Factors controlling the spatial distribution of flash flooding in the complex environment of a metropolitan urban area. The case of Athens 2013 flash flood event. (September 2016)

• Record Type: Journal Article
• Authors: Diakakis, Michalis; Deligiannakis, Georgios; Pallikarakis, Aggelos; Skordoulis, Michalis
• Is Part Of: International journal of disaster risk reduction. Volume 18(2016)
• Page Start: 171
• Page End: 180
• Publication Date: 2016-09
• Subjects: Flash flood -- Urban flood -- Logistic regression -- Flood susceptibility -- Athens -- Greece; Emergency management -- Periodicals; Risk management -- Periodicals; Disaster relief -- Periodicals; Hazard mitigation -- Periodicals; 363.34
• DOI: 10.1016/j.ijdrr.2016.06.010
• Languages: English
• ISSNs: 2212-4209

Abstract

Abstract: This work studies the influence of various factors on the spatial distribution of flash flooding in urban areas. It examines an extreme precipitation event that occurred in 2013, in Athens, Greece, causing flooding and damages in various parts of the city. GIS was used to determine the values of potentially flood-influencing parameters, while logistic regression was used to examine the statistical correlation between their values and the respective values at the locations of 1138 flooded buildings. Statistical significance and other metrics were used to measure the influence of these factors on the distribution of flooding. Results showed that certain factors, including the degree of soil sealing, accumulated rainfall, slope and others, influence the distribution of flooding to significantly different degrees. Incorporation of building characteristics into the regression model increased further its coherence. The study showed a strong relationship between a combination of factors and flood occurrence with R 2 reaching 0.63 and a prediction accuracy of 82.9%. The degree of soil sealing and slope were the most influential in determining the distribution of flooding. However, when building parameters were incorporated, the level of a building's lower opening in comparison with ground level was by a significant margin the most important parameter. Highlights: Spatial correlation of different factors with flood occurrence is examined. Logistic regression is used to examine and quantify their statistical significance. Results show strong positive or negative correlation of certain factors. Meteorological, geomorphological and building factors have influence. Using these factors model can predict at a rate of 82.9% flood occurrence locations.