This study presents data from a large-scale Before After Control Impact (BACI) design field experiment that measured the sediment reduction achieved by remediating large alluvial gullies. The study was carried out on Bonnie Doon Creek on the lower Burdekin River, in Queensland Australia. Prior to remediation, the four large alluvial gully complexes (active area of-17ha) were conservatively estimated to be delivering 5800 +/- 1500 t of fine sediment (<20 mu m) per year (20 year average). The experiment demonstrated that the average remediation effectiveness across 10 different treatments was a 96%-99% reduction in fine sediment yield (or an annualised reduction of-5500t). High resolution lidar DEM of Difference (DoD) derived sediment yields in the unremediated gullies were found to be, on average, 58% lower than yields derived from monitored suspended sediment concentration (SSC) data, albeit with some uncertainty. These data support the notion that even high resolution (0.1m) lidar DoD yields are missing erosion driven by rainfall driven downwearing across all internal gully surfaces that is below the limit of detection (LOD) of the lidar. The results highlight that the greatest uncertainty in predicting the sediment abatement from gully remediation is associated with the determination of the baseline sediment yield of each gully. Future research effort should be focused on improving our understanding of baseline (multi-decadal) sediment yields, and monitored (annual) yields in different types of unremediated gullies. This is dependent on developing a detailed understanding of how these gullies evolve through time, and what the processes are that drive ongoing gully growth. (c) 2024 International Research and Training Center on Erosion and Sedimentation, China Water and Power Press, and China Institute of Water Resources and Hydropower Research. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under the CC BY- NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Avulsion of rivers can be a gradual process that is associated with a metamorphosis of channel pattern or changed channel characteristics. The processes controlling avulsion, and hence anastomosis, often operate too slowly to study by measuring active river systems, and hence well preserved Late Quaternary rivers offer one of the best ways to study the long-term development of avulsive systems. The modern and ancient channels of the Murrumbidgee River provide a classic example of long-lived, semi-static anastomosis, operating on timescales that include stadial and interstadial climate cycles. Over the last glacial cycle, regional avulsions have occurred every similar to 12 ka while maintaining an anastomosing pattern through the slow infill of abandoned channels. The Yanco Creek Palaeochannel System emerged from a period of high discharge linked to snowmelt in the terminal Pleistocene. Here, geomorphological mapping of the Yanco System was conducted together with single-grain, optically stimulated luminescence dating of sediments in the channel belt. Since the main phase of channel construction during the Last Glacial Maximum, the Yanco system has functioned as a flood conduit and minor anabranch of the Murrumbidgee River, with reworking of ancient channel sediments by an underfit stream that is ongoing to the present day. Our new ages of similar to 13-14 ka are interpreted as channel sedimentation during an underfit phase following avulsion. The prevalence of full and partial avulsion in this environment may be complicating palaeohydrological interpretations of ages for channel activity, and reworking has gone unrecognised. We contend that some previous interpretations of the significance of ages for sediments collected from palaeochannels in the Riverine Plain may need revision.

ABSTRACTThe combination of high and often intense rainfall, steep slopes, and a major loss of forest cover in the high standing and tectonically active islands of the Coral Triangle produces high erosion rates with large amounts of riverine sediment that poses a threat to coastal and near shore biodiversity and hence the livelihoods of many people. While Timor-Leste is not as tectonically active as other islands in the region, it has all of the other characteristics that give rise to large sediment loads. Because of the paucity of data, estimates of rates and rate changes of erosion and sediment transport to the coast of Timor-Leste are here derived from surrogates and ‘global’ relationships, in conjunction with locally collected data, some of which is from local people. Reduction of vegetation cover for agriculture and timber harvesting, particularly in the mountains, has increased erosion rates by factors up to 120, mainly by landsliding, with impacts on rivers and at the coast. Revegetation is therefore likely to reduce these impacts, and it should be planned to also enhance the livelihoods of people in the mountains. Sound science-based management of the Coral Triangle is at risk if similar studies are not performed. In this paper the methods that have been successful, semi-successful or unsuccessful have all been included for the benefit of those who wish to pursue this line of research.

