Since 2011, Structure-from-Motion Multi-View Stereo Photogrammetry (SfM or SfM-MVS) has gone from an overlooked computer vision technique to an emerging methodology for collecting low-cost, high spatial resolution three-dimensional data for topographic or surface modeling in many academic fields. This dissertation examines the applications of SfM to the field of fluvial geomorphology. My research objectives for this dissertation were to determine the error and uncertainty that are inherent in SfM datasets, the use of SfM to map and monitor geomorphic change in a small river restoration project, and the use of SfM to map and extract data to examine multi-scale geomorphic patterns for 32 kilometers of the Middle Fork John Day River. SfM provides extremely consistent results, although there are systematic errors that result from certain survey patterns that need to be accounted for in future applications. Monitoring change on small restoration stream channels with SfM gave a more complete spatial perspective than traditional cross sections on small-scale geomorphic change. Helicopter-based SfM was an excellent platform for low-cost, large scale fluvial remote sensing, and the data extracted from the imagery provided multi-scalar perspectives of downstream patterns of channel morphology. This dissertation makes many recommendations for better and more efficient SfM surveys at all of the spatial scales surveyed. By implementing the improvements laid out here and by other authors, SfM will be a powerful tool that will make 3D data collection more accessible to the wider geomorphic community.

O estudo das formas do relevo terrestre teve um avanço expressivo com a utilização das modernas ferramentas de sensoriamento remoto. Na região Amazônica, devido à vasta cobertura de nuvens, bruma e fumaça, quase que o ano inteiro, de certa forma, ocasiona dificuldade para levantamento com sensores ópticos. As melhores imagens para a região, portanto, são as imagens provenientes de radar, pois esse sistema tem a vantagem de adquirir imagens, independente, da presença de nuvens ou fumaça. Desta forma, este trabalho teve por objetivo principal o uso das imagens do Radar de Abertura Sintética (SAR) R99B para o mapeamento geomorfológico ao longo do furo do Ariaú no município de Iranduba. A escolha desta área se justifica por apresentar características naturais e socioeconômicas bem peculiares. Na questão natural, Iranduba está encravado entre os dois maiores caudal da região: rio Negro e Solimões/Amazonas e composto de um ambiente de terra firme e várzea. Na questão socioeconômica, é nesse município que se encontra, também, o maior polo oleiro do estado do Amazonas. A metodologia empregada neste trabalho foi baseada na adaptação dos métodos propostos por Lima (1995), composta de três fases: identificação, análise e interpretação. O método proposto permitiu a interpretação das feições geomorfológicas baseadas na textura das imagens SAR/R99B com múltiplas polarizações para as terras baixas com altitude não ultrapassando 100 m. A fim de buscar entender a dinâmica fluvial da área do Ariaú foi realizado trabalho de campo e levantamento pedológico. Como resultado desta pesquisa, obteve-se mapa da drenagem, mapa geomorfológico das feições fluviais e os principais nutrientes do solo na calha do Ariaú. As imagens do SAR mostraram-se eficientes para as diferentes feições geomorfológicas para o mapeamento de média escala, no entanto, a falta de imagens nos diversos períodos hidrológicos, seca, enchente, cheia e vazante, de certa forma, compromete um melhor entendimento do ambiente. Este trabalho, portanto, coloca-se como uma perspectiva para novas interpretações geomorfológicas dos ambientes fluviais nas áreas baixas da Amazônia. / The study of the forms of terrestrial landscape had a significant progress with the use of modern remote sensing tools. In the Amazon region, due to extensive cloud cover, haze and smoke, almost the entire year, somehow, causes difficulty imaging with optical sensors. The best images for the region, are the pictures from Radar because this system has the advantage of acquiring images, independent of the presence of clouds or smoke. Thus, this work aimed to use pictures of the Synthetic Aperture Radar (SAR) R99B for geomorphological mapping along the Furo do Ariaú in the municipality of Iranduba. The choice of this area is justified due to its peculiar natural and socioeconomic characteristics. In the natural matter, Iranduba is wedged between the two major flow region: Rio Negro and Solimões /Amazonas and composed of an environment of upland and floodplain. In the socioeconomic issue, it is in this city that lays the largest polo potter in the state of Amazonas. The methodology used in this study was based on the adaptation of the methods proposed by Lima (1995), composed of three phases: identification, analysis and interpretation. The proposed method allowed the interpretation of geomorphological features based on texture of SAR/R99B images with multiple polarizations to the lowlands to an altitude not exceeding 100m. In order to try to understand the dynamics of the Ariaú river area fieldwork and soil survey were conducted. As a result of this research a map was obtained, including the drainage map, geomorphological map of the river and the main nutrients in the soil trough Ariaú. The SAR images were effective for different geomorphological mapping to medium scale, however, the lack of pictures in different hydrological periods, drought, flooding and receding somewhat, compromises a better understanding of the environment. This work, therefore, arises as a prospect for new interpretations of fluvial geomorphic low-lying areas of the Amazon.

