하고, CNN 모형을 훈련시켰다. 모형의 검증에는 2014년과 2020년의 두 가지 과거 침수 사상에 대한 물리모형 대비 최대 침수심 해석 재현성을 HR (Hit Rate), FAR (False Alarm Ratio), CSI (Critical Success Index) 등의 정량지표로 평가하였다. 또한, 딥러닝 모형에서 강수량의 시간 변동성이 미치는 영향을 평가하기 위해 강수 입력의 시간 범위(강수 윈도우)에 따른 민감도 분석을 수행하였다. 모의 결과, 2014년 침수사상은 HR 0.98, FAR 0.12, CSI 0.85, 2020년 침수사상은 HR 0.92, FAR 0.19, CSI 0.76 로 CNN 모형이 기존 물리모형 대비 침수 공간분포를 적절하게 재현함을 확인하였다. 강수 윈도우 분석 결과, 2014년 사상에서는 6시간, 2020년 사상에서는 2시간 범위를 입력할 때 CNN 모형의 성능이 최적이었다. 또한, 딥러닝 기반 모형은 물리과정 기반 모형의 순차 계산 대비 연산 시간을 약 98%, 병렬 계산 대비 약 90% 단축할 수 있어 준실시간에 가까운 효율적 침수 예측이 가능함을 확인하였다. 제안된 딥러닝 기반 모형은 계산 효율성과 정확도 측면에서 물리과정 기반 모형의 대안으로 활용 가능하며, 도시 침수 예측 및 조기 경보 시스템의 운영을 위한 효율적인 도구가 될 수 있을 것으로 기대된다.

키워드
인공지능
딥러닝
물리정보
도시침수
홍수 예측
References
1

Berkhahn, S., Fuchs, L., and Neuweiler, I. (2019). "An ensemble neural network model for real-time prediction of urban floods." Journal of Hydrology, Elsevier, Vol. 575, pp. 743-754.

10.1016/j.jhydrol.2019.05.066
2

Burrichter, B., Hofmann, J., Koltermann da Silva, J., Niemann, A., and Quirmbach, M. (2023). "A spatiotemporal deep learning approach for urban pluvial flood forecasting with multi-source data." Water, MDPI, Vol. 15, No. 9, 1760.

10.3390/w15091760
3

Busan (2004). Oncheon-cheon river basic management plan report. p. 27.

4

Busan Ilbo (BI) (2020). Busan turned into a 'Sea of Water' Overnight... 3 dead due to flooding and collapses, accessed 12 August 2024, <https://www.busan.com/view/busan/view.php?code=2020072409302417962>.

5

Choi, H., Lee, S., Woo, H., and Noh, S.J. (2023a) "High-resolution urban flood modeling using cellular automata-based WCA2D in the Oncheon-cheon catchment in Busan, South Korea." Journal of Civil and Environmental Engineering Research, KSCE, Vol. 43, No. 5, pp. 587-599.

6

Choi, H., Lee, S., Woo, H., Kim, M., and Noh, S.J. (2023b). "Applying deep learning based super-resolution technique for high- resolution urban flood analysis." Journal of Korea Water Resources Association, KWRA, Vol. 56, No. 10, pp. 641-653.

7

Environment Agency (EA) (2013). What is the updated flood map for surface water? Bristol, UK, p. 26.

8

Guidolin, M., Chen, A.S., Ghimire, B., Keedwell, E.C., Djordjević, S., and Savić, D.A. (2016). "A weighted cellular automata 2D inundation model for rapid flood analysis." Environmental Modelling & Software, Vol. 84, pp. 378-394.

10.1016/j.envsoft.2016.07.008
9

Guo, Z., Leitão, J.P., Simões, N.E., and Moosavi, V. (2021). "Data- driven flood emulation: Speeding up urban flood predictions by deep convolutional neural networks." Journal of Flood Risk Management, Vol. 14, No. 1, e12684.

10.1111/jfr3.12684
10

Gyeonggi Metropolitan News (IMNEWS) (2024). The impact of climate change on the agricultural sector in Gyeonggi Province. accessed 19 June 2024, <http://www.iwnews.co.kr/bbs/board.php?bo_table=news&wr_id=22324>.

11

Hofmann, J., and Schüttrumpf, H. (2021). "floodGAN: Using deep adversarial learning to predict pluvial flooding in real time." Water, MDPI, Vol. 13, No. 16, 2255.

10.3390/w13162255
12

Hosseiny, H. (2021). "A deep learning model for predicting river flood depth and extent." Environmental Modelling & Software, Vol. 145, 105186.

