Joshua Larsen

Materials Contributed through SERC-hosted Projects

Other Contributions (2)

Climate change in the dead heart of Australia part of Vignettes:Vignette Collection
Despite the absence of large-scale glaciation, the Australian continent has experienced substantial environmental change throughout the Quaternary period. This is especially pronounced in central Australia, where one seventh of the continent is drained internally to the depocentre, and lowest point in Australia, Lake Eyre (Figure 1). Research has shown that at one time, large sandy braided and meandering rivers carried water through dunefields to a large freshwater lake system. Today, the rivers are hostage to the dunefield, and floodwaters might only reach Lake Eyre once every ten years or so. In order to understand the development of this arid desert landscape, and how changing climates have affected it, we need to find out when rivers and dunes were active, and how they are related to each other. Large volumes of fluvial and aeolian quartz sand have been deposited and reworked within central Australia throughout the Quaternary. This is an ideal material for luminescence dating techniques, which allow us to determine the length of time these grains have been buried. If we combine this dating method with detailed stratigraphy of the rivers and dunes, a picture of changing wet and dry environments begins to emerge. This approach has been taken in the Strzelecki Desert, where Cooper and Strzelecki Creeks run through an expansive linear dunefield. Looking from above, be it from a satellite image or from an aircraft, it is obvious that contemporary river patterns are affected by the position of the linear dunes (Figures 1 & 2). It is also apparent that the linear dunes are grouped into discrete clusters, with a flat muddy floodplain in between (Figures 2 & 3). Interestingly, if we dig beneath the linear dunes and floodplain, we find abundant fluvial sands, which stand in stark contrast to the mud dominated fluvial system of today (Figure 4). Why then, do we have so much fluvial sand beneath muddy floodplains and linear dunes? Clearly Cooper and Strzelecki Creeks were much more active in the past, and the substantially higher discharges allowed transportation of coarse sands in braided and meandering streams. Most of this fluvial sand was transported during previous interglacials (i.e. warm stages), albeit with declining amounts with each successive interglacial. However these rivers also maintained active channels throughout glacial periods (i.e. cold stages) as well, suggesting that although rainfall and discharge had declined, central Australia was not completely dry during glacial periods. It is also clear that fluvial transportation of sand was the dominant source of sand for the linear dunefields. The rivers were probably seasonal, and this allowed sand to be blown from the channel and into 'source bordering' or transverse dune features running parallel to the channel. These large source bordering dunes were then reworked into the linear dune clusters that cover the Strzelecki Desert. The ages of the linear dunes do not seem to reveal any significant growth during glacial or interglacial periods, in fact, they have a very steady, linear growth rate for at least the last 100,000 years. This suggests that the development of linear dunes in central Australia was not directly driven by changes in global climate, but rather by local and regional changes in sediment supply. What does all this tell us about climate change and dead heart of Australia? Large rivers once flowed across the Strzelecki Desert delivering sand, which could then be used to build the dunes. Discharge has steadily declined over the Quaternary, such that today, the rivers are a shadow of their former selves. The linear dunes have probably been slowly accreting throughout this time, building first from fluvial sands, and then from the re-working of smaller dunes. Central Australia has, therefore, been able to build its dunefields for at least the last 100,000 years, and probably much longer. It was also much wetter, especially during interglacials, and the rivers would have supported a diverse riparian habitat and fed large freshwater lakes. However, during the current interglacial, central Australia is just as arid, if not more arid, than during any previous glacial period. Why is this so? No one is really sure. Some researchers believe human arrival dramatically altered the vegetation which in turn altered the climate, whereas others believe the climate has changed all on its own as the continent has slowly drifted northwards. Whichever the case, it seems the dead heart of Australia has much more to reveal about climate change.

