Keith Smettem

Understanding how summer low flows in a Mediterranean climate are influenced by climate and land use is critical for managing both water resources and in-stream ecohydrological health. The Eucalyptus forest ecosystems of southwestern Australia are experiencing a drying and warming climate, with a regional step decline in rainfall in the mid-1970s. Reductions in catchment water storage may be exacerbated by the deep rooting habit of key overstorey species (>30 m has been reported), which can buffer against drought during dry years. Root exploitation of deep soil moisture reserves and/or groundwater can accelerate the long term decline in summer low flows, with a trend towards more ephemeral flow regimes. In contrast, conversion of forests to agricultural land in some catchments can lead to counter-trends of increased low flows due to a rise in groundwater pressure. These are invariably associated with an increase in stream salinity as regolith stores of salt are mobilized. There has also been extennsive reforestation of farmland in some catchments. In this study we perform a detailed analysis of changes to annual summer seven day low flow trends in perennial catchments and flow duration curves in ephemeral catchments across 39 catchments in south-western Australia that have long term records of runoff, rainfall and land cover. Results showed that 15% of catchments exhibited increased low flows and 85% decreased flows or decreased flow days since the 1970s. Significant downward step changes in low flows were observed in 17 catchments (44%). The earliest downward step changes occurred in three catchments between 1981-82 (a lag of one decade after the rainfall decline), with the most recent step changes for five catchments occurring in 2001-2004 (three decades after rainfall decline). Eleven catchments were already ephemeral in the 1970s, but exhibited continued declines in the number of annual flow days over subsequent decades. Step changes occur when groundwater becomes disconnected or reconnected to the stream invert, with disconnection associated with rainfall decline and vegetative water use. The statistical methods we used in this study can be applied to any catchment in order to aid land and water managers assess the impact of climate change and land cover manipulation on low flow response.

There is an established need to measure soil salinity, and wireless sensor networks offer the potential to achieve this, coupled with a suitable sensor. However, suitable sensors, up until very recently, have not been available. In this paper we report on the fabrication and calibration of a new low-cost, robust, screen-printed sensor for detecting chloride ions. We also report on two experiments using this sensor. The first is a laboratory-based experiment that shows how sensors can be used to validate modeling results by installing several sensors in a soil column and tracking the vertical migration of a chloride pulse in real time. The second is a trial of multiple sensors installed in a fluvarium (stream simulator) showing that distributed sensors are able to monitor real time changes in horizontal chloride flux in an emulated natural environment. We report on results from both surface flows as well as from sensors at a depth of a few mm in the fluvarium sediment, and differences and trends between the two are discussed. As an example of how such sensors are useful, we note that for the flow regime and sediment type tested, penetration of surface chloride into the river bed is unexpectedly slow and raises questions regarding the dynamics of pollutants in such systems. We conclude that such sensors, coupled with a distributed network, offer a new paradigm in hydrological monitoring and will enable new applications, such as irrigation using mixtures of potable and brackish water with significant cost and resource saving.

Surface water and groundwater (SW-GW) interaction plays a key role in replenishing alluvial aquifers and sustaining the ecology, quality, and quantity of streams and rivers. The interaction, mainly through the hyporheic zone (HZ), also plays a role in the nutrient exchange between terrestrial and aquatic ecosystems. Hyporheic water exchange in many sandy coastal plain drains presents an intermittent hydrological regime, as often shifts occur in their hydraulic functioning from a losing to a gaining stream condition upon the position of the surrounding water table. This work documented the existence and the complex hydrodynamics of HZ water exchange in an artificial drain typical of a coastal plain area (Mayfield drain, Harvey River) in Western Australia (WA), which resulted from a combination of highly responsive water level regime in the drain and a seasonal-transient shallow water table developed in a duplex soil (sand over clay) setting. A novel hydrometric approach using a rugged field camera (water levels on the drain), automatic water level sensors (bores), and a set of temperature sensors (drain?s bed and bank) provided a robust data set to explore vertical water exchange (fluxes) under baseflow and storm event conditions for different hydraulic scenarios (water stages for drain and water table) across the wet season. Water fluxes and direction in the HZ were computed using the one dimensional (1D) heat transport model for pore water (VFLUX) from temperature data, and further verified by a standard hydraulic approach using Darcy?s law and hydrometric data. The results indicated that under baseflow conditions, vertical water flux estimates were directed downwards, of drain water into a shallow sandy layer (< 0.4 m), but with an upward flux from the underneath clay layer (0.4 - 0.7 m). The magnitude and temporal variability of the fluxes corresponded with the drain water stage as increasing downward fluxes due to higher drain stage resulted in decreasing upward fluxes from the clay layer. During storm event conditions, a similar water exchange direction was observed but a substantial increase in the downward flux (by an order of magnitude) of drain water occurred without any change on the upward flux. This dynamics showed a dependency on both mean drain storm water stage and the duration of high flow conditions. These results and the observations of high water table level at the bank indicated that the excess of water into the HZ was mainly mobilized to downstream locations along the drain bed. Further analysis of the gaining and losing water condition over a 620 m drain reach, using flow measurements and water quality surveys, supported VFLUX results. This work identified the presence of a shallow HZ confined vertically by a clay layer (typical feature of duplex soils in the area) and laterally by a high water table in the bank, under both baseflow and stormflow conditions present in the drain. The findings also highlighted the importance of flux estimation using thermal records to complement traditional hydraulic approaches due to lack of conclusive results provided by the latter.

