No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Sampling Procedures for Coarse Fluvial Sediments
... Most authors investigating the sizes and shapes of coarse-grained material have focused on modern J o u r n a l P r e -p r o o f Journal Pre-proof 4 systems where individual grains are lying flat on gravel bars and where the a-/b-axes plane can be viewed from above (Johansson, 1976;Brayshaw, 1984;Strom et al., 2010). However, the partial hiding of clasts due to imbrication or burial of individual grains poses major challenges when collecting grain size data from photos (e.g., Kellerhals and Bray, 1971;Adams, 1979;Graham et al., 2010). Furthermore, since photos display the grains as projections in 2D, they cannot resolve the full 3D-view of a single grain, which introduces an additional bias during the collection of such grain size datasets (e.g., Warrick et al., 2009;Stähly et al., 2017). ...
... Over the past decades, the quantification of coarse material in modern streams has undergone a significant development. Time-consuming in-situ class counting (e.g., Wolman, 1954) and sieving techniques (e.g., Batel, 1960) were partially substituted by manual collections of grain size datasets on photos (e.g., Ritter and Helley, 1969;Kellerhals and Bray, 1971;Adams, 1979) and approaches where clasts were semiautomatically measured (e.g., Butler et al, 2001;Buscombe, 2008;Graham et al., 2010;Purinton and Bookhagen, 2019). Grain measurements on photos (both manually or semi-automatic) are usually accomplished on a selection of grains only, using either grid-by-area (e.g., Ibbeken and Schleyer, 1986; Church et al., 1987) or grid-by-number concepts (i.e., class-based; e.g., Wolman, 1954;Kellerhals and Bray, 1971). ...
... Time-consuming in-situ class counting (e.g., Wolman, 1954) and sieving techniques (e.g., Batel, 1960) were partially substituted by manual collections of grain size datasets on photos (e.g., Ritter and Helley, 1969;Kellerhals and Bray, 1971;Adams, 1979) and approaches where clasts were semiautomatically measured (e.g., Butler et al, 2001;Buscombe, 2008;Graham et al., 2010;Purinton and Bookhagen, 2019). Grain measurements on photos (both manually or semi-automatic) are usually accomplished on a selection of grains only, using either grid-by-area (e.g., Ibbeken and Schleyer, 1986; Church et al., 1987) or grid-by-number concepts (i.e., class-based; e.g., Wolman, 1954;Kellerhals and Bray, 1971). Nowadays the flourishing use of uncrewed aerial vehicles (i.e., drones) allows simple and rapid surveys of large areas. ...
The size of grains in gravel and conglomerate deposits is most easily measured on photos taken from related outcrops. However, the occlusion of grains by the sedimentary matrix or other grains, and possible distortions of photos, could introduce a bias in such datasets. Here, we explore the uncertainties associated with datasets where the lengths of the grains were measured on photos. To this end, we analysed coarse-grained (>2 mm) fluvial material from a gravel pit (Bern, Switzerland). We compared grain size data collected from digital photos with the results where the same material was measured with a calliper and mechanically sieved. Our analyses reveal that the percentile values such as the D16, D50 and D84 of datasets where the grains’ longest visible axes were measured on digital photos best correlate to the corresponding percentile values of data collected through sieving. We also find that the longest visible axes of grains measured on digital photos are c. 17% smaller than the lengths of the intermediate b-axes of grains measured with a calliper. We therefore suggest to measure the longest visible axes on digital photos, and to correct the data by a corresponding factor such as +17% for the target grain size percentiles.
... However, the measurement of grain size has been time-consuming and laborious especially in mountain rivers due to the wide range of grain size classes, diverse grain lithology, the hiding of grains, diverse structures and the influence of organic materials. The most widely used grain sizing method is sieving (Kellerhals and Bray, 1971) which is used as a benchmark to other methods when reliable sediment samples are able to be collected (Church et al., 1987). Wolman (1954) proposed a pebble count method (Wolman method) that samples a minimum of 100 pebbles from the riverbed surface with a grid-based system. ...
