Relationship between vehicle probe data and flooding conditions for developing flood inundation monitoring method

Heavy rainfall disasters caused by climate change have become increasingly severe and frequent, causing extensive flood damage worldwide. Understanding when and where river flooding occurs and how far it has spread in real time is essential for mitigating human suffering. This study aims to clarify the relationship between vehicle traffic information obtained from vehicle probe data and actual flood inundation conditions for real-time flood inundation monitoring using vehicle probe data. To this end, large-scale flood inundation in the Kuma River in Kyushu District, Japan (July 2020) was set as the target flood condition. The inundation depth was extracted from the results of the integrated analysis of river and inundation flows in and around the Kuma River using a numerical model and compared with vehicle traffic data. The results show that the presence or absence of a vehicle traffic area corresponds to the expansion or contraction of the inundation area during flooding, verifying the validity of the inundation area estimation using vehicle traffic information. The results indicate that there is a negative correlation between vehicle speed, flow velocity, and inundation depth, and vehicle speed is an important indicator to understand flood inundation conditions. Road links with no vehicle traffic or vehicle speeds below 20 km/h at selected road classes are considered inundated and organized into a 250 m resolution mesh. The results quantitatively show that the inundation area could be approximately estimated, suggesting the validity of this inundation area estimation method. • Developed flood inundation monitoring method using vehicle probe data and verified the accuracy. • Clarified relation between vehicle traffic data and flood inundation conditions. • Negative correlation exists between vehicle speed, flow velocity, and inundation depth.