A dynamic model of Phosphorus (P) movement through the Peel-Harvey Watershed in South Western Australia was developed using STELLA dynamic modelling software. The model was developed to illustrate watershed P flux and to predict future P loss rates under a range of management scenarios. Input parameters were sourced from surveys of local agricultural practices and regional soil testing data. Model P-routing routines were developed from the known interactions between the various watershed P compartments and fluxes between various P stores. P-retention characteristics of a variety of management practices were determined from field trials where available and published values where not. The model simulated a 200 year time frame to reflect 100 years to the present day since initial land development, and forecast 100 years into the future. Although the watershed has an annual P loss target of 70 tonnes per annum (tpa), the measured present day loss is double this amount (140 tpa) and this is projected to rise to 1300 tpa if current land management practices continue. Even if broad-scale BMP implementation occurs, P losses are likely to increase to approximately 200 tpa. This has significant implications for both future land use and subsequent water quality in the watershed.

A biophysical model WAVES, developed by CSIRO, has recently been implemented as the modelling engine to provide recharge input and discharge output to an integrated regional groundwater model for the Perth region, namely, the Perth Regional Aquifer Modelling System (PRAMS). Application of the WAVES model over such a large area requires an extensive set of data that are not known with certainty. This paper describes a method based on the mean value first order reliability analysis method to determine the parameters that significantly affect uncertainty in the recharge estimates. The fraction of model output (recharge) variance (FOV) contributed by each basic parameter is determined using the sensitivity coefficients and uncertainty (measured by the variance or the coefficient of variation) of the model parameters. The FOV provides a quantitative means to rank the order of importance of parameters that affect the reliability of the recharge estimate and can be used to prioritise data collection to improve the estimate. Application of the proposed method to the local conditions on the Gnangara Mound found that the critical parameters for WAVES in the modelling framework of PRAMS are: rainfall, vegetation density measured by leaf area index, soil water holding capacity of the soil root zone, maximum root depth and maximum carbon assimilation rate.

Salinization threatens up to 17 million ha of Australian farmland, major fresh water resources, biodiversity and built infrastructure. In higher rainfall (>600 mm yr-1) areas of south-western Australia a market-based approach has resulted in the reforestation of 250,000 ha of farmland with Eucalyptus globulus pulpwood plantations. This has had significant collateral environmental benefits in terms of reducing salinity and restoring landscape function in several key water supply watersheds. This success has not been replicated in the lower (300-600 mm yr-1) rainfall areas of this region, which is a global biodiversity hotspot. Wood yields are lower and there is often a land-holder preference to maintain existing agricultural activities. Several new forest products, such as sequestered carbon and biomass for renewable energy generation, are being evaluated as it is considered that a multi-product approach is more likely to be profitable than timber production alone. Carbon sequestration may occur both as an adjunct to wood production systems and also where restoration of biodiversity is the primary aim. The development of full markets for these products is dependent on the establishment of national emissions and renewable energy policies and targets, and in the case of liquid biofuels further technological development. Three broad approaches to integrating trees into the dryland farming systems are being assessed viz. (a) belts of trees with farming maintained in the alleys, (b) blocks of trees located on areas of water accumulation or of high recharge, and (c) short phases (3-5 years) of trees alternated with cropping. Both the alley and phase farming systems offer the prospect of producing biofuels from farmland without either using food-grains or displacing farming production. Paradoxically, for systems that are attempting to stabilize landscape hydrology, a major issue for reforestation is water management and this can be manipulated through species selection, tree placement and canopy management. If inappropriate species or planting densities are used, in relation to a site’s water supply, the trees will die. Although this is an aim of the phase farming system, for the other systems, avoidance of annual and periodic drought is a prime consideration.