Since the damming and channelization of the lower Long Tom River in the 1940s and 1950s, the quality and quantity of habitat for coastal cutthroat trout and spring Chinook salmon in the watershed has dramatically diminished. In order to better understand the potential for stream restoration, this study uses 2D hydraulic modeling to determine the impact of reconnecting historical meander bends to the main stem of the lower Long Tom River on localized flooding, sediment erosion and deposition, and salmonid physical habitat. These models compare the current conditions to two restoration scenarios that allow for fish passage given 1, 2, and 5-year flood events at two study sites. This study reveals important variations in the impact of restoration between the study sites and the reconnection methods. It also suggests that there is the potential for a large increase in the area of accessible habitat with stream restoration.

The archaeological record varies with the fluvial style and sedimentation of a river, thus controlling the location, preservation, and recognition of archaeological sites. This project identifies archaeological site patterning and preservation along a fluvial system in Iowa that has been relatively stable since the last glaciation. The Upper Des Moines River in Iowa formed approximately 12,500 yr. cal BP as an englacial channel routing glacial meltwater south along the Des Moines Lobe. During the last glacial retreat, the channel incised forming what is now called the High Terrace (TH). Early post-glacial occupations in Iowa would have had access to this the high terrace and uplands. Thus, the Paleoindian and early to middle Archaic site distributions along the Upper Des Moines River are found in upland and upper terrace locations. The intermediate terrace (TI) formed between 4000 and 1000 yr. cal BP and represents the elevation of the river flood plain during this period. Late Archaic people had access to this surface as well as the TH and uplands. By 1000 yr. cal BP, the Upper Des Moines River had eroded down to bedrock and established a floodplain at the level of the lower terrace. Subsequent erosion produced the modern channel configuration with an established series of dated terraces (High Terrace (TH), Intermediate Terrace (TI), and Low Terrace (TL)). The Woodland period peoples would have had access to the current/modern fluvial landscape. To assess site locations strategies of populations that inhabited the Upper Des Moines River valley in the Holocene, 721 archaeological sites were examined and classified using site reports and artifacts as Paleoindian, Archaic, Woodland, Historic and Unknown. The sites were then mapped and associated with geomorphic features in the valley. As expected, Paleoindian sites and Early to Middle Archaic sites were all located on the upper terrace and uplands because the intermediate and lower terraces had not yet formed. Late Archaic people had access to the floodplain that formed intermediate terrace as well as the uplands and upper terrace. Woodland period sites occurred on high (upland and TH), intermediate (TI), and low (TL and floodplain) elevation landforms along the river. The study demonstrated there was a preference for burials/ceremonial sites being placed at higher elevations (TI and higher) and habitation sites being placed at lower elevations (TL and current floodplain) near the main river channel.