Despite over a century of investigations into gullies and gully erosion, the characterization and categorization of gullies and the varied definitions, nomenclatures and terminology used has caused some confusion in understanding and communicating the relationships of gully forms and processes around the world. We firstly review the gully literature and highlight how a lack of consistency in gully definition and characterization prevents unifying theory from being developed within this important field of research, since it is often unclear whether different landscape features being discussed are comparable. We propose that conventionally employed qualitative planform and cross‐sectional characteristics of gullies alone are inadequate to define gully types, yet both these features remain central to most modern gully descriptions. We discuss the need to revise and augment these basic characteristics with clearly defined morphogenetic attributes such as landscape context, soil material characteristics, erosion processes, hydrological integrity, modes of development, and head/side‐wall morphology for an effective, practicable, generic gully classification scheme. Central to a gully classification scheme is the need for a clear definition of what a gully is – and is not – for which geomorphological criteria are proposed to differentiate a ‘gully’ from other ‘incisional land surface forms’. This gully definition hinges largely on the identification of a retreating head scarp and the internal erosion by mass‐movement and other sidewall slope erosion processes, coupled with the transport of the soil materials from the gully void. By defining a gully and synthesizing descriptions of gully ‘types’ from the literature and our own experience, we propose key morphogenetic attributes of gullies necessary to form a framework for a systematic gully classification scheme. An initial, eclectic classification framework is presented as both a summation and a synthesis of the literature review, and as a progenitor to a dynamic generic classification scheme that is proposed in a follow‐up article.

Buildings and monuments constructed from stone provide some of the best‐preserved surface archaeology, but their construction ages can be difficult to determine using radiocarbon techniques. In Australia, stone arrangements are recognised as architectural or symbolic features belonging to Aboriginal societies. The structures are predominantly inorganic with shallow infill, hampering attempts to determine their antiquity. Optically stimulated luminescence (OSL) techniques have the potential to date these features, but their complex geometry requires careful consideration of the background radiation. Here, we present the first ages for Australian Aboriginal stone construction using single‐grain OSL techniques on quartz from stone arrangements in central Australia. Beta and gamma dose rates and the cosmic ray dose have been estimated from mapping the gross geometry of stone and sand courses. The resulting OSL ages are internally consistent and, together with fallout radionuclides 137 Cs and 210 Pb, indicate a minimum age for construction between 1959 and 1981 AD. We demonstrate that single‐grain OSL techniques can be used to determine the age of emplaced sand between stones and, assuming a stable substrate, can be used to date stone building construction as well as building occupation, providing chronologies for sites where organic material for radiocarbon analysis is limited or unavailable.

Modern Homo sapiens engage in substantial ecosystem modification, but it is difficult to detect the origins or early consequences of these behaviors. Archaeological, geochronological, geomorphological, and paleoenvironmental data from northern Malawi document a changing relationship between forager presence, ecosystem organization, and alluvial fan formation in the Late Pleistocene. Dense concentrations of Middle Stone Age artifacts and alluvial fan systems formed after ca. 92 thousand years ago, within a paleoecological context with no analog in the preceding half-million-year record. Archaeological data and principal coordinates analysis indicate that early anthropogenic fire relaxed seasonal constraints on ignitions, influencing vegetation composition and erosion. This operated in tandem with climate-driven changes in precipitation to culminate in an ecological transition to an early, pre-agricultural anthropogenic landscape.

