The frequency of localized short-term torrential rains that exceed the planned rainfall is increasing along with inundation damage due to inland flooding. Stepwise inundation measures utilizing existing stock and disaster prevention/mitigation for excessive rainfall are required. In this study, we describe the results of empirical research using a statistical model constructed based on rainfall and water level observation data as a highly accurate water level prediction method suitable for real-time prediction. This is aimed at application in flood control activities and operation support of pump facilities. By comparing and verifying the prediction accuracy between the water level prediction model and the statistical model by Convolutional Neural Network (CNN), which is generally used as an image recognition technology, the usefulness of the statistical model was confirmed. Further improvement in accuracy and widespread use of these water level prediction models are expected.

1. [1] Y. Tachikawa, G. Nagatani, and K. Takara, “Development of stage-discharge relationship equation incorporating saturated-unsaturated flow mechanism,” Proc. of Hydraulic Engineering, Vol.48, pp. 7-12, 2004 (in Japanese).
2. [2] O. Suga, “Real inundation map system and method for inundation forecast and drainage control,” Japan Patent, JP2002-298063A, 2002.
3. [3] K. Mizukusa and K. Hiroki, “Development of the flood analysis model incorporating the sewage system,” Civil Engineering J., Vol.45, No.12, pp. 40-45, 2003 (in Japanese).
4. [4] M. Takeda, Y. Morita, and N. Matsuo, “Application of inundation analysis considering sewer system to measure for flood disaster prevention,” Proc. of Hydraulic Engineering, Vol.51, pp. 529-534, 2007 (in Japanese).
5. [5] J. Akiyama, M. Shigeeda, Y. Kozono, and H. Kusano, “Numerical simulations of inundation flows with flood control system in Iizuka City and examination of its effects on flood hazard mitigation,” J. of Japan Society of Civil Engineers, Ser. B1 (Hydraulic Engineering), Vol.67, No.4, pp. I_943-I_948, 2011 (in Japanese).
6. [6] M. Kimura, Y. Kido, and E. Nakakita, “Fundamental study on real-time flood forecasting method for locally heavy rainfall in urban drainage areas,” J. of Japan Society of Civil Engineers, Ser. B1 (Hydraulic Engineering), Vol.67, No.4, pp. I_931-I_936, 2011 (in Japanese).
7. [7] M. Kimura, Y. Kido, and E. Nakakita, “Simplification method on real-time flood forecasting for locally heavy rainfall in urban drainage areas,” J. of Japan Society of Civil Engineers, Ser. B1 (Hydraulic Engineering), Vol.68, No.4, pp. I_985-I_990, 2012 (in Japanese).
8. [8] A. N. A. Schellart, W. J. Shepherd, and A. J. Saul, “Influence of rainfall estimation error and spatial variability on sewer flow prediction at a small urban scale,” Advances in Water Resources, Vol.45, pp. 65-75, 2012.
9. [9] M. Kitano, T. Wada, and S. Itoshiro, “Flood analysis method for water disasters in urban area,” Proc. of the 6th Symp. on Urban Flood Disasters, pp. 17-23, 2007 (in Japanese).
10. [10] S. V. Ackere, J. Verbeurgt, L. D. Sloover, S. Gautama, A. D. Wulf, and P. D. Maeyer, “A Review of the Internet of Floods: Near Real-Time Detection of a Flood Event and its Impact,” Water, Vol.11, No.11, 2019.
11. [11] D. Zhang, G. Lindholm, and H. Ratnaweera, “Use long short-term memory to enhance Internet of Things for combined sewer overflow monitoring,” J. of Hydrology, Vol.556, pp. 409-418, 2018.
12. [12] B. Kerkez, C. Gruden, M. Lewis, L. Montestruque, M. Quigley, B. Wong, A. Bedig, R. Kertesz, T. Braun, O. Cadwalader, A. Poresky, and C. Pak, “Smarter Stormwater Systems,” Environmental Science & Technology, Vol.50, No.14, pp. 7267-7273, 2016.
13. [13] H. R. Maier and G. C. Dandy, “Neural networks for the prediction and forecasting of water resources variables: a review of modelling issues and applications,” Environmental Modelling & Software, Vol.15, No.1, pp. 101-124, 2000.
14. [14] H. R. Maier, A. Jain, G. C. Dandy, and K. P. Sudheer, “Methods used for the development of neural networks for the prediction of water resource variables in river systems: Current status and future directions,” Environmental modelling & Software, Vol.25, No.8, pp. 891-909, 2010.
15. [15] T. Yamada, “A neural network for rainwater inflow amount predicting device,” Japan Patent, JP2955413, 1999.
16. [16] ASCE Task Committee on Application of Artificial Neural Networks in Hydrology, “Artificial neural networks in hydrology. II: Hydrologic Applications,” J. of Hydrologic Engineering, Vol.5, No.2, pp. 124-137, 2000.
17. [17] C. W. Dawson and R. L. Wilby, “Hydrological modeling using artificial neural networks,” Progress in Physical Geography, Vol.25, No.1, pp. 80-108, 2001.
18. [18] M. Hitokoto, M. Sakuraba, and S. Sakamoto, “Method comparison for reduction of the flood prediction uncertainty using real-time hydrological observation data,” Advances in River Engineering, Vol.21, pp. 431-436, 2015 (in Japanese).
19. [19] M. Hitokoto, M. Sakuraba, and Y. Sei, “Development of the real-time river stage prediction method using deep learning,” J. of Japan Society of Civil Engineers, Ser. B1 (Hydraulic Engineering), Vol.72, No.4, pp. I_187-I_192, 2016 (in Japanese).
20. [20] M. Hitokoto and M. Sakuraba, “Development of deep learning flood prediction model and future perspectives,” Nippon Koei Technical Forum, Vol.26, pp. 1-7, 2018 (in Japanese).
21. [21] Y. Kimura, N. Takase, H. Fukai, M. Niwakawa, and M. Nakashima, “Prediction of Water Level in Sewer Pipe using Convolutional Neural Network,” The Papers of Technical Meeting on “Perception Information,” IEE Japan, PI, pp. 1-3, 2018 (in Japanese).