Doctor of Philosophy(PhD) / It is generally accepted that tree roots can reinforce soil and improve the stability of vegetated slopes. Tree root reinforcement is also recognised in riverbanks although the contribution that the roots make to bank stability has rarely been assessed due to the reluctance of geomorphologists to examine riverbank stability by geomechanical methods that allow for the inclusion of quantified biotechnical parameters. This study investigates the interaction between alluvial soil and the roots of four southeastern Australian riparian trees. It quantifies the amount and distribution of root reinforcement present beneath typically vegetated riverbanks of the upper Nepean River, New South Wales, and examines the effect of the reinforcement on the stability of these banks. The ability of a tree to reinforce the soil is limited by the spatial distribution of its root system and the strength that the roots impart to the soil during shear. These two parameters were determined for the following four species of native riparian tree: Casuarina glauca, Eucalyptus amplifolia, Eucalyptus elata, and Acacia floribunda. The four species all exhibit a progressive reduction in the quantity of root material both with increasing depth and with increasing lateral distance from the tree stem. In the vertical direction there are two distinct zones that can be described. The first occurs from between 0 and approximately 15 % of the maximum vertical depth and consists of approximately 80 % of the total root material quantity. In this zone the root system consists of both vertical and lateral roots, the size and density of which varies between species. The second zone occurs below approximately 15 % of the maximum vertical depth and consists primarily of vertical roots. The quantity of root material in this zone decreases exponentially with depth due to the taper of individual roots. The earth reinforcement potential in terms of both geometric extent and the quantity of root material expressed as the Root Area Ratio (RAR) varies significantly from species to species. E. elata exhibited the highest values of RAR in soil zones beneath it while E. amplifolia reinforced a greater volume of soil than any of the other species examined. The increased shear resistance (Sr) of alluvial soil containing roots was measured by direct in-situ shear tests on soil blocks beneath a plantation. For three of the species (C. glauca, E. amplifolia, E. elata) Sr increased with increasing RAR measured at the shear plane, in a similar linear relationship. The shear resistance provided by A. floribunda roots also increased with increasing RAR at the shear plane but at a much greater rate than for the other three species. This is attributable to A. floribunda’s greater root tensile strength and therefore pull-out resistance, as well as its smaller root diameters at comparative RARs which resulted in a greater proportion of roots reaching full tensile strength within the confines of the test. Tree roots fail progressively in this system. Therefore determining the increased shear strength from the sum of the pull-out or tensile strengths of all individual roots and Waldron’s (1977) and Wu et al’s (1979) simple root model, would result in substantial over estimates of the overall strength of the soil-root system. The average difference between Sr calculated in this manner and that measured from direct in-situ shear tests is 10.9 kPa for C. glauca, 19.0 kPa for E. amplifolia, 19.3 kPa for E. elata, and 8.8 kPa for A. floribunda. A riverbank stability analysis incorporating the root reinforcement effect was conducted using a predictive model of the spatial distribution of root reinforcement beneath riparian trees within the study area. The model is based on measurements of juveniles and observations of the rooting habits of mature trees. It indicates that while the presence of vegetation on riverbank profiles has the potential to increase stability by up to 105 %, the relative increase depends heavily on the actual vegetation type, density, and location on the bank profile. Of the species examined in this study the greatest potential for improved riverbank stability is provided by E. amplifolia, followed by E. elata, A. floribunda, and C. glauca. The presence of trees on banks of the Nepean River has the potential to raise the critical factor of safety (FoS) from a value that is very unstable (0.85) to significantly above 1.00 even when the banks are completely saturated and subject to rapid draw-down. It is likely then that the period of intense bank instability observed within this environment between 1947 and 1992 would not have taken place had the riparian vegetation not been cleared prior to the onset of wetter climatic conditions. Typical ‘present-day’ profiles are critically to marginally stable. The introduction of vegetation could improve stability by raising the FoS up to 1.68 however the selection of revegetation species is crucial. With the placement of a large growing Eucalypt at a suitable spacing (around 3-5 m) the choice of smaller understorey trees and shrubs is less important. The effect of riparian vegetation on bank stability has important implications for channel morphological change. This study quantifies the mechanical earth reinforcing effect of some native riparian trees, thus allowing for improved deterministic assessment of historical channel change and an improved basis for future riverine management.