Explanations for the Upper Pleistocene extinction of megafauna from Sahul (Australia and New Guinea) remain unresolved. Extinction hypotheses have advanced climate or human-driven scenarios, in spite of over three quarters of Sahul lacking reliable biogeographic or chronologic data. Here we present new megafauna from north-eastern Australia that suffered extinction sometime after 40,100 (±1700) years ago. Megafauna fossils preserved alongside leaves, seeds, pollen and insects, indicate a sclerophyllous forest with heathy understorey that was home to aquatic and terrestrial carnivorous reptiles and megaherbivores, including the world’s largest kangaroo. Megafauna species diversity is greater compared to southern sites of similar age, which is contrary to expectations if extinctions followed proposed migration routes for people across Sahul. Our results do not support rapid or synchronous human-mediated continental-wide extinction, or the proposed timing of peak extinction events. Instead, megafauna extinctions coincide with regionally staggered spatio-temporal deterioration in hydroclimate coupled with sustained environmental change.

ABSTRACTKiacatoo Man, a large, rugged Aboriginal adult buried in the Lachlan riverine plains of southeastern Australia, was discovered in 2011. Laser‐ablation uranium series analysis on bone yielded a minimum age for the burial of 27.4 ± 0.4 ka (2σ). Single‐grain, optically stimulated luminescence ages on quartz sediment in which the grave had been dug gave a weighted mean age of 26.4 ± 1.5 ka (1σ). Luminescence samples from the grave infill and from sediment beneath the grave exhibit overdispersed dose distributions consistent with bioturbation or other disturbance, which has obscured the burial signal. The overlap between the minimum (U‐series) and maximum (luminescence) ages places the burial between 27.0 and 29.4 ka (2σ). Luminescence ages obtained from the channel belt of between 28 ± 2 and 25 ± 3 ka indicate that fluvial sedimentation was occurring before the Last Glacial Maximum, which is consistent with the broader geomorphic setting. Together, these results are internally and regionally consistent, and indicate that Kiacatoo Man was one of the more ancient individuals so far identified in Australia. His remains are important to our understanding of patterns of biological variation and other processes that have shaped people in the Murray‐Darling Basin through time. Copyright © 2019 John Wiley & Sons, Ltd.

In this paper we outline a worked example of the combined use of genetic data and archaeological evidence. The project focuses on Queensland’s Cape York Peninsula and has two goals. One is to shed new light on the population history of the region. The other is to develop a methodology to facilitate repatriation of the remains of Aboriginal Australians. After providing some background to the project and outlining its main activities, we summarize our key findings to date. Subsequently, we discuss what the project has taught us about the prehistory of Cape York, the potential for DNA research and isotope chemistry to assist research institutions and Aboriginal communities with the repatriation of unaffiliated remains, and the process of conducting combined genetic and archaeological research.