10.1016/j.envsoft.2021.105186
13

Jung, J., Mo, H., Lee, J., Yoo, Y., and Kim, H.S. (2021). "Flood stage forecasting at the Gurye-Gyo station in Sumjin River using LSTM-based deep learning models." Journal of the Korean Society of Hazard Mitigation, KOSHAM, Vol. 21, No. 3, pp. 193-201.

10.9798/KOSHAM.2021.21.3.193
14

Jung, S., Cho, H., Kim, J and Lee, G. (2018). "Prediction of water level in a tidal river using a deep-learning based LSTM model." Journal of Korea Water Resources Association, KWRA, Vol. 51, No. 12, pp. 1207-1216.

15

Kabir, S., Patidar, S., Xia, X., Liang, Q., Neal, J., and Pender, G. (2020). "A deep convolutional neural network model for rapid prediction of fluvial flood inundation." Journal of Hydrology, Elsevier, Vol. 590, 125481.

10.1016/j.jhydrol.2020.125481
16

Kim, H.I., Han, K.Y., and Lee, J.Y. (2020). "Prediction of urban flood extent by LSTM model and logistic regression." Journal of Civil and Environmental Engineering Research, KSCE, Vol. 40, No. 3, pp. 273-283.

17

Kim, H.I., Keum, H.J., and Han, K.Y. (2019). "Real-time urban inundation prediction combining hydraulic and probabilistic methods." Water, MDPI, Vol. 11, No. 2, 293.

10.3390/w11020293
18

Latt, Z.Z., and Wittenberg, H. (2014). "Improving flood forecasting in a developing country: A comparative study of stepwise multiple linear regression and artificial neural network." Water Resource Manage, Vol. 28, pp. 2109-2128.

10.1007/s11269-014-0600-8
19

Lee, S., Choi, H., Woo, H., Kim, M., Lee, E., Kim, S., and Noh, S.J. (2024). "Development and application of cellular automata- based urban inundation and water cycle model CAW." Journal of Korea Water Resources Association, KWRA, Vol. 57, No. 3, pp. 165-179.

20

Lee, S., Nakagawa, H., Kawaike, K., and Zhang, H. (2016). "Urban inundation simulation considering road network and building configurations." Journal of Flood Risk Management, Vol. 9, No. 3, pp. 224-233.

10.1111/jfr3.12165
21

Liao, Y., Wang, Z., Chen, X., Lai, C. (2023). "Fast simulation and prediction of urban pluvial floods using a deep convolutional neural network model." Journal of Hydrology, Elsevier, Vol. 624, 129945.

10.1016/j.jhydrol.2023.129945
22

Löwe, R., Böhm, J., Jensen, D.G., Leandro, J., and Rasmussen, S.H. (2021). "U-FLOOD - Topographic deep learning for predicting urban pluvial flood water depth." Journal of Hydrology, Elsevier, Vol. 603, 126898.

10.1016/j.jhydrol.2021.126898
23

Nearing, G., Cohen, D., Dube, V., Gauch, M., Gilon, O., Harrigan, S., Hassidim, A., Klotz, D., Kratzert, F., Metzger, A., Nevo, S., Pappenberger, F., Prudhomme, C., Shalev, G., Shenzis, S., Tekalign, T.Y., Weitzner, D., Matias, Y. (2024). "Global prediction of extreme floods in ungauged watersheds." Nature, Vol. 627, pp. 559-563.

10.1038/s41586-024-07145-138509278PMC10954541
24

News1 (N1) (2020). Busan hit by 90mm 'Water bomb' per hour: Floods and collapses widespread." accessed 24 July 2020, <https://www.news1.kr/local/busan-gyeongnam/4005486>.

25

Newsis (2023). Busan city establishes comprehensive manual for preventing river accidents. accessed 7 November 2023, <https://www.newsis.com/view/?id=NISX20231107_0002511761&pc_view=1>.

26

Noh, S.J., Lee, J.-H., Lee, S., and Seo, D.-J. (2019). "Retrospective dynamic inundation mapping of hurricane harvey flooding in the Houston Metropolitan Area using high-resolution modeling and high-performance computing." Water, MDPI, Vol. 11, No. 3, 597.

10.3390/w11030597
27

Park, C.E. (2024). "Hourly water level simulation in Tancheon River using an LSTM." Journal of the Korean Society of Agricultural Engineers, KSAE, Vol. 66, No. 4, pp. 51-57.

28

Samantaray, S., Sahoo, A., and Agnihotri, A. (2023). "Prediction of flood discharge using Hybrid PSO-SVM algorithm in Barak River Basin." MethodsX, Vol. 10, 102060.

10.1016/j.mex.2023.10206036865648PMC9972406
29

Schmid, F., and Leandro, J. (2023). "A feature-informed data-driven approach for predicting maximum flood inundation extends." Geosciences, MDPI, Vol. 13, No. 12, 384.