Healthy riparian vegetation has a positive impact on the adjacent river. Unfortunately, riparian vegetation is often threatened by human impacts such as dam construction and clearing. To gain the knowledge underlying the effects of such impacts and to aid riparian rehabilitation, the objective of this thesis was: to determine riparian vegetation association with, and response to, variation in fluvial geomorphology over several scales and consequently to fluvial disturbance. Only woody riparian plant species were considered. Flood disturbance was the unifying theme of this thesis. Linked to this theme and arising from the main objective was the supposition that plant interactions with the abiotic environment, but not biotic interactions between species, control riparian species distribution because of frequent fluvial disturbances. Woody riparian vegetation and riverine environmental variables were recorded along the upper Murrumbidgee River at three spatial scales based on a geomorphic hierarchy for Chapter 2. Multivariate analysis was used to group species and to associate environmental variables with vegetation at the three spatial scales. Observations at the two larger scales, of river segment (site) and riparian reach (transect), identified a river-longitudinal speciescomposition gradient associated with geology, river width and stream channel slope. Observations at the smallest scale of geomorphic units (plot) identified a lateral riparian gradient and also the longitudinal gradient; these gradients were associated with geomorphic variation, land use, plot elevation and also river longitudinal variables. Using the same data set, but varying the spatial scale of analysis caused the species composition pattern to change between scales. Increase in scale of observation, that is from geomorphic unit to reach and segment scales, resulted in disproportionate importance of rarer species and decreased importance of some key riparian species at the larger scales. It would appear that in this instance the geomorphic unit scale best described patches of different species composition because this scale had high spatial resolution and was also able to identify multiple gradients of environmental variation. It was recommended that riparian sampling take place at scales that represent dominant gradients in the riparian zone. These gradients are represented by geomorphic scales, indicating the appropriateness of using geomorphic based scales for observation of riparian vegetation. Chapter 3 considered whether there is a geomorphic template upon which riparian vegetation is patterned and whether it is associated with process variables, such as flooding and soil type. This question was investigated at different spatial scales in three ways: i) by an experiment to determine whether soil nutrient condition affects plant growth; ii) by graphical analysis of trends between geomorphic units, species and process variables; and iii) by analysis of vegetation distribution data. The smallest scale (meso) found experimental differences in plant growth because of soil type. Plants growing in sand had the lowest performance, with an average plant Relative Growth Rate (RGR) of 0.01, compared to plants growing in soils with small amounts of silt or clay particles, with an average plant RGR of 0.04. This pattern was attributed to differences in nutrients. Clear relationships were demonstrated at the larger geomorphic unit scale between species distribution and process variables. For example, hydrology and substratum type were found to be associated with geomorphic units and species. The largest scale considered in Chapter 3 was the riparian reach scale. At this scale species were clearly grouped around reach type. Therefore, geomorphology was considered to be a template for riparian species distribution. Findings in this chapter suggested that geomorphic variables should be good predictors of riparian species distribution. This hypothesis was tested and supported in Chapter 6. The experiments reported in Chapter 4 aimed to determine whether inundation depth and duration affected plant performance and survival for five common riparian zone species. Riparian seedling patterns in the field were also compared with experimental results to test whether species performance was reflected by field distribution. The experiments that were conducted included an inundation period and depth experiment, and a survival period test whilst under complete inundation. Biomass and height relative growth rates were determined, and the results were analysed using factorial Analysis of Variance. Obligate riparian species (Callistemon sieberi, Casuarina Cunninghamiana, Leptospermum obovatum) were found to be tolerant of inundation duration and depth, to the point where inundation provided a growth subsidy. On the other hand, non-obligate riparian species (Acacia dealbata, Kunzea ericoides) were either just tolerant of inundation or showed a negative growth response. For instance, C. sieberi demonstrated an average height RGR of 0.04 after complete inundation and 0.007 when not inundated, while A. dealbata had an average height RGR of 0.001 after complete inundation and 0.01 when not inundated. These experimental findings were found to closely reflect both seedling and adult plant distribution in the field such that inundation tolerant species were found close to the river and intolerant species further away. Thus, the conclusion was drawn that riparian species establishment and distribution is affected by inundation