... Wolman (1954) proposed a pebble count method (Wolman method) that samples a minimum of 100 pebbles from the riverbed surface with a grid-based system. Limited to material > 8 mm (Kellerhals and Bray, 1971), the Wolman method has been especially popular in the field due to the limited equipment required and its benefit of reducing sampling times while providing a relatively valid estimation of reach-scale grain size distribution. Since then, various versions of the Wolman method have been proposed with different approaches to collecting stones such as the random walk approach for particle collection (Leopold, 1970), superim-posing gravel templates upon the sedimentological unit for reduced operator error (Bunte and Abt, 2001) and imagebased Wolman method analysis . ...
... For a predicted image, the a axis (major axis) of a grain was defined as the maximum Euclidean distance between two pixels on the grain boundary, and the b axis (minor axis) was calculated as the maximum intercept to the grain along a line perpendicular to the a axis. Based on the b axis and gridby-area method (Kellerhals and Bray, 1971), sediment percentiles D 5 , D 16 , D 50 , D 84 and D 95 were calculated for the results of manual labeling, GrainID and BASEGRAIN. The sediment percentiles of the Wolman method were calculated based on a grid-by-number method equivalent to the grid-byarea method demonstrated by Kellerhals and Bray (1971). ...
Image-based grain sizing has been used to measure grain size more efficiently compared with traditional methods (e.g., sieving and Wolman pebble count). However, current methods to automatically detect individual grains are largely based on detecting grain interstices from image intensity which not only require a significant level of expertise for parameter tuning but also underperform when they are applied to suboptimal environments (e.g., dense organic debris, various sediment lithology). We proposed a model (GrainID) based on convolutional neural networks to measure grain size in a diverse range of fluvial environments. A dataset of more than 125 000 grains from flume and field measurements were compiled to develop GrainID. Tests were performed to compare the predictive ability of GrainID with sieving, manual labeling, Wolman pebble counts (Wolman, 1954) and BASEGRAIN (Detert and Weitbrecht, 2012). When compared with the sieving results for a sandy-gravel bed, GrainID yielded high predictive accuracy (comparable to the performance of manual labeling) and outperformed BASEGRAIN and Wolman pebble counts (especially for small grains). For the entire evaluation dataset, GrainID once again showed fewer predictive errors and significantly lower variation in results in comparison with BASEGRAIN and Wolman pebble counts and maintained this advantage even in uncalibrated rivers with drone images. Moreover, the existence of vegetation and noise have little influence on the performance of GrainID. Analysis indicated that GrainID performed optimally when the image resolution is higher than 1.8 mm pixel−1, the image tile size is 512×512 pixels and the grain area truncation values (the area of smallest detectable grains) were equal to 18–25 pixels.
... However, the measurement of grain size has 25 been time-consuming and laborious especially in mountain rivers due to the wide range of grain size classes, diverse grain lithology, the hiding of grains, diverse structures and the influence of organic materials. The most widely used grain-sizing method is sieving (Kellerhals and Bray, 1971) and is used as a benchmark to other methods when reliable sediment samples are able to be collected (Church et al., 1987). Wolman (1954) proposed a pebble count method (Wolman method) that samples a minimum of 100 pebbles from the riverbed surface with a grid-based system. ...
... The image-based Wolman method samples 100 grains based on an equidistant grid on the image where the sediment distribution was calculated via a grid-by-number approach that has been applied in many literatures (Kellerhals and Bray, 1971;195 Hassan et al., 2020). ...
... The a-axis (major-axis) of a grain was defined as the maximum Euclidean distance between two points on the grain boundary, and the b-axis (minor-axis) was calculated as the maximum intercept to the grain along a line perpendicular to the a-axis. Based on the b-axis and grid-by-area method (Kellerhals and Bray, 1971), sediment percentiles D5, D16, D50, D84 and D95 were calculated for the results of manual labeling, GrainID and BASEGRAIN. The sediment percentiles of the Wolman method were calculated based on a grid-by-210 number method equivalent to the grid-by-area method demonstrated by Kellerhals and Bray (1971). ...