10.3390/geosciences13120384
30

Seleem, O., Ayzel, G., Bronstert, A., and Heistermann, M. (2023). "Transferability of data-driven models to predict urban pluvial flood water depth in Berlin, Germany." Natural Hazards and Earth System Sciences, Vol. 23, No. 2, pp. 809-822.

10.5194/nhess-23-809-2023
31

Seoul Metropolitan Government (SMG) (2023). Seoul implements the nation's first flood warning system 'Companion partners' support evacuation for the vulnerable. accessed 10 March 2023, < https://news.seoul.go.kr/env/archives/522983>.

32

Son, T.S., Kang,D.H., Jang,J.K and Shin, H.S. (2010). "A study of assessment for internal inundation vulnerability in urban area using SWMM." Journal of The Korean Society of Hazard Mitigation, KOSHAM, Vol. 10, No. 4, pp. 105-117.

33

The Kukmin Daily (TKD) (2014). Busan woke up to a 'Mud field'... Is it a 'Special disaster area'?, accessed 12 August 2024, <https://www.kmib.co.kr/article/view.asp?arcid=0008622104>.

34

Tsakiri, K., Marsellos, A., and Kapetanakis, S. (2018). "Artificial neural network and multiple linear regression for flood prediction in Mohawk River, New York." Water, MDPI, Vol.10, 1158.

10.3390/w10091158
35

Wang, Y., Chen, A.S., Fu, G., Djordjević, S., Zhang, C., and Savić, D.A. (2018). "An integrated framework for high-resolution urban flood modelling considering multiple information sources and urban features." Environmental Modelling & Software, Vol. 107, pp. 85-95.

10.1016/j.envsoft.2018.06.010
36

Wei, C.-C. (2020). "Comparison of River basin water level forecasting methods: Sequential neural networks and multiple- input functional neural networks." Remote Sensing, MDPI, Vol. 12, 4172.

10.3390/rs12244172
37

Wing, O.E.J., Bates, P.D., Sampson, C.C., Smith, A.M., Johnson, K.A., and Erickson, T.A. (2017). "Validation of a 30 m resolution flood hazard model of the conterminous United States." Water Resources Research, Vol. 53, No. 9, pp. 7968-7986.

10.1002/2017WR020917
38

Xu, L., and Gao, L. (2024). "A hybrid surrogate model for real-time coastal urban flood prediction: An application to Macao." Journal of Hydrology, Elsevier, Vol. 642, 131863.

10.1016/j.jhydrol.2024.131863
39

Yamashita, R., Nishio, M., Do, R.K.G., and Togashi, K. (2018). "Convolutional neural networks: An overview and application in radiology." Insights Into Imaging, Vol. 9, No. 4, pp. 611-629.

10.1007/s13244-018-0639-929934920PMC6108980
40

Yonhap News Agency (YNA) (2014). Heavy rain of 130 mm per hour in Busan... Landslides and flood damage occurring one after another, accessed 12 August 2024, <https://www.yna.co.kr/view/AKR20140825155800051>.

41

Zhang, L., Qin, H., Mao, J., Cao, X., and Fu, G. (2023). "High temporal resolution urban flood prediction using attention-based LSTM models." Journal of Hydrology, Vol. 620, 129499.

10.1016/j.jhydrol.2023.129499
42

Zhou, Q., Teng, S., Situ, Z., Liao, X., Feng, J., Chen, G., Zhang, J., and Lu, Z. (2023). "A deep-learning-technique-based data- driven model for accurate and rapid flood predictions in temporal and spatial dimensions." Hydrology and Earth System Sciences, Vol. 27, No. 9, pp. 1791-1808.

10.5194/hess-27-1791-2023
43

Zhou, Y., Wu, W., Nathan, R., and Wang, Q.J. (2021). "A rapid flood inundation modelling framework using deep learning with spatial reduction and reconstruction." Environmental Modelling & Software, Vol. 143, 105112.

10.1016/j.envsoft.2021.105112
44

Zhu, H., Leandro, J., and Lin, Q. (2021). "Optimization of Artificial Neural Network (ANN) for maximum flood inundation forecasts." Water, MDPI, Vol. 13, No. 16, 2252.

10.3390/w13162252
Information
  • Publisher :KOREA WATER RESOURECES ASSOCIATION
  • Publisher(Ko) :한국수자원학회
  • Journal Title :Journal of Korea Water Resources Association
  • Journal Title(Ko) :한국수자원학회 논문집
  • Volume : 57
  • No :12
  • Pages :989-1001
  • Received Date : 2024-08-23
  • Revised Date : 2024-10-18
  • Accepted Date : 2024-11-07
  • DOI :https://doi.org/10.3741/JKWRA.2024.57.12.989
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