THE CRESSBROOK CREEK ALLUVIAL AQUIFER SYSTEM , SOUTHEAST QUEENSLAND : HYDROCHEMISTRY AND ISOTOPES TO DETERMINE HYDROLOGICAL PROCESSES AND RESPONSE TO FLOODS Adam C. King Bachelor of Science Honours (University of Tasmania) - 2005 Thesis submitted in accordance with the regulations for the Degree of Doctor of Philosophy School of Earth, Environmental and Biological Sciences Science and Engineering Faculty Queensland University of Technology September 2014 Keywords Alluvial aquifers, aquifer connectivity, Australia, barometric pressure, Cressbrook Creek, conceptual hydrogeological model, groundwater recharge, hydrochemistry, inter-aquifer connectivity, isotopic tracers, multivariate statistical analysis, southeast Queensland, groundwater/surface water interactions, three-dimensional (3D) geological modelling, tritium, water-table fluctuation method. i Abstract Alluvial aquifer systems are major components of many catchments, are vital for the maintenance of healthy stream ecosystems, and provide an essential water source for multiple uses. Sustainable management of these alluvial aquifers is critical, but to enable this, a solid understanding of hydrological processes is required. The alluvial system of Cressbrook Creek in subtropical southeast Queensland is a good example of an aquifer system that is an essential water supply for intensive irrigation. However, this water supply is subject to an increasing number of competing pressures, including the construction of a dam in the catchment headwaters, which has contributed to reduced creek flow and lower groundwater levels in the alluvium. Continued groundwater abstraction for crop irrigation further decreased water storages during an extended drought, before groundwater storage was replenished in December 2010 and January 2011 as a result of heavy rain and associated flooding. The frequency and severity of these climatic extremes are predicted to increase, and such events will continue to impact on alluvial groundwater resources. These conditions confirm that an understanding of the hydrological processes associated with such extreme climatic events is vital for the continued groundwater resource management of this alluvial aquifer system and for groundwater resources in similar settings. Therefore, it is critical that methods are developed for the analysis of these climatically induced hydrological processes, particularly in relation to episodic recharge and interactions between surface waters and groundwater. The hydrological processes described in this thesis are potentially analogous to those that occur in comparable locations, and the methods developed here may be applied to other catchments. This thesis is structured around three linked papers, which provide a holistic assessment of hydrological processes, particularly in relation to surface and groundwater interactions, and the identification of various recharge processes, including diffuse recharge, channel leakage and bedrock seepage. An important aspect of the study is the response of the alluvial aquifer to climatic variations, including the protracted drought from the late 1990 until 2009, and the subsequent period of heavy rain and flooding in 2010 to 2011. iii Paper 1 outlines the catchment character and identifies and explains hydrochemical changes to groundwater composition that result from the observed climatic variations. An extensive geological dataset is integrated to produce a three- dimensional (3D) geological model of the alluvium and the broader catchment, which provides a framework enabling assessment of geological controls on hydrological processes. This assessment is coupled with a multivariate statistical analysis (MSA) of time series hydrochemical data, which are used to characterise hydrochemical assemblages. Hydrochemical processes were then identified and described using the geological framework. Paper 2 assesses the response of aquifers to flooding, together with the inter-aquifer connectivity between the alluvium and the underlying bedrock. A suite of isotopic tracers is employed ( δ2H, δ18 O, 87 Sr/ 86 Sr, tritium and 14 C) in conjunction with a comprehensive assessment of hydrochemical patterns to understand the link between alluvial groundwaters and the aquifer materials. In paper 3, multiple methods are applied to quantify recharge rates in the alluvial aquifer, including the water budget method, the chloride mass balance method, tritium, and the water-table fluctuation (WTF) method. The alluvium is characterised by a fining-upwards sequence, which is typically composed of basal sands and gravels, overlain by silts and clays. The thickness of the upper, low permeability layer increases with distance downstream, whereas the thickness of the basal high permeability layer decreases down-gradient. These variations are related to depositional processes, but also indicate that the lower parts of the catchment may receive less diffuse recharge (vertical infiltration from rainfall or irrigation returns) than the upper parts. Alluvial groundwaters of Cressbrook Creek are generally fresh to brackish with no dominant cation and with low SO 4 concentrations. The alluvial groundwaters generally evolve chemically with distance downstream and with distance from the creek. This evolution is marked by an increase in salinity, longer groundwater 2 18 residence times and higher Cl/HCO 3 ratios. Stable isotopes (δ H and δ O) and Cl concentrations confirm that these more evolved groundwaters have also been subjected to higher degrees of evapotranspiration, and the elevated Ca and Mg concentrations result from the dissolution of silicate minerals. During the period of this study, baseflow for Cressbrook Creek was typically generated within the upper parts of the catchment, whereas the drainage system was predominately losing in the lower reaches. However, hydrograph data covering iv longer periods show that this relationship is variable, and highly dependent on antecedent rainfall. In the wider, lower parts of the catchment, channel leakage is a key process for the maintenance of groundwater quality and storages. The hydrological effects of this high recharge rate during floods and periods of high stream-flow are particularly important near the confluence of Cressbrook Creek and the larger Brisbane River. In this downstream area, diffuse surface recharge is limited by the relatively thick, low permeability unsaturated zone. In comparison, infiltration rates are relatively high below the streambed, where the unsaturated zone has been incised by the creek and the high permeability basal sediments of the alluvium are in hydraulic connection with the streambed. Bedrock discharge to the alluvium was identified from isotopes and hydrochemical data at several locations in the lower part of the catchment where recharge rates are relatively low. In these areas, bedrock seepage into the alluvium is enabled by lower groundwater levels in the alluvium. This lower water-table is due to the combined effects of reduced surface water flow following upstream dam construction, the extended dry period and the continued abstraction of groundwater by irrigators. The lower parts of the catchment are particularly vulnerable to changes in surface water flows, as channel leakage is a major source of recharge in this area. Estimates of total recharge for the alluvial system of Cressbrook Creek were variable, depending on the method of investigation. The catchment-wide recharge estimate from the water budget method was approximately 41-56 mm/year (corresponding to 5-7% of average annual rainfall), representing the hydrological conditions that occurred during periods of average rainfall. In contrast, recharge estimates from tritium and the WTF method generated a catchment-wide recharge rate of approximately 86 mm/year (corresponding to 10% of average annual rainfall). This result demonstrated the effect that variations in rainfall can have on recharge estimates that are obtained using chemical tracers (i.e. tritium) and the WTF method. There are significant spatial variations in recharge mechanisms, soil characteristics and recharge rates. In the upper parts of the catchment, alluvial groundwater is predominately recharged by diffuse infiltration of rainfall, and recharge rates vary from approximately 130-250 mm/year. In the lower parts of the catchment, recharge rates vary depending on proximity to the creek. For wells located close to the creek, recharge typically ranged from 80-190 mm/year. For wells located distal to the creek, recharge rates range from 4-90 mm/year, but they are v generally <30 mm/year, highlighting the importance of channel leakage for the replenishment of the alluvial aquifer. Significant outcomes from this study include the development and refinement of methodologies relating to the assessment of hydrogeological processes. The first paper demonstrates that a multivariate statistical analysis of time series hydrochemical data is greatly enhanced by the integration of a three-dimensional geological model, as this enables a better hydrogeological conceptualisation. It also underscores the influence of episodic climatic events and catchment geomorphological characteristics on groundwater recharge and chemistry. The second paper highlights the importance of a comprehensive hydrochemical assessment for the interpretation of results from multiple isotope techniques. In particular, this paper also demonstrates the value of time series rainfall stable isotope data for the identification of hydrological processes that