Hypotheses to explain the source of the 1011 tons of salt in groundwaters of the Murray Basin, south-eastern Australia, are evaluated; these are (a) mixing with original sea water, (b) dissolution of salt deposits, (c) weathering of aquifer minerals and (d) acquisition of solutes via rainfall. The total salinity and chemistry of many groundwater samples are similar to sea-water composition. However, their stable isotopic compositions (δ18O= –6.5 ‰; δ2H = –35) are typical of mean winter rainfall, indicating that all the original sea water has been flushed out of the aquifer. Br/Cl mass ratios are approximately the same as sea water (3.57 x 10-3) indicating that NaCl evaporites (which have Br/Cl<10-4) are not a significant contributor to Cl in the groundwater. Similarly, very low abundances of Cl in aquifer minerals preclude rock weathering as a significant source of Cl. About 1.5 million tons of new salt is deposited in the Murray–Darling Basin each year by rainfall.The groundwater chemistry has evolved by a combination of atmospheric fallout of marine and continentally derived solutes and removal of water by evapo-transpiration over tens of thousands of years of relative aridity. Carbonate dissolution/precipitation, cation exchange and reconstitution of secondary clay minerals in the aquifers results in a groundwater chemistry that retains a ‘sea-water-like’ character.

s u m m a r yAccurate quantification of evaporative losses to the atmosphere from surface water bodies is essential for calibration and validation of hydrological models, particularly in remote arid and semi-arid regions, where intermittent rivers are generally minimally gauged. Analyses of the stable hydrogen and oxygen isotope composition of water can be used to estimate evaporative losses from individual pools in such regions in the absence of instrumental data but calculations can be complex, especially in highly variable systems. In this study, we reviewed and combined the most recent equations required for estimation of evaporative losses based on the revised Craig-Gordon model. The updated procedure is presented stepby-step, increasing ease of replication of all calculations. The main constraints and sources of uncertainties in the model were also evaluated. Based on this procedure we have designed a new software, Hydrocalculator, that allows quick and robust estimation of evaporative losses based on isotopic composition of water. The software was validated against measures of field pan evaporation under arid conditions in northwest Australia as well as published data from other regions. We found that the major factor contributing to the overall uncertainty in evaporative loss calculations using this method is uncertainty in estimation of the isotope composition of ambient air moisture.

No abstract

[1] Dryland rivers function as strongly linked ecologic-hydrologic systems, including both extended periods of drought and episodic flooding events. However, few studies have combined hydrologic and biogeochemical measurements to better understand the ecology of pools within dryland rivers. We used d 2 H and d 18 O values of pool water, rainfall, and groundwater combined with pool water measurements of C, N, and P and dissolved organic matter (DOM) fluorescence characteristics to determine (1) the concentration and chemical composition of DOM and (2) the origin of surface water in 16 pools of a dryland river in northern Western Australia. Parallel factor analysis of excitation-emission matrices showed that humic-like components derived mainly from terrestrial plant material dominated total DOM fluorescence for all pools. Evaporation models using d H and d18 O showed a variety of pool hydrologic regimes, including pools with moderate to high evaporative water loss that were largely isolated from shallow alluvium water inputs and pools with consistent alluvium water throughflow and low evaporation. Concentrations of C, N, and P as well as total DOM fluorescence were generally greater in pools with high evaporative loss and lower in pools with alluvium water inputs. Pool d H and d18 O values were also significantly correlated with DOM fluorescence characteristics and C, N, and P concentrations, providing quantitative evidence of the hydrologic influence on DOM biogeochemistry. Taken together, our findings suggest that individual pools function as distinct ecosystems within the riverine environment.