Image-based grain sizing has been used to measure grain size more efficiently compared to traditional methods (e.g. sieving and Wolman pebble count). However, current methods (e.g. BASEGRAIN) are largely based on detecting grain interstices from image intensity which not only require a significant level of expertise for parameter tuning but also underperform when they are applied to sub-optimal environments (e.g. dense organic debris, various sediment lithology). We proposed a model (GrainID) based on convolutional neural networks to measure grain size in a diverse range of fluvial environments. A data set of more than 125,000 grains from flume and field measurements were compiled to develop GrainID. Tests were performed to compare the predictive ability of GrainID with sieving, manual labeling, Wolman pebble counts and BASEGRAIN. When compared with the sieving results for a sandy-gravel bed, GrainID yielded high predictive accuracy (comparable to the performance of manual labeling) and outperformed BASEGRAIN and Wolman Pebble counts (especially for small grains). For the entire evaluation dataset, GrainID once again showed fewer predictive errors and significantly lower variation in results in comparison to BASEGRAIN and Wolman pebble counts and maintained this advantage even in uncalibrated rivers with drone images. Moreover, the existence of vegetation and noise have little influence on the performance of GrainID. Analysis indicated that GrainID performed optimally when the image resolution is higher than 1.8 mm/pixel, the image tile size is 512*512 pixel*pixel and the grain area truncation values (the area of smallest detectable grains) were equal to 18–25 pixels.
... Since the output of the DGS algorithm is a distribution of line-by-number grain diameters (see Kellerhals and Bray [39], Church et al. [40] for descriptions of the types particle size distributions), a conversion factor is needed in order for the DGS output to be comparable to (i.e., dimensionally consistent with) output from a sieve-type analysis. A commonly used conversion formula is [37,39,41]: ...
... where p 1,i is the known proportion of the ith size fraction obtained using the input measure, p 2,i is the proportion of the ith size fraction in units consistent with the desired output measure, D i is the grain diameter of the ith size fraction, and the exponent x is a conversion constant whose value is empirically dependent upon the grain size distribution. Equation (A1) is based on the voidless cube model from Kellerhals and Bray [39]. The Kellerhals and Bray [39] conversion is based on purely dimensional arguments, and does not depend upon an idealisation of the material. ...
... Equation (A1) is based on the voidless cube model from Kellerhals and Bray [39]. The Kellerhals and Bray [39] conversion is based on purely dimensional arguments, and does not depend upon an idealisation of the material. Thus, though the parameter x can be theoretically defined based only on knowledge of the input and output measures, it is best employed as an empirically defined tuning parameter. ...
On mixed sand–gravel beaches, impacts from gravel- and cobble-sized grains—mobilized by the energetic shorebreak—limit the utility of in situ instrumentation for measuring the small-scale response of the beach face on wave period time scales. We present field observations of swash zone morpho-sedimentary dynamics at a steep, megatidal mixed sand–gravel beach using aeroacoustic and optical remote sensing. Coincident observations of bed level and mean surficial sediment grain size in the swash zone were obtained using an array of optical cameras paired with acoustic range sensors. Lagrangian tracking of swash-transported cobbles was carried out using an additional downward-oriented camera. The principal objective of the study was to investigate linkages between sediment grain size dynamics and swash zone morphological change. In general, data from the range sensor and camera array show that increases in bed level corresponded to increases in mean grain size. Finer-scale structures in the bed level and mean grain size signals were observable over timescales of minutes, including signatures of bands of coarse-grained material that migrated shoreward with the leading edge of the swash prior to high tide berm formation. The direction and magnitude of cobble transport in the swash varied with cross-shore position, and with the composition of the underlying bed. These results demonstrate that close-range remote sensing techniques can provide valuable insights into the roles of cobble-sized versus sand-sized particle dynamics in the swash zone on mixed sand–gravel beaches.
