Describing the recurrence interval of extreme floods using nonextensive thermodynamics and Tsallis statistics

Abstract

This paper introduces Tsallis statistics and the q-exponential distribution as a means of analysing hydrological phenomena. The basic framework is introduced and then the method is used to derive a distribution for flood recurrence intervals using recently published data from the River Po, Italy. This fits the data much more effectively than the simple power-law applied in a previous study. Hence, a distribution derived from power-law considerations is more appropriate than the power-law itself. Nonextensive statistical mechanics has the potential for much broader utility in hydrology than is demonstrated here. Some potential avenues for future study are introduced. © 2005 Elsevier Ltd. All rights reserved.