Modelling subsurface flow conditions in a salinized catchment in south-western Australia, with a view to improving management practices

Abstract

Abstract:Finite element modelling of the saturated±unsaturated surface±subsurface¯ow mechanisms operative in a small salinized catchment in south-western Australia was used to help de®ne the¯ow system and explain the causes of waterlogging and salinization there. Data available at the site from a previous study were used to obtain a ®rst approximation to the¯ow system. Altering the properties of some of the strata gave a closer calibration. It was found that the modelled saturated hydraulic conductivity of the B horizon in the duplex soil zone needed to be at least an order of magnitude lower than that measured in order to reproduce the perching conditions observed in the ®eld. Also, the model indicated the in¯uence of a doleritic dyke, whose presence was con®rmed by ®eld measurement.Our analysis showed that there were two main¯ow systems operating in the hillslope. The ®rst, and most dominant, was the recharge occurring through the upslope gradational soil zone and percolating down to both the deeply weathered regolith and the basal aquifer. The second¯ow system is an unsaturated¯ow system operating in the high permeability A horizon in the downslope duplex soil zone. The ®rst system is primarily responsible for the saline seepage zone in the valley bottom. The second contributes to the waterlogging and perching occurring upslope of the seepage zone.Vertical¯ow through the higher permeability B horizon in the gradational soil zone in the upper slopes is a major contributor of recharge. Recharge by¯ow through macropores occurs where, but only where, perched aquifers develop and allow the macropores to be activated. Areas with perched aquifers occurred in downslope locations and near a doleritic dyke located upslope. Thus, the area where macropore recharge occurred was not large.The recharge rate required to maintain the piezometric levels at present values is only about 30 mm/yr (about 5% of the annual rainfall). The piezometric levels under the upper part of the catchment varied greatly with only small changes in recharge rate. A 50% reduction in recharge rate had the eect of reducing the length of the seepage zone at the end of winter by 40%. Changes in recharge rate had little eect on the extent of the perched aquifer at the end of winter. Deep-rooted perennial forages, shrubs or trees on the gradational soil zone in the upper part of the catchment and on the zones upslope of geological barriers to¯ow would be required to reduce the recharge and to allow for rehabilitation of the saline valley¯oor. Waterlogging associated with the perched water table in the bottom part of the catchment would be best addressed by tree plantations located just upslope of the salinized zone in the valley¯oor.

Influence of antecedent soil moisture on hydraulic conductivity in a series of texture‐contrast soils

“…Past efforts to model subsurface lateral flow in texturecontrast soils have usually represented the impeding layer of the B2 horizons as a single uniform layer, with a single value for hydraulic conductivity (Stolte et al, 1999;Cook and Rassam, 2002;Ticehurst et al, 2003a;Ticehurst et al, 2003b). While a number of studies have simulated water movement in texture-contrast soils using either single porosity or tipping bucket models (Silberstein et al, 1999;Stolte et al, 1999;Cook and Rassam, 2002;Ticehurst et al, 2003a;Ticehurst et al, 2003b), this study has demonstrated these approaches are not suitable for simulating water and solute movement in soils that contain vertic subsoils, water repellent topsoils or soluble silica bridging such as the texture-contrast soils investigated in this study. As Silberstein et al (1999) demonstrated in the Ucarro catchment, Western Australia, inability to simulate a reduction in subsoil hydraulic conductivity resulting from seasonal swelling of texturecontrast clay subsoils resulted in inadequate simulation of seasonal perched watertables and subsurface flow.…”

“…(), Stolte et al . () and Ticehurst et al . () have used simulation models to better predict or understand the occurrence of perched watertables and subsurface flows in texture‐contrast soils.…”

“…Studies such as Cook and Rassam (2002), Hatton et al (2002), Silberstein et al (1999), Stolte et al (1999) and Ticehurst et al (2003aTicehurst et al ( , 2003b have used simulation models to better predict or understand the occurrence of perched watertables and subsurface flows in texturecontrast soils. Current modelling of subsurface lateral flows, waterlogging and solute transport in texturecontrast soils is limited by simple model conceptualisation and poor understanding of the mechanisms by which antecedent soil moisture influences hydraulic conductivity.…”

“…Current modelling of subsurface lateral flows, waterlogging and solute transport in texturecontrast soils is limited by simple model conceptualisation and poor understanding of the mechanisms by which antecedent soil moisture influences hydraulic conductivity. Past modelling efforts have represented the B horizon as a uniform soil layer, parallel to the soil surface, in which hydraulic conductivity does not vary with seasonal soil moisture content, and macroporosity is either ignored or averaged within spatial variance of the bulk soil layer (Stolte et al, 1999;Ticehurst et al, 2003a;Ticehurst et al, 2003b).…”

Hydrological and salinity impacts of a bio‐drainage strategy application in the Yizre'el Valley, Israel

“…In light of theoretical considerations (Freeze and Cherry, 1979), field observations such as those in the Yizre'el Valley and our present measurements of off-plot lateral draw-down of the water table, we conclude that the primary impact of the bio-drainage is on the zones immediately beneath and just outside the plantation. Therefore, the location with the best prospects of controlling rising groundwater and the associated salinization by planting trees is in the discharge zone, and not some distance upslope in the recharge zone (Morris, 1991;Schofield and Bari, 1991;Stole et al, 1999;Benyon et al, 2001), particularly if the flows in the groundwater system are on a regional scale (Stirzaker et al, 2002). All the hills around the Yizre'el Valley are densely covered with planted forests, but this did not reduce, let alone prevent, the groundwater rise and associated widespread salinization that have plagued the valley since the mid 1980s.…”