Long-term variations and temporal scaling of hydroclimatic time series with focus on the German part of the Elbe River Basin



Long-term variations and temporal scaling of mean monthly time series of river flow, precipitation, temperature, relative humidity, air pressure, duration of bright sunshine, degree of cloud cover, short wave radiation, wind speed and potential evaporation within or in vicinity of the German part of the Elbe River Basin are analyzed. Statistically significant correlations between the 2–15 year scale-averaged wavelet spectra of the hydroclimatic variables and the North Atlantic Oscillation- and Arctic Oscillation index are found which suggests that such long-term patterns in hydroclimatic time series are externally forced. The Hurst parameter estimates (H) based on the Detrended Fluctuation Analysis (DFA) indicate persistence for discharge, precipitation, wind speed, air pressure and the degree of cloud cover, all having an annual cycle and a broad low-frequency distribution. Also, DFA H parameter estimates are higher for discharge than for precipitation. The major long-term quasi-periodic variability modes of precipitation detected using Singular Spectrum Analysis coincide with those detected in the discharge time series. Upon subtraction of these low-frequency quasi-periodic modes, the DFA H parameter estimates suggest absence of the persistence for both precipitation and discharge. Copyright © 2013 John Wiley & Sons, Ltd.