... A number of methods exist for extracting GSDs from images, including manual digitizing (e.g., Ibbeken & Schleyer, 1986), automated pixel-based methods (e.g., Buscombe, 2013), and automated segmentation methods (e.g., Butler et al., 2001;Kozakiewicz, 2018). Methods for extracting grain size data from rover images used to date have been some form of grid-by-numbers (e.g., Jerolmack et al., 2006or Ewing et al., 2017 methodology adapted from Kellerhals & Bray, 1971) or clustered/targeted sampling strategy (Banham et al., 2018;Weitz et al., 2018) paired with a manually digitized line of the user-estimated short and/or long axis of grains as they appear in the image. Crucially, presenting the grain count results of a grid-by-numbers approach, where a grid is overlain on an image and grains at the grid intersections are measured, can directly produce volumetrically equivalent data because the increased probability of a larger area grain falling on a grid point proxies volume (Kellerhals & Bray, 1971). ...
... Methods for extracting grain size data from rover images used to date have been some form of grid-by-numbers (e.g., Jerolmack et al., 2006or Ewing et al., 2017 methodology adapted from Kellerhals & Bray, 1971) or clustered/targeted sampling strategy (Banham et al., 2018;Weitz et al., 2018) paired with a manually digitized line of the user-estimated short and/or long axis of grains as they appear in the image. Crucially, presenting the grain count results of a grid-by-numbers approach, where a grid is overlain on an image and grains at the grid intersections are measured, can directly produce volumetrically equivalent data because the increased probability of a larger area grain falling on a grid point proxies volume (Kellerhals & Bray, 1971). However, doing the same grain count procedure for targeted areas of adjacent grains will not (Bunte & Abt, 2001;Johnson, 1994). ...
Plain Language Summary
Mars and Earth have many different sizes of sand ripples and sand dunes. The Curiosity rover took pictures of large sand ripples with one meter spacing between crests. On Earth these large ripples are often called “megaripples”. However, it is hard to know if the meter‐scale ripples observed by Curiosity form in the same way as Earth’s megaripples. To help solve this problem, we measured more than twenty thousand sand grains from close‐up images of the ripples taken by the Mars Hand Lens Imager on Curiosity. The method we used to measure the imaged grains simulates the measurement method used for real sand samples on Earth. Using a method like this is important because it allows for direct comparison between Earth sand and Mars sand. With this method, we find that the large ripples observed by Curiosity have larger sand grains than the smaller ripples. Earth’s megaripples also have larger grains than smaller ripples. The difference in sand grain size we measured provides evidence that large Martian ripples form in a way that is similar to Earth’s “megaripples”.
... To compare this distribution to the Wolman field counts, it must be converted to a gridby-number distribution, which is considered equivalent to a volumetric grain-size distribution. Conversion factors have been proposed to convert grain-size data acquired with one approach to another one, based on geometrical arguments (Kellerhals and Bray, 1971;Church et al., 1987;Diplas and Fripp, 1992). For example, converting an area-by-number (or areal) distribution to a grid-by-number (or volumetric; e.g., Wolman) distribution requires multiplying the frequency of all the particle classes by the factor D 2 . ...
The grain-scale morphology and size distribution of sediments are important factors controlling the erosion efficiency, sediment transport and the aquatic ecosystem quality. In turn, characterizing the spatial evolution of grain size and shape can help understand the dynamics of erosion and sediment transport in coastal, hillslope and fluvial environments. However, the size distribution of sediments is generally assessed using insufficiently representative field measurements, and determining the grain-scale shape of sediments remains a real challenge in geomorphology. Here we determine the size distribution and grain-scale shape of sediments located in coastal and river environments with a new methodology based on the segmentation and geometric fitting of 3D point clouds. Point cloud segmentation of individual grains is performed using a watershed algorithm applied here to 3D point clouds. Once the grains are segmented into several sub-clouds, each grain-scale morphology is determined by fitting a 3D geometrical model applied to each sub-cloud. If different geometrical models can be tested, this study focuses mostly on ellipsoids to describe the geometry of grains. G3Point is a semi-automatic approach that requires a trial-and-error approach to determine the best combination of parameter values. Validation of the results is performed either by comparing the obtained size distribution to independent measurements (e.g., hand measurements) or by visually inspecting the quality of the segmented grains. The main benefits of this semi-automatic and non-destructive method are that it provides access to (1) an un-biased estimate of surface grain-size distribution on a large range of scales, from centimeters to meters; (2) a very large number of data, mostly limited by the number of grains in the point cloud data set; (3) the 3D morphology of grains, in turn allowing the development of new metrics that characterize the size and shape of grains; and (4) the in situ orientation and organization of grains. The main limit of this method is that it is only able to detect grains with a characteristic size significantly greater than the resolution of the point cloud.
