Bank permeability in an Australian ephemeral dry‐land stream: variation with stage resulting from mud deposition and sediment clogging

Abstract

Percolation of flood waters into the bed and banks of ephemeral streams provides one of the key mechanisms responsible for transmission loss. However, there are very few published estimates of the rates at which water can enter stream-bank sediments, and little is known about the variation in bank permeability with elevation above the bed and the resulting effects on transmission loss in floods of different magnitudes. This paper presents the results of 69 field determinations of bank infiltrability made on Fowlers Creek, an ephemeral dryland stream located in arid western New South Wales, Australia. Fowlers Creek carries high concentrations of suspended sediments, which are deposited as mud drapes on the bed, banks and floodplain. Results demonstrate that infiltration rates are lowest at the base of the banks, and tend to increase steadily with elevation on the bank, even above the apparent upper limit of mud drapes. In parallel, the texture of the bank sediments (assessed from samples of the uppermost 10 cm) becomes coarser with elevation above the bed. This pattern is inferred to relate to the delivery of silts and clays into pore spaces in the bank sediments by percolating flood waters. The patterns of infiltration rate and sediment texture mapped in the field are reasoned to be the product of many clogging episodes in past flood events having different peak stages. The increase in infiltration rate and mean particle size up the banks reflects lower frequencies of submergence and clogging of the upper banks by large floods, and more frequent inundation and clogging of the lower banks by sub-bank-full flows. The stage-related changes in bank permeability provide a mechanism that can drive variations in transmission loss among floods having different peak stages and hydrograph shapes.larger; see, e.g., Coes and Pool, 2005, for data from south-eastern Arizona). Thus, the percolation of water into the bed and banks is the dominant transmission loss mechanism during the hours-to-days durations typical of flood flows in ephemeral streams, and these processes form the subject of the present paper. Infiltration and percolation are dependent upon a range of local site conditions, including the texture and hydraulic properties of the channel boundary sediments (sorting, particle sizes, porosity, permeability), features of the flow hydrograph, including the stage and duration of the flow peak and of the entire hydrograph, and the suspended sediment load carried by the flow (particle size distribution, sediment concentration). In very large river systems, such as the Darling River of inland Australia, flood waters take months to progress downstream toward terminal lake basins, and under these circumstances evapotranspiration losses presumably become an increasingly significant component of water loss. Additionally, in these very large exogenic river systems crossing low-gradient alluvial plains, the channel boundary sediments are fine textured and this restricts the seepage losses in comparison with sand and gravel bed s...

Distinctiveness and Diversity of Arid Zone River Systems

“…size of the wetted perimeter, porosity and initial moisture content of the perimeter sediments, stratigraphy of the channel fill) (e.g. Sorman and Abdulrazzak, 1993;Knighton and Nanson, 1994b, 1997Sharma, Murthy and Dhir, 1994;Dick, Anderson and Sampson, 1997;Greenbaum et al, 1998;Lekach et al, 1998;Dunkerley and Brown, 1999;Lange, 2005;Dunkerley, 2008b). Storm characteristics and resulting hydrograph events are important in very large catchments or in catchments subject to discrete convective storms, for individual storms are unlikely to wet the entire catchment while successive storms are likely to wet different parts of the catchment and may produce compound rather than single floodwaves (Cooke, Warren and Goudie, 1993;Knighton and Nanson, 2001).…”

A semi-analytical generalized Hvorslev formula for estimating riverbed hydraulic conductivity with an open-ended standpipe permeameter

“…have been widely applied to investigate the hydraulic properties of riverbed sediments (Cheong et al, 2008;Kalbus et al, 2006;Wang et al, 2015), the accurate estimation of riverbed K values remains a challenge. One of the challenging aspects of estimating riverbed K is associated with its high spatial and temporal variability across measurement scales due to heterogeneity in the riverbed sediments (Chen et al, 2010), scouring and depositional processes during flooding events (Dunkerley, 2008;Hatch et al, 2010), and diurnal and seasonal changes in stream flow temperature (Constantz, 1998). Additionally, the successful application of the aforementioned methods is highly dependent on the assumptions and limitations of the applied methods, the specific equipment, and the design of the measurements (Shanafield and Cook, 2014).…”

Where does all the water go? Partitioning water transmission losses in a data-sparse, multi-channel and low-gradient dryland river system using modelling and remote sensing

“…Estimating total transmission losses and/or individual components in dryland river systems has previously been undertaken using three main approaches (Cataldo et al, 2004;Cataldo et al, 2010): (i) small-scale field experiments (Dahan et al, 2008;Dunkerley and Brown, 1999;Dunkerley, 2008;Maurer, 2002;Parsons et al, 1999); (ii) interpolation of sparse streamflow networks using simple regression and/or differential equations (Arnott et al, 2009;Costelloe et al, 2006;Knighton and Nanson, 1994;Knighton and Nanson, 2001;McCallum et al, 2012;Schmadel et al, 2010); and (iii) water balance modelling to allow estimation of total and component transmission losses (Morin et al, 2009). Key papers for these approaches are summarised in Table 1, and includes examples where hydrodynamic modelling has incorporated remotely sensed data in order to: (i) provide input data; (ii) calibrate and validate such models; and (iii) estimate various components of transmission losses (Karim et al, 2011;Milewski et al, 2009;Sharma and Murthy, 1994).…”