Billabongs (waterholes), unique geomorphology and hydrology in action in arid Australia part of Vignettes:Vignette Collection
Much on the centre of Australia is an arid or semi-arid landscape, and the low average annual rainfalls make it difficult for the rivers to run and the lakes to fill. Some of the rivers that drain the centre of the continent however, have their headwaters within the reach of the summer monsoon, which has the ability to deliver substantial rainfall to these rivers, and sometimes to the centre of the continent. These large rain and flooding events are also critical to the ecology of the arid zone, which has adapted to the 'boom and bust' nature of resource availability. The summer rains that feed the headwaters of these rivers are subject to large scale atmospheric 'forcings', such as the El Nino Southern Oscillation and the Indian Ocean Dipole, which results in substantial variation in monsoonal rainfall. As a result, the large rivers which drain into the centre of the continent, such as Cooper Creek, have some of the highest discharge variability in the world. This in turn has led to the development of a muddy multiple channel-floodplain (anabranching) river system, which typically have very low slopes over very large distances (Figure 1 & 2). A unique feature of this anabranching system, are the series of billabongs, or waterholes, which are enlarged channel segments along the main course of the river (Figure 3). Billabongs typically occur at the confluence of two smaller channels, but are unique because they do not reach their maximum width/depth ratio until much further downstream, and then a short distance later, the channel abruptly terminates. Channel termination typically occurs in rivers when a clear slope advantage cannot be found, or alternatively due to a downstream decline in discharge. However, many billabongs initiate and terminate along the course of Cooper Creek, which suggests the overall slope of the system is not declining, and that discharge remains sufficient to maintain and indeed initiate new billabongs along its length, all without any significant tributary contribution. How then do Billabongs form? The answer it seems lies in the unique hydrology of the system. We know that over ~400km of its length (between Windorah and Nappa Merrie), Cooper Creek experiences, on average, transmission losses (e.g. evaporation, floodplain routing, groundwater seepage) in the order of 75%. This is perhaps not surprising given the distance or the arid climatic setting. However, an analysis of the long-term longitudinal chemical trends reveals that the concentration of dissolved salts does not increase with increasing chloride concentrations. If evaporation were responsible for flow transmission losses over this reach, then we should find the opposite, chloride should increase with increasing dissolved salt concentrations. If evaporation does not play a major role, where then does all this water go? Let us consider the sedimentology and stratigraphy of the system. Cooper Creek transports mud as both suspended and bed load (in the form of mud aggregates), and has done so for most of the Holocene and possible late Pleistocene. These deposits have formed a thick mud floodplain, beneath which lies older, almost homogenous coarse to medium quartz sands deposited when Cooper Creek sustained much higher discharges during the Quaternary (Nanson et al., 2008). If a channel has sufficient stream power, it can scour the mud base of the channel, and expose the clean sands and allowing discharge of stream water into the underlying water table. Once the floodwaters cease, the mud carried in suspension settles to the base of the channel once more, thus sealing the channel and isolating it from the water table. This mechanism allows for local reductions in channel discharge which may explain the abrupt termination of some billabongs. But is this plausible? Groundwater investigations reveal the water table is ~10-12m beneath the floodplain, and ~2-3m beneath the maximum bankfull depth of the channels, thus water retained in the billabongs following flow cessation is 'perched' above the local aquifer. Given the thickness of the floodplain mud, it is unlikely that floodwaters can recharge through the floodplain, even if large soil desiccation cracks exist. This is further supported by the salinity of the regional floodplain groundwater, which in places approaches that of seawater, indicating much of this groundwater has not been recharged with freshwater in quite some time. If however, we look at groundwater below and adjacent to the billabongs where their depth is greatest, we find remarkably fresh groundwater with very similar chemistry to the surface waters (Figure 4). This confirms that during large flood events, the muddy channel base can be scoured, and freshwater recharged into the shallow groundwater system, which in turn develops freshwater lenses that sit above and adjacent to much saltier, regional groundwater. These processes may be important in other large alluvial rivers with highly variably discharge, and has implications for the water resources of the local populations.