Seasonal fractal-scaling of floods in two US water resources regions

Understanding the behavior and estimating the magnitude of floods with specific recurrence intervals are important tasks for various applications such as flood protection strategies. Fractal analysis has proven useful in characterization of flood frequency behavior. We employ a systematic fractal approach which enables dividing streamflow data into different behavior regimes and, in particular, identifying flood regimes. Since seasonality is a key factor in flood-formation scenarios, we incorporate this concept in our analysis through generating two separate streamflow data sets for summer and winter, and next performing associated fractal analysis on each. To illustrate our approach and see how it enhances our flood magnitude estimations, we study a large number of watersheds chosen from two adjacent water resources regions in the United States. Besides improving flood-magnitude estimates, this seasonal fractal analysis can suggest patterns through which one may obtain more fine-grained information about regional behavior of floods. (C) 2016 Elsevier B.V. All rights reserved.