Continuous baseflow separation from time series of daily and monthly streamflow data

Abstract

Continuous basefiow separation procedures have been frequently used to differentiate total flows into the high-frequency, low-amplitude 'baseflow' component and the low-frequency, high-amplitude 'flood' flows. In the past, such procedures have normally been applied to streamflow time-series data with time steps of 1 day or less. However, there are applications in South Africa (notably related to setting instream flow requirements) where the only available data for natural flow conditions are monthly flow volumes. A relatively experienced hydrologist can be expected to successfully calibrate a separation model using daily data, coupled with a conceptual understanding of the hydrological processes prevailing in the catchment. The same cannot be said for monthly data, as the majority of the information on short-term flow variability has been lost. As part of a regional study covering the whole of South Africa, this paper presents some example results of comparisons between daily and monthly separations. While it can be concluded that it is possible to determine regionalised parameters for monthly data separations that are useful, further information on the processes involved would be of great value to validate the methods and parameter values. This information could also form the basis for the further develonment of baseflow separation methods for South African flow regimes.