Artificial Intelligence and Data Science
Online ISSN : 2435-9262
Development of an inner water level forecasting and gate/drainage operation optimi-zation method for urban rivers linked with storm surge predictions
Kenta SAITO, Atsuki OBA, Soma WATAHIKI, Hiroki TAKIKAWA, Shunya OKUNO, Yui SATO
Author information
JOURNAL OPEN ACCESS

2026 Volume 7 Issue 3 Pages 618-629

Details
Abstract

In urban canals exposed to the combined risks of storm surges and inland flooding, timely and appropri-ate operation of disaster prevention facilities, such as storm surge gates and drainage pump stations, is extremely critical. To support rapid on-site decision-making, this study develops a real-time decision-support system that seamlessly integrates a highly accurate outer water level (tidal level) forecasting model based on dynamical systems theory, a tank-model-based physical model for inner water level prediction, and a facility operation optimization algorithm.

Focusing on the canals in the Koto and Shibaura districts under the jurisdiction of the Bureau of Port and Harbor, Tokyo Metropolitan Government, we refine the discharge forecasting from sewerage facilities using data assimilation and Bayesian regression with past flood data. In addition, a logic is developed to deterministically calculate appropriate facility operations linked directly to the predicted water level timelines. Validation results using past storm events demonstrated that the proposed system forecasts both inner and outer water levels as well as facility operation timelines with high accuracy sufficient for practical operations, confirming its applicability as a decision-support tool.

Content from these authors
© 2026 Japan Society of Civil Engineers
Previous article Next article
feedback
Top