Long-term temporal structure analysis of sediment fluxes in the Upper Changjiang (Yangtze River) using a wavelet transformation

Temporal fluctuations of suspended sediment fluxes are highly related with the flow and sediment regime of river networks. Flow dynamics are mainly triggered by climatic forces. Conversely, sediment regime is majorly controlled by land cover/land use changes and by human infrastructures (e.g. dams). Resulting sediment fluxes thus show a non-linear and non-stationary nature.  The wavelet transformation technique emerges as a skilful tool capable of dealing with this type of data. It allows to distinguish within the full spectrum of time-scales, those which are dominant and thus govern the overall trend of the time-series. Furthermore, the temporal multiresolution wavelet technique allows to accurately denoise the discharge and SSC time-series, filtering out the high frequency intermittent physical processes which are superimposed on the main signal. In this research, we analyse a long-term dataset of discharge and sediment fluxes time-series for the Upper Changjiang (Yangtze River, China).