Flood prioritization of basins based on geomorphometric properties using principal component analysis, morphometric analysis and Redvan’s priority methods: A case study of Harşit River basin

Abstract

Flood events in the Harşit River Basin, which is located in the Eastern Black Sea Region and actively experienced one of the most flood events in Turkey, were discussed based on drainage basin morphometry using three different methods. In this study, 26 sub-basins over 10 km2 are taken into account and flood dynamics of sub-basins are evaluated depending on morphometric properties. The results obtained according to the morphometric parameters are evaluated based on statistical techniques, and the flood priority is determined. A total of 20 different indices were used in this study. The obtained morphometry results were evaluated according to the probability of occurrence of floods according to three different methods including morphometric priority, flood priority according to principal component analysis method and Redvan’s priority ranking method. Each method is evaluated within itself and flood priorities of sub-basins have been determined. With respect to the 26 sub-basins of the Harşit River and according to the morphometric priority method and results obtained from this method, 8 sub-basins have low flood priority, 12 basins have medium flood priority, and 7 basins have high flood priority. By examining the results obtained from the PCA priority method, 11 sub-basins have low flood priority, 6 sub-basins have medium flood priority, while 9 basins have high flood priority. Finally, according to the Redvan’s priority method and its classification, 11 sub-basins are located in the low priority class, 6 sub-basins have medium flood priority. According to this method, the number of sub-basins with high priority is 9. Priorities obtained using three different techniques were checked for accuracy with 6 different statistical parameters based on predicted and inventory datasets. Accordingly, the accuracy value for the RPRM and PCA methods are higher than the MA method. On the other hand, the RPRM method has the highest TN value, which is equal to 7 sub-basins, which shows the correct prediction of sub-basins with the low flood risk. Based on the results, it was seen that the obtained values have a high consistency in basin morphometry. The RPRM, which is a new suggested technique in determining basin flood priority, shows that common basins give very similar results with the results obtained according to morphometry and PCA method. Sub-basins with common flood risk, due to outputs of used methods, were compared with recorded floods map of Turkey.