... Garcia et al., 1999;Vericat et al., 2008Vericat et al., , 2020 and would therefore entrain, if present. Within this context, in those sections where the presence of fine particles is considered significant, combining pebble-count and area by weight samples (Kellerhals and Bray, 1971) may provide a GSD representative of the full spectrum of sizes on the bed (e.g. Vericat et al., 2006). ...
Hydropeaking, by artificially generated flow peaks, influences hydro-sedimentary dynamics on rivers and, consequently, affects bed material entrainment and transport. This study examines the onset of motion of sediment particles in four sections of a Pyrenean gravel-to-cobble bed river exposed to frequent hydropeaking (once per day, on average). Five criteria of particle entrainment have been used to assess the prediction of the initiation of grain motion at-a-section scale. Theoretical entrainment conditions were validated using real observations of mobility by means of tracers. It was found that the maximum flow discharged by the hydropower plant mostly affects the furthest downstream section, located almost 17 km downstream, in which the finer fractions of the bed are entrained. The mobile grain sizes include up to coarse gravels (≈ 30 mm). Differences in sediment supply (imposed by tributaries), the value of the bed slope and the structure of the coarse surface layer decisively control the downstream variability of incipient particle motion between sections. Results from a 17 km study segment indicated that hydropeaking generate partial transport, that is, a partially size-selective transport that occurs downstream from the hydropower plant and winnows the sand and small gravel further downstream, increasing armouring and depleting fine sediments.
... A comparison of the D max at each dredge site with the total sampled mass shows that 73 % of all sites were within 5 % of this rule and 30 % of all sites were within 1 % of this rule. According to Kellerhals and Bray (1971), the particle-size distributions determined from volume-by-weight and grid-by-number samples are equivalent in the sense that there is no need to apply a correction factor when comparing counts by number with counts by volume or by mass (Bunte and Abt, 2001). Nevertheless, to the extent that sample procedures are different (and submitted to different biases), in this work we restricted ourselves to comparisons of data collected D. Vázquez-Tarrío et al. using the same procedure: data from Wolman counts or data from dredge samples. ...
Present-day river forms and processes are in many cases conditioned by the consequences of anthropogenic modifications such as dams, embankments and gravel mining. Fluvial geomorphologists have typically investigated the effects of these human impacts using a so-called expert-based approach, whereby observed association or synchronicity between geomorphological changes and a given, preidentified impact, are interpreted as evidence of causation. This approach has important limitations when the effects of multiple human interventions interact along the same river corridor or overlap with the legacy of natural changes affecting the sediment - water balance. In such situations, the establishment of causal links between channel morphology and single impacts is not as straightforward as commonly assumed and the conclusions are susceptible to ‘confirmation biases’. In this paper we highlight this risk through an assessment of human impacts on the Rhône River within a multi-driver context. The French Rhône is an excellent example of an Anthropocene river impacted by two main development phases during the twentieth century: embankments (1890s–1930s) followed by a series of multiple dams (1950s–1990s). We began by laying out several geomorphologically consistent hypotheses for the geomorphological trajectory of the Rhône over the twentieth century. Next, we tested these hypotheses against grain size data collected in the field in a structured and hypothesis-oriented way. Using this hypothesis-driven and deductive attribution analysis we identified the relative impacts of the different development phases on the present-day grain size distribution and in particular on armouring in the Rhône River, and proposed a hierarchy of dominant drivers of geomorphological change along the Rhone over the last century and a half. Our results led us to conclude that in the case of the Rhône, the effect of dams on armouring was negligible compared to a legacy of natural heritages and embankments.
