A measure for transformation of the input signal in the soil-groundwater-stream continuum and implications for modelling

Abstract:

Most of hydrological research in fact focuses on the interplay between structure and behaviour. To what degree does the dynamics of stream discharge or soil water content depend on properties of the catchment like land use, stream network or soil texture? The structure of a catchment can be determined by different approaches. However, hydrological behaviour that manifests itself in hydrological time series has so far only been analysed using either forward modelling approaches or crude assessments that make use of only a minor fraction of the information provided by observed hydrological time series. Consequently, analysing hydrological behaviour in a quantitative way and comparing behaviour at different sites has been a tedious task so far.

We developed a method to measure the degree and the kind of transformation of the input signal at different depths and sites within in a single or in adjacent catchments. It is based on a dimensionality reduction of joint sets of time series of soil matrix potential, soil water content, groundwater head, lake water level and/or stream discharge using either linear or nonlinear approaches. So far, published results concern different data sets of groundwater head, lake water level and stream discharge data (Lewandowski et al. 2009, Lischeid et al. 2010, Page et al. 2012, Thomas et al. 2012).

Results for soil hydrological data from various catchments strongly suggest that the kind of transformation in these soils is approximately the same irrespective of differing soil texture. Soils differed only with respect to the degree of transformation per depth interval. We did not find any evidence that preferential flow processes had a substantial impact on the kind of transformation, although they clearly reduced the degree of transformation per depth interval.

It could be shown that time series of groundwater head and stream discharge can be regarded as a superposition of the contribution of different flowpaths in the subsoil, whereas processes in the aquifer did not seem to play any role. We conclude that transformation of the input signal throughout a catchment essentially follows a single common trajectory throughout the soil-groundwater-stream continuum.

This methodology can be used to measure transformation of the input signal directly on any observed time series of soil matrix potential, soil water content, groundwater head or stream discharge without requiring any other assumption than that the output signal is a superposition of independent components. Thus, observed behaviour at different sites can be compared and can be used for analysis of the soil-groundwater-stream continuum. These findings have substantial implications for modelling hydrological processes in catchments. They can be used both for thorough model tests as well as for developing parsimonious models that make efficient use of the observed data. Besides, our findings clearly point to a continuum of catchment runoff processes and challenge the common distinction of different runoff processes.

References:

Lewandowski, J., Lischeid, G., Nützmann, G. (2009): Drivers of water level

fluctuations and hydrological exchange between groundwater and surface water at the lowland River Spree (Germany): field study and statistical analyses. Hydrological Processes 23: 2117-2128, DOI: 10.1002/hyp.7277

Lischeid, G., Natkhin, M., Steidl, J., Dietrich, O., Dannowski, R., Merz, C. (2010): Assessing coupling between lakes and layered aquifers in a complex Pleistocene landscape based on water level dynamics. Advances in Water Resources 33: 1331-1339, DOI: 10.1016/j.advwatres.2010.08.002

Page, R., Lischeid, G., Epting, J., Huggenberger, P. (2012): Principal component analysis of time series for identifying indicator variables for riverine groundwater extraction. Journal of Hydrology 432-433: 137-144, DOI: 10.1016/j.jhydrol.2012.02.025

Thomas, B., Lischeid, G., Steidl, J., Dannowski, R. (2012): Regional catchment classification with respect to low flow risk in a Pleistocene landscape. Journal of Hydrology 475: 392-402, DOI: 10.1016/j.jhydrol.2012.10.020