Extraction of flooding features with multifractal analysis for the Wujiang River of South China

Abstract

Flooding is often affected by multiple factors such as climate, hydrology, infrastructures, etc. This paper attempted to examine multifractal methodology as a tool to characterize the indicators for flood magnitude. Results from this study have shown that the multifractal analysis can be an encouraging method to identify the dynamic features of flood process, particularly for rivers in the mountainous region such as the Wujiang River in this study. The flood peak, total flood volume, the 24-, 48-, and 72-h flood volumes were selected as flood magnitude indicators in this study. It was found that the five indicators showed strong multifractal feature. The singularity spectrum showed the flood volume had the highest fluctuation among all five indicators. The span of the multifractal singularity showed the flood volume had stronger multifractal characteristic. The flood volume was sensitive to local fluctuations, but other four indicators were not sensitive to fluctuations. This study also showed, for the Wujiang River, multifractality were in the order from greatest multifractal strength to the smallest were the flood volume, the maximum 24-h flood volume, the peak flood, the maximum 48-h flood volume and the maximum 72-h flood volume, which was due to the terrain features in Wujiang River and its drainage area.