... Sediment grain size distributions (GSDs) vary substantially in different Henry Mountains channels, because of spatial variability in coarse sediment that is eroding from older localized pediment remnants and from igneous intrusions outcropping upstream in some but not all watersheds in the area (Johnson et al., 2009). GSDs were measured by random-walk point counts (corresponding to "grid by number", Kellerhals & Bray, 1971) of 200 intermediate particle diameters in reaches away from springs (i.e., sparse vegetation reaches) in both Trail and Woodruff Canyons. A fine-tipped stylus (e.g., pencil) was used to touch the ground without looking. ...
A better understanding of how vegetation influences alluvial channels could improve (a) assessments of channel stability and flood risks, (b) applications of vegetation as a river management tool, and (c) predictions of channel responses to climate change and other human impacts. We take advantage of a natural field experiment in the semi‐arid to arid Henry Mountains, Utah, USA: large spatial differences in bed and bank vegetation are found along some alluvial channels due to localized perennial springs caused by aquicludes in the underlying bedrock. Airborne LiDAR topography and flood modeling are used to constrain channel morphology, vegetation density, and flow velocity at different flood discharges for three spring‐fed reaches along intermittently flowing streams. The spatial distribution of vegetation quantitatively influences both the magnitude and direction of channel adjustment. Reaches with abundant bed vegetation are significantly wider (by an average of ≈50%), with shallower flows and lower velocities, than reaches with little bed vegetation. Reaches with dense channel bank vegetation are ≈25% narrower and ≈25% deeper than sparsely vegetated reaches. We interpret that sediment grain size influences the spatial distribution of vegetation within spring reaches, but that bank vegetation may be more important than grain size for “threshold” width adjustments. Widths, depths, and velocities are fairly insensitive to whether local hydraulic roughness is parameterized in terms of local vegetation density or is assumed spatially constant, suggesting that the underlying “bare earth” topography of the channel bed and banks exerts more control on local flow than does local vegetation density in this landscape.
... Because previous studies had three sampling methods (i.e., grid, areal, and volumetric sampling) and two analyzed categories (i.e., by number and by weight), various conversion methods were proposed to obtain size distributions with the same sample and analysis categories (e.g., Bunte & Abt, 2001). Kellerhals and Bray (1971) used a model deposit comprising a mixture of three cube sizes packed without voids (i.e., a voidless cube model) to propose conversion methods between the various combinations of sampling methods and sample analyses. They concluded that the grid-by-number procedure is the only surface-oriented procedure directly comparable (equivalent) to customary bulk sieve analysis. ...
Landslide‐induced debris flows can travel long distances. Many field studies, laboratory experiments, and theoretical studies have been conducted to clarify the mechanisms of long‐runout landslides. However, information concerning landslide‐induced debris flows is still inadequate despite several explanations proposed. Thus, we collected various forms of data, including video images and light detection and ranging data, for a landslide‐induced debris flow that occurred on 28 July 2015, in Fukaminato River, Japan, and clarified the debris flow behavior from initiation to deposition. We showed that the water content of the debris flow should only be approximately 30% immediately after landslides occurred at the site. However, a fluid phase‐dominated flow flowed over 1.2 km, expanding the damaged area. We inferred that the fine sediments behaved as part of the interstitial fluid in the debris flow and that, during debris flow deposition, some fine sediments were stored in the interstitial space of the riverbed and did not contribute to forming the skeleton of deposits. We conducted a numerical simulation to test the processes of fluid phase‐dominated flow propagation, focusing on the flow and deposition of fine sediments. Our simulation results suggest that fine sediments' flow and depositional processes significantly influence both the initiation of fluid phase‐dominated debris flow and the travel distance of landslide‐induced debris flow.
... GSDs were obtained using grid-by-number sampling, following established protocols developed for measuring riverine GSDs (see Kellerhals and Bray, 1971) and subsequently applied to landslide deposits (Casagli et al., 2003;Attal and Lavé, 2006). We extended survey tapes along an elevation contour over a substantial portion of the deposit width (10 to 50 m) and sampled grains along the tape at a constant interval, recording the size bin of the b axis, measured with rulers. ...
