Application of Open Building Footprints Data in Flood Impact Assessment, a Case Study of Can Tho Province

Abstract

Floods, the most common disaster in Vietnam, occur due to climatic events (such as storms and rains), discharges from multinational basins, and tidal activities. The Mekong Delta is particularly known for the convergence of these three water sources. In recent years, a significant amount of research has been dedicated to flood modeling and risk assessment, utilizing data from satellites and hydraulic modeling. Residential data plays a crucial role in flood risk assessment, but its sources are limited. Recently, Google and Microsoft have released two sources of residential building data that are freely available to the public. This paper employs Google’s Open Building Footprint data and Building data extracted from the Landuse map to assess the impact of floods on the built-up area in Can Tho city. The flood data, provided by MIKE hydraulic modeling, includes return periods of 100, 50, 20, and 5 years (1%, 2%, 5%, and 20% respectively). The current flood data is classified using the Australian Flood Hazard system, which consists of five levels of hazard ranging from Caution to Danger for all. The two building datasets have been overlaid with the flood data in a GIS environment using GIS operations. A geostatistical function has been applied to determine the impact of the flood on the buildings. The results indicate that the area impacted by the flood is smaller for the Google Open Building than for the Landuse data. The primary difference is that the Google Open Building Footprint was created using satellite imagery and focuses on building footprints, while the Landuse data represents the type of land use assigned to the land parcel, which may be significantly larger than the building footprint. In conclusion, the Open Building Footprint demonstrates high potential for use in flood impact assessment.