Evaluation of Hurst exponent for precipitation time series

A major issue in time series analysis and particularly in the study of meteorological time series behaviour is the long range dependence (LRD). Various estimators of LRD have been proposed. Their accuracy have been generally tested by using simulated time series since sometimes only their asymptotic property are known, or worse, no asymptotic property have been proved. It is well – known that the Hurst exponent (H) is a statistical measure used to classify time series. In this article we determine the Hurst exponent for precipitation time series collected in Dobrudja region, for 41 years and we compare the results. Key-Words: Long – range dependence, Hurst coefficient, precipitation, detrended fluctuation