The size of grains delivered to rivers by hillslope processes is thought to be a key factor controlling sediment transport, long-term erosion and the information recorded in sedimentary archives. Recently, models have been developed to estimate the grain size distribution produced in soil, but these models may not apply to active orogens where high erosion rates on hillslopes are driven by landsliding. To date, relatively few studies have focused on landslide grain size distributions. Here, we present grain size distributions (GSDs) obtained by grid-by-number sampling on 17 recent landslide deposits in Taiwan, and we compare these GSDs to the geometrical and physical properties of the landslides, such as their width, area, rock type, drop height and estimated scar depth. All slides occurred in slightly metamorphosed sedimentary units, except two, which occurred in younger unmetamorphosed shales, with a rock strength that is expected to be 3–10 times weaker than their metamorphosed counterparts. For 11 landslides, we did not observe substantial spatial variations in the GSD over the deposit. However, four landslides displayed a strong grain size segregation on their deposit, with the overall GSD of the downslope toe sectors being 3–10 times coarser than apex sectors. In three cases, we could also measure the GSD inside incised sectors of the landslides deposits, which presented percentiles that were 3–10 times finer than the surface of the deposit. Both observations could be due to either kinetic sieving or deposit reworking after the landslide failure, but we cannot explain why only some deposits had strong segregation. Averaging this spatial variability, we found the median grain size of the deposits to be strongly negatively correlated with drop height, scar width and depth. However, previous work suggests that regolith particles and bedrock blocks should coarsen with increasing depth, which is the inverse of our observations. Accounting for a model of regolith coarsening with depth, we found that the ratio of the estimated original bedrock block size to the deposit median grain size (D50) of the deposit was proportional to the potential energy of the landslide normalized to its bedrock strength. Thus, the studied landslides agree well with a published, simple fragmentation model, even if that model was calibrated on rock avalanches with larger volume and stronger bedrock than those featured in our dataset. Therefore, this scaling may serve for future modeling of grain size transfer from hillslopes to rivers, with the aim to better understanding landslide sediment evacuation and coupling to river erosional dynamics.
... The most general conventional sampling of surveying riverbed material is by field methods, such as grid-by-number or volumetric methods [5,6]. Field methods require huge labor, time, and cost because samples are collected on-site and measured or sieved. ...
Riverbed material has multiple functions in river ecosystems, such as habitats, feeding grounds, spawning grounds, and shelters for aquatic organisms, and particle size of riverbed material reflects the tractive force of the channel flow. Therefore, regular surveys of riverbed material are conducted for environmental protection and river flood control projects. The field method is the most conventional riverbed material survey. However, conventional surveys of particle size of riverbed material require much labor, time, and cost to collect material on site. Furthermore, its spatial representativeness is also a problem because of the limited survey area against a wide riverbank. As a further solution to these problems, in this study, we tried an automatic classification of riverbed conditions using aerial photography with an unmanned aerial vehicle (UAV) and image recognition with artificial intelligence (AI) to improve survey efficiency. Due to using AI for image processing, a large number of images can be handled regardless of whether they are of fine or coarse particles. We tried a classification of aerial riverbed images that have the difference of particle size characteristics with a convolutional neural network (CNN). GoogLeNet, Alexnet, VGG-16 and ResNet, the common pre-trained networks, were retrained to perform the new task with the 70 riverbed images using transfer learning. Among the networks tested, GoogleNet showed the best performance for this study. The overall accuracy of the image classification reached 95.4%. On the other hand, it was supposed that shadows of the gravels caused the error of the classification. The network retrained with the images taken in the uniform temporal period gives higher accuracy for classifying the images taken in the same period as the training data. The results suggest the potential of evaluating riverbed materials using aerial photography with UAV and image recognition with CNN.
... Since grains are not physically handled and only viewed top-down, tilting, different axes exposed, difficult lighting, partial burial