and that change to the flood regime could have serious impacts on riparian zone plant composition. The other aim of this chapter was to determine whether optimum germination temperatures were associated with flood or rainfall. Growth chamber germination trials were conducted at air temperatures of 15�C, 20�C and 25�C to determine the 'best' germination temperature. These germination patterns at different temperatures were then related to annual variation in field temperature, flooding period and rainfall. No evidence was found to suggest a relationship between ideal germination temperature and flood season, rather it was suggested that germination was patchy through time and may simply reflect recent rainfall. Investigations that were reported in Chapter 5 aimed to elucidate relationships between species and flow velocity variables. Two experiments were conducted: i) a flume experiment to determine the effect of flow velocity on plant growth; and ii) an experiment to observe the response of plants to damage (imitating flood damage) and inundation. Field observations of species distribution and flow velocity related variables were also conducted to put the flume results into a real-world context. Treatments for the flume experiment were fast flow velocity (0.74 m s-1), slow velocity (0.22 m s-1) and no velocity (control) but still inundated. All treatments were flooded completely for four days. Subsequent biomass and height relative growth rates were determined, and the results were analysed using factorial Analysis of Variance. Results were unexpected, given that obligate species exposed to the fastest velocity had the highest growth rate with an average height RGR of 0.046, compared to plants in still water, which grew the least with an average height RGR of 0.013. It was hypothesised that this response was because relatively greater carbon dioxide and oxygen levels were available in the moving water compared to the still water. With regard to shoot damage, the species that were nonobligate riparian species lost more leaves from velocity treatment than the obligate riparian species. The cut and flood experiment found growth of the obligate species (Casuarina cunninghamiana) to be greater after cutting than the non-obligate species. Flooding was not found to have an effect in the cut and flood experiment, probably because the period to sampling after flood treatment was longer (4 weeks) than other flooding experiments (3 weeks). Field observations were found to support the experimental findings, with a gradient of species across the riparian zone that reflected potential flood velocities. Therefore, velocity is one of a suite of riparian hydrological factors that are partially responsible for the gradient of species across the riparian zone. Potentially the absence of flooding could result in a homogeneous mix of species, rather than a gradient, except on the very edge of the river. The study that was reported in Chapter 6 investigated a technique for predicting riparian vegetation distribution. One of the aims of this investigation was to address a current riparian rehabilitation shortfall, which was how to objectively select species to plant for rehabilitation. Field data were collected from three confined river valleys in south-eastern New South Wales. Using data on plant species occurrence and site and plot measures of soils, hydrology and climate, an AUSRIVAS-style statistical model, based on cluster and discriminant analysis, was developed to predict the probability of species occurrence. The prediction accuracy was 85 % when tested with a separate set of plots not used in model construction. Problems were encountered with the prediction of rarer species, but if the probability of selection was varied according to the frequency of species occurrence then rarer species would be predicted more often. Various models were tested for accuracy including three rivers combined at the geomorphic unit (plot) scale and riparian reach (transect) scale in addition to a Murrumbidgee River plot scale model. Surprisingly, the predictive accuracy of the all rivers and single river models were approximately the same. However, the difference between the large scale and small scale models pointed to the importance of including small scale flood-related parameters to predict riparian vegetation. When these riparian predictions were compared to predictive outcomes from a hill slope model, which was assumed to be affected by fewer disturbances (i.e. flooding), predictive accuracies were not very different. Overall though, predictive accuracy for riparian vegetation was high, but not good enough to support the supposition that riparian vegetation is abiotically controlled because of frequent flood disturbance. Nevertheless, geomorphology and consequently flood effects are still important for the determination of the riparian community composition. Overall, riparian vegetation was found to be closely linked to its environment (evidenced in Chapters 2, 3, 4, 5) in a predictable manner (Chapter 6). Species pattern relied on flood disturbance affecting species distribution. Some riparian species were found to be highly tolerant of flooding and gained a growth advantage after flooding (Chapters 4 and 5). Therefore, flood tolerance was important for the formation of a species gradient across the riparian zone. These species tolerances and growth requirements reflect riparian geomorphic pattern (Chapter 3), which was suggested to form a template on which riparian vegetation is structured.