Traditional approaches to flood management have often conflicted with what is now regarded as being best practice for river and catchment management focused on ecological and water quality outcomes. Maximising the extent of riparian vegetation, including in-channel vegetation, has long been recognised as critical for improving aquatic ecosystem complexity and diversity, and for stabilising channels, and thereby reducing erosion. The conventional wisdom amongst floodplain managers has been that optimal flood mitigation is best achieved through having the channel network designed to be as hydraulically efficient as possible, and typically this has meant clearing in-channel vegetation, straightening and armouring channels. However, it has long been appreciated that these types of activities can lead to increased channel instability, thereby threatening assets, and are deleterious to many ecosystem processes within the river channel network. There is also increasing evidence that by not accounting for the geomorphic feedbacks between in-channel vegetation and channel stability or catchment scale flood wave dynamics, that vegetation removal or suppression (intentional or otherwise) can lead to increased downstream flooding and a cascade of unintended consequences. It has also become increasingly apparent that across government there are policies focused on different aspects of river and floodplain management that can sometimes conflict, and when viewed collectively are unnecessarily complicated. In this paper, based on work in the Hunter Valley, we propose a simplified unifying framework for viewing all aspects of river and floodplain management through the lens of the maximisation of in-channel woody vegetation. By maximising in-channel (ideally native) woody vegetation four complementary objectives can be achieved: I. Minimising channel erosion, thereby protecting riparian land and reducing sources of sediment to the downstream waterbodies. II. Maximising sediment deposition and nutrient retention within the channel network, improving downstream water quality and reduced “leakiness” of ecological pathways (e.g. carbon and nitrogen cycling). III. Maximising ecosystem functioning through increased habitat complexity (e.g. increased woody debris; low flow pool habitat availability) IV. Minimising flooding in the lower catchment, through increasing-channel roughness in upstream channels with reduced flood celerity in key tributaries. 2018 Floodplain Management Australia National Conference Gold Coast 2 Riparian Vegetation Management as a Unifying River management Framework. Introduction Catchment, river and flood management are complex policy areas for various levels of government, which often result in a myriad of overlapping and sometimes contradictory policy frameworks. In a recent review of the management of unregulated rivers in the Hunter (Brooks., et al., 2016; Kemp et al., 2017) it became apparent that, despite a daunting array of policy and management frameworks, most management objectives could be practically achieved through a focus on the management of riparian vegetation. We concluded, therefore, that by maximising in-channel (ideally native) woody vegetation four complementary objectives can be achieved: I. Minimising channel erosion, thereby protecting riparian land and reducing sources of sediment to the downstream waterbodies. II. Maximising sediment deposition and nutrient retention within the channel network, improving downstream water quality and reduced “leakiness” of ecological pathways (e.g. carbon and nitrogen cycling). III. Maximising ecosystem functioning through increased habitat complexity (e.g. increased woody debris; low flow pool habitat availability) IV. Minimising flooding in the lower catchment, through increasing-channel roughness in upstream channels with reduced flood celerity in key tributaries. In this paper, we review evidence underpinning these conclusions, along with knowledge gaps, associated with the role that riparian vegetation plays, and could play, as a flood management tool in the Hunter. The evidence underpinning the first three aspects is less controversial than that surrounding the role of vegetation as a flood management tool, and will not be the primary focus of this paper. For a recent review see Kemp et al. (2017). The Role of Riparian Vegetation There is now abundant scientific evidence demonstrating that the maximisation of inchannel and riparian vegetation can help to achieve the first three of the objectives outlined above. In the Hunter, where riparian vegetation regrowth has occurred in some tributaries as much by happenstance as by design, there is anecdotal and some robust scientific evidence of its associated benefits to channel stability, post-flood channel recovery, increased baseflow, the persistence of pools, and improved water quality and ecological function. As graphically demonstrated in Figure 1, the channel of Wollombi Brook was largely devoid of riparian vegetation in the 1960s, and at this time channel and bank erosion was a major problem throughout the Hunter, delivering considerable volumes of sediment to the lowland reaches of the Hunter River and Newcastle Harbour. By the early 2000s, this situation had changed dramatically, as extensive natural vegetation regeneration has very effectively stabilised the channel. As is evident from the images, the large “Pasha Bulker” flood of June 2007 caused very little channel erosion, resulting if anything in sediment deposition within the channel (David Outhet, pers. Comm., 2007). Also apparent from the set of images is the considerable degree of in-channel roughness associated with the mass of in-channel vegetation. In the remainder of this paper we explore the implications of all of this extra roughness on flood behavior. 2018 Floodplain Management Australia National Conference Gold Coast 3 Figure 1. Sequence of photographs taken from the Warkworth bridge over the Wollombi Brook, in 1967, Feb and June 2007, showing the dramatic increase in riparian vegetation over four decades, and the associated improvement in channel stability and in-stream ecosystem diversity. 2018 Floodplain Management Australia National Conference Gold Coast 4 The role of riparian vegetation in flood mitigation Local Context of Flooding in the Hunter The last thirty years have seen an abeyance in catastrophic flooding in the Lower Hunter, partly owing to episodes of drought and protection offered by reservoirs upstream. Despite this, the threat of flooding remains high, particularly for population centres at Maitland and surrounding agricultural land (Figure 2). Glenbawn Dam and Lake St Clair (Figure 2), which together regulate some 1600 km or 7% of the total Hunter catchment, offer modest protection again flooding in the lower Hunter, but large areas of the catchment remain unregulated. In the lower Hunter, flood control works including levees were constructed after the destructive flood in 1955, which in Maitland alone killed 24 people and inundated 2180 houses (WMA Water, 2015). However, the levees do not protect against floods on the scale of 1955 or even lower floods, and dams at spillways upstream of Maitland were overtopped in 1991, 1997 and 2007. Singleton is located almost entirely on the floodplain of the lower Hunter River. Its flood protection levees are designed to accommodate floods at the 1955 flood level (14.31 m) (Paterson Consultants, 2012), but residential buildings are required to be 0.48 m higher to allow for climate change effects on flood levels. It is noteworthy that the flood height in June 2007 was 13.35 m, which was the second highest recorded stage since gauge recordings commenced in 1913. Model results of flooding in Wollombi Brook (BMT WBM, 2016) noted inundation of the township of Broke by floods with average return intervals of more than 100 yr, with considerable flooding extent downstream of Brickmans Bridge (stn. 210135, Figure 2). However, the model was sensitive to channel roughness (Manning’s n parameter), particularly in areas upstream of Brickmans Bridge. This represents the actual sensitivity of floods in the Wollombi, among other things, to changes in floodplain and in-channel vegetation, and sediment transport conditions that result in a change in channel dimensions. An increase of 1.0 m was recommended to Flood Planning Levels and Flood Planning Areas in this part of Wollombi Brook (BMT WMB, 2016). Figure 2 Hunter catchment showing places discussed in text. 2018 Floodplain Management Australia National Conference Gold Coast 5 Flood Attenuation Attenuation is the natural dissipation of a flood wave in a downstream direction. It produces a downstream decrease in peak flow velocity and flood stage, and an increase in flood duration. The degree of attenuation increases with drainage area, sinuosity, and overbank flooding, which increases the storage of flood waters and slows the celerity or speed of the flood wave. Celerity is defined as the distance travelled divided by the peak to peak travel time with units of velocity in m s. Riparian vegetation, particularly forest, that is established both within the channel and on the floodplain is a substantial component of channel roughness (Acrement and Schneider, 1989) and increases flood attenuation. Attenuation may be counterbalanced to some extent with tributary inflows downstream. In the Hunter, a historical reduction in the attenuation of flood peaks has resulted from historical attempts to reduce flow resistance through channel clearing and desnagging, channel straightening, and the construction of levees. These interventions combined with indirect feedback effects and land use change resulted in bank and bed erosion and the creation of enlarged, incised channels in the Upper Hunter, which isolated the channel from its floodplain except during severe floods. In the Lower Hunter, sand liberated by channel erosion upstream was deposited in

Sediment runoff has been cited as a major contributor to the declining health of the Great Barrier Reef (GBR), however, climate and land use drivers have not been jointly evaluated. This study used alluvial archives from fluvial benches in two tributaries of the Upper Burdekin catchment together with the best available land use history and climate proxy records to provide insights into the timing of depositional events in this region over the past 500 years. This study suggests that mining and the increased runoff variability in the latter half of the nineteenth century are the likely sources of the original excess sediment that was used to build the bench features in these catchments. Grazing also contributed to increased bench sedimentation prior to 1900, however, the contribution of grazing was likely more significant in the second half of the 20th century, and continues to be a dominant land use contributor today.

This paper reports the results of jet tester experiments on soil samples of uniform properties which allow quantitative application of the new theory proposed in part 1 of these publications. This theory explores the possibly that a more adequate indicator of soil erodibility may be obtained by using the mass (and so volume) of soil eroded by the jet and the depth of scour penetration, rather than by using penetration depth alone, as assumed in the commonly-used data interpretation method. It is shown that scour geometry can be well described using a generalized form of the Gaussian function, defined by its standard deviation and maximum depth. Using a published expression for jet kinetic energy flux, the new theory divides this flux into that used to erode soil, and the remainder which is dissipated in a variety of ways. Jet experiments on a specially-prepared uniform soil sample are reported which provide the key to determining the spatial variability in the profile resistance to erosion offered by field soils. This resistance is expressed in the work required to erode unit mass of soil, denoted as J (in J/kg). The paper also gives results obtained on the profile variation in J for jet tests carried out at riverine sites on the upper Brisbane River, Queensland, Australia. As expected in most natural soil profiles, the results show an increase in J with depth in the profile. The soil resistance (J) is compared to the traditional interpretation of soil erodibility, (k(d)). The graphical comparison of these two indicators illustrates the inverse type of relationship between them which is expected from their respective definitions, but this relationship is associated with significant scatter. Possible reasons for this scatter are given, together with comments on jet tester experience in a wide variety of soil types. Copyright (C) 2017 John Wiley & Sons, Ltd.

Climatic forcing of fluvial systems has been a pre-occupation of geomorphological studies in Australia since the 1940s. In the Riverine Plain, southeastern Australia, the stable tectonic setting and absence of glaciation have combined to produce sediment loads that are amongst the lowest in the world. Surficial sediments and landforms exceed 140,000 yr in age, and geomorphological change recorded in the fluvial, fluvio-lacustrine and aeolian features have provided a well-studied record of Quaternary environmental change over the last glacial cycle. The region includes the Willandra Lakes, whose distinctive lunette lakes preserve a history of water-level variations and ecological change that is the cornerstone of Australian Quaternary chronostratigraphy. The lunette sediments also contain an ancient record of human occupation that includes the earliest human fossils yet found on the Australian continent. To date, the lake-level and palaeochannel records in the Lachlan-Willandra system have not been fully integrated, making it difficult to establish the regional significance of hydrological change. Here, we compare the Willandra Lakes environmental record with the morphology and location of fluvial systems in the lower Lachlan. An ancient channel belt of the Lachlan, Willandra Creek, acted as the main feeder channel to Willandra Lakes before channel avulsion caused the lakes to dry out in the late Pleistocene. Electromagnetic surveys, geomorphological and sedimentary evidence are used to reconstruct the evolution of the first new channel belt following the avulsion. Single grain optical dating of floodplain sediments indicates that sedimentation in the new Middle Billabong Palaeochannel had commenced before 18.4 +/- 1.1 ka. A second avulsion shifted its upper reaches to the location of the present Lachlan River by 16.2 +/- 0.9 ka. The timing of these events is consistent with palaeohydrological records reconstructed from Willandra Lakes and with the record of palaeochannels on the Lachlan River upstream. Willandra Lakes shows high inflows during the Last Glacial Maximum (similar to 22 ka), but their subsequent drying between 20.5 ka and 19 ka was caused by river avulsion rather than regional aridity. This case study highlights the benefits of combining fluvial with lacustrine archives to build complementary records of hydrological change in lowland riverine plains. (C) 2017 Elsevier Ltd. All rights reserved.

The spectral signature of individual quartz grains were measured using a high-sensitivity thermoluminescence imaging system based on a charge-coupled device (CCD) camera and custom optics. Luminescence emission behaviour was characterised for individual quartz grains (180–212 μm), with single grains shown to emit luminescence strongly across multiple signal bands. The spectral profiles of three quartz samples from contrasting geological contexts were then compared, with clear distinction in the spectral signatures of individual grains originating from single deposits within each provenance.

