Abstract

A long-term database with information on precipitation, discharge and sediment yield dynamics was analyzed to characterize the temporal structure response of a catchment in a humid mountain badland area. Due to the non-linearity and non-stationary nature of the data, a combination of frequency analysis and wavelet reconstruction techniques were initially used to determine the dominant time scales of the precipitation, discharge and sediment yield. Within the full spectrum of time scales, the spectral frequency analysis could distinguish the dominant ones governing the overall trend of the time-series. Furthermore, the temporal multiresolution wavelet technique enabled accurate de-noising of the hydrometric and sediment time-series, by filtering out the high frequency intermittent processes superimposed on the main signal. Geomorphological metrics were subsequently applied to the filtered data and link to the temporal structure of the catchment response. The combined frequency-geomorphological analysis provides a physical explanation of the complete temporal structure of the catchment response to precipitation. This analysis features components from intra-annual, annual and multi-year scales. Intra-annual and annual time scales are led by climatological characteristics of the catchment site (seasonal rainfall patterns of a mountain Pyrenean catchment). The multi-year response related to the sediment yield reveals the importance of the sediment storage/depletion cycle in the catchment: although the main driver of the sediment yield is the discharge, weathering processes and sediment storage are also major control factors. Therefore, correct management of the catchment requires the combined multi-temporal response of water and sediment fluxes to be controlled. The temporal spectral analysis herein outlined provides a detailed analysis of the long-term temporal structure of databases and can be accepted as an adequate tool for catchment management in terms of flood forecasting or reservoir operation.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call