This dissertation reports the results of a field based study examining sediment transport and bed mobility in a low-ordered, ephemeral watershed. Runoff and sediment transport concentrations were sampled at the watershed outlet to determine flow discharge and sediment flux during approximately 21 flow events, from 1998 - 2007. Sediment collected in flow was measured for grain-size distribution to determine if specific grain-size fractions behave differently while in transport. The coarse sediment yield was measured for mass and grain-size distribution at the watershed outlet for two years, 2005 - 2006. Further, the arrangement and composition of the channel bed material was comprehensively mapped using terrestrial-based photogrammetry for the years, 2005 - 2006. Results show that patterns of sediment transport are complex, controlled in part by flow hydraulics but also by other phenomena. Some of the variation in sediment transport is determined by grain-size. Grain-sizes with different sources within the watershed and that transported by different transport modes were observed to follow different patterns of transport. Also, the channel bed, which serves as the source for the coarse fraction of the sediment transport, was observed to change in grain composition during periods of flow. This tendency for the bed material to evolve in time likely affected the amount and composition of the sediment grains that were entrained from it.An additional objective of this dissertation was to determine how unique the observed patterns of sediment transport were to low-order ephemeral channels. Sediment transport and yield were modeled using bed load transport formulae designed to capture the physical mechanics of transport as observed in perennial streams. Results show that contemporary transport models predict transport within the field site with similar accuracy as that in many perennial systems but not well enough to rely on their predictions for many engineering applications.

The Mgeni River is the major water resource for the eThekwini Metropolitan and Msunduzi Municipalities. At the end of 2002, the Mooi-Mgeni Transfer Scheme Phase 1, which transfers water from the Mooi River into the Mgeni catchment to augment the water supply to this region, was completed. The interbasin transfer of water resulted in the loss of habitat, erosion of the stream channel and transformation of the riparian zone in the receiving streams. Stream regulation resulting in an altered flow regime is considered the greatest threat to a riverine environment. An Environmental Management Plan (EMP), incorporating fluvial geomorphological monitoring procedures, was implemented to monitor the impact of the transfer on the receiving streams, the Mpofana and Lions Rivers, and to determine the rate and magnitude of erosion. A comparison of the geomorphological monitoring procedure of the EMP with best practice geomorphological monitoring derived from a review of the national and international stream geomorphological literature was conducted in this study. In addition, the implementation of the EMP geomorphological monitoring procedures was described and onsite observations of physical impacts on the receiving streams were completed. The geomorphological monitoring of the EMP included the use of erosion pins, survey of stream cross-sections and fixed-point photography. Photographs and data were collected from February 2003 to June 2006. The comparison of these monitoring methods against stream assessment best practices revealed the strengths and weaknesses of the geomorphological monitoring implemented in the receiving streams. Several key weaknesses were revealed. Firstly, an inadequate number of stream cross sections was included in the monitoring procedures. Secondly, although the erosion pins indicated some general trends in the erosion of the stream channel, they did not give a true impression of the rate and magnitude of change in slope and channel width of the stream, and the location of the erosion pins sites did not take into account the actual direction of flow during transfer as erosion pin sites were selected during low flow conditions. In addition, it was difficult to determine whether the erosion pins had been lost due to erosion or to turbulence. The results were difficult to assess and did not show whether the erosion was localised at the pins or the section of bank or stream profile. Thirdly, analysis of platform changes in the stream channel (e.g. through a comparison of aerial photograph sets) was lacking and no attempt was made to integrate the results from the different methods. Overall, the study concluded that the geomorphological monitoring of the EMP was limited, and it did not highlight the rate and magnitude of erosion in the receiving streams. Based on the findings of this study, recommendations are provided for geomorphological monitoring of the receiving streams of the Mooi Mgeni Transfer Scheme. / Thesis (M.Env.Dev.)-University of KwaZulu-Natal, Pietermaritzburg, 2009.

Hydraulic geometry relationships and regional curves are used in natural channel design to assist engineers, biologists, and fluvial geomorphologists in the efforts undertaken to ameliorate previous activities that have diminished, impaired or destroyed the structure and function of stream systems. Bankfull channel characteristics were assessed for 14 United States Geological Survey (USGS) gaged sites in the Inner Bluegrass and 15 USGS gaged sites in the Outer Bluegrass Regions of Kentucky. Hydraulic geometry relationships and regional curves were developed for the aforementioned regions. Analysis of the regression relationships showed that bankfull discharge is a good explanatory variable for bankfull parameters such as area, width and depth. The hydraulic geometry relationships developed produced high R2 values up to 0.95. The relationships were also compared to other studies and show strong relationships to both theoretical and empirical data. Regional curves, relating drainage area to bankfull parameters, were developed and show that drainage area is a good explanatory variable for bankfull parameters. R2 values for the regional curves were as high as 0.98.