Along low gradient rivers in northern Australia, there is widespread gully erosion into unconfined alluvial deposits of active and inactive floodplains. On the Mitchell River fluvial megafan in northern Queensland, river incision and fan-head trenching into Pleistocene and Holocene megafan units with sodic soils created the potential energy for a secondary cycle of erosion. In this study, rates of alluvial gully erosion into incipiently-unstable channel banks and/or pre-existing floodplain features were quantified to assess the influence of land use change following European settlement. Alluvial gully scarp retreat rates were quantified at 18 sites across the megafan using recent GPS surveys and historic air photos, demonstrating rapid increases in gully area of 1.2 to 10 times their 1949 values. Extrapolation of gully area growth trends backward in time suggested that the current widespread phase of gullying initiated between 1880 and 1950, which is post-European settlement. This is supported by young optically stimulated luminescence (OSL) dates of gully inset-floodplain deposits, LiDAR terrain analysis, historic explorer accounts of earlier gully types, and archival records of cattle numbers and land management. It is deduced that intense cattle grazing and associated disturbance concentrated in the riparian zones during the dry season promoted gully erosion in the wet season along steep banks, adjacent floodplain hollows and precursor gullies. This is a result of reduced native grass cover, increased physical disturbance of soils, and the concentration of water runoff along cattle tracks, in addition to fire regime modifications, episodic drought, and the establishment of exotic weed and grass species. Geomorphic processes operating over geologic time across the fluvial megafan predisposed the landscape to being pushed by land used change across an intrinsically close geomorphic threshold towards instability. The evolution of these alluvial gullies is discussed in terms of their initiation, development, future growth, and stabilisation, and the numerous natural and anthropogenic factors influencing their erosion.

ABSTRACTThis paper reconstructs past flooding from a range of settings in Lockyer Creek, a key tributary of the mid‐Brisbane River, which experienced extreme flood events in AD 2011 and AD 2013. Optically stimulated luminescence samples (n = 110) were collected from alluvial material preserved in within‐channel benches and floodplains. Age distributions from material in the bedrock reaches confirm an event ∼ 300 years ago which stripped the valley alluvium to bedrock. In the unconfined reaches floodplain deposits indicate lateral stability over the past 6000 years. Marked differences in the inundation patterns of the AD 2011 event highlighted changes in downstream channel geometry. The age distribution of alluvium in reaches not inundated during AD 2011 was older, ∼12 000 years, with no preserved evidence of deposition during the past 1000 years. A relatively continuous record of floodplain deposition in reaches which were inundated in AD 2011 identifies a major peak in flood activity also around 300 years ago (∼AD 1730) with five additional peaks occurring at approximately AD 1962, AD 1897, AD 1300, AD 550 and 5400 BC. The main climatic driver of changes in flooding over this timescale is oscillations in El Niño Southern Oscillation and although proxy records are scarce for this region, some correlations with high‐resolution records of rainfall variability are apparent.

This study describes the use of linearly modulated optically stimulated luminescence (LM‐OSL) to distinguish surface‐soil derived sediments from those derived from channel bank erosion. LM‐OSL signals from quartz extracted from 15 surface‐soil and five channel bank samples were analysed and compared to signals from samples collected from two downstream river sites. Discriminant analysis showed that the detrapping probabilities of fast, first slow and second slow components of the LM‐OSL signal can be used to differentiate between the samples collected from the channel bank and surface‐soil sources. We show that for each of these source end members these components are all normally distributed. These distributions are then used to estimate the relative contribution of surface‐soil derived and channel bank derived sediment to the river bed sediments. The results indicate that channel bank derived sediments dominate the sediment sources at both sites, with 90.1 ± 3% and 91.9 ± 1.9% contributions. These results are in agreement with a previous study which used measurements of 137Cs and 210Pbex fallout radionuclides to estimate the relative contribution from these two sources. This result shows that LM‐OSL may be a useful method, at least in the studied catchment, to estimate the relative contribution of surface soil and channel erosion to river sediments. However, further research in different settings is required to test the difference of OSL signals in distinguishing these sediment sources. And if generally acceptable, this technique may provide an alternative to the use of fallout radionuclides for source tracing. Copyright © 2015 John Wiley & Sons, Ltd.