Doctor of Philosophy(PhD) / It is generally accepted that tree roots can reinforce soil and improve the stability of vegetated slopes. Tree root reinforcement is also recognised in riverbanks although the contribution that the roots make to bank stability has rarely been assessed due to the reluctance of geomorphologists to examine riverbank stability by geomechanical methods that allow for the inclusion of quantified biotechnical parameters. This study investigates the interaction between alluvial soil and the roots of four southeastern Australian riparian trees. It quantifies the amount and distribution of root reinforcement present beneath typically vegetated riverbanks of the upper Nepean River, New South Wales, and examines the effect of the reinforcement on the stability of these banks. The ability of a tree to reinforce the soil is limited by the spatial distribution of its root system and the strength that the roots impart to the soil during shear. These two parameters were determined for the following four species of native riparian tree: Casuarina glauca, Eucalyptus amplifolia, Eucalyptus elata, and Acacia floribunda. The four species all exhibit a progressive reduction in the quantity of root material both with increasing depth and with increasing lateral distance from the tree stem. In the vertical direction there are two distinct zones that can be described. The first occurs from between 0 and approximately 15 % of the maximum vertical depth and consists of approximately 80 % of the total root material quantity. In this zone the root system consists of both vertical and lateral roots, the size and density of which varies between species. The second zone occurs below approximately 15 % of the maximum vertical depth and consists primarily of vertical roots. The quantity of root material in this zone decreases exponentially with depth due to the taper of individual roots. The earth reinforcement potential in terms of both geometric extent and the quantity of root material expressed as the Root Area Ratio (RAR) varies significantly from species to species. E. elata exhibited the highest values of RAR in soil zones beneath it while E. amplifolia reinforced a greater volume of soil than any of the other species examined. The increased shear resistance (Sr) of alluvial soil containing roots was measured by direct in-situ shear tests on soil blocks beneath a plantation. For three of the species (C. glauca, E. amplifolia, E. elata) Sr increased with increasing RAR measured at the shear plane, in a similar linear relationship. The shear resistance provided by A. floribunda roots also increased with increasing RAR at the shear plane but at a much greater rate than for the other three species. This is attributable to A. floribunda’s greater root tensile strength and therefore pull-out resistance, as well as its smaller root diameters at comparative RARs which resulted in a greater proportion of roots reaching full tensile strength within the confines of the test. Tree roots fail progressively in this system. Therefore determining the increased shear strength from the sum of the pull-out or tensile strengths of all individual roots and Waldron’s (1977) and Wu et al’s (1979) simple root model, would result in substantial over estimates of the overall strength of the soil-root system. The average difference between Sr calculated in this manner and that measured from direct in-situ shear tests is 10.9 kPa for C. glauca, 19.0 kPa for E. amplifolia, 19.3 kPa for E. elata, and 8.8 kPa for A. floribunda. A riverbank stability analysis incorporating the root reinforcement effect was conducted using a predictive model of the spatial distribution of root reinforcement beneath riparian trees within the study area. The model is based on measurements of juveniles and observations of the rooting habits of mature trees. It indicates that while the presence of vegetation on riverbank profiles has the potential to increase stability by up to 105 %, the relative increase depends heavily on the actual vegetation type, density, and location on the bank profile. Of the species examined in this study the greatest potential for improved riverbank stability is provided by E. amplifolia, followed by E. elata, A. floribunda, and C. glauca. The presence of trees on banks of the Nepean River has the potential to raise the critical factor of safety (FoS) from a value that is very unstable (0.85) to significantly above 1.00 even when the banks are completely saturated and subject to rapid draw-down. It is likely then that the period of intense bank instability observed within this environment between 1947 and 1992 would not have taken place had the riparian vegetation not been cleared prior to the onset of wetter climatic conditions. Typical ‘present-day’ profiles are critically to marginally stable. The introduction of vegetation could improve stability by raising the FoS up to 1.68 however the selection of revegetation species is crucial. With the placement of a large growing Eucalypt at a suitable spacing (around 3-5 m) the choice of smaller understorey trees and shrubs is less important. The effect of riparian vegetation on bank stability has important implications for channel morphological change. This study quantifies the mechanical earth reinforcing effect of some native riparian trees, thus allowing for improved deterministic assessment of historical channel change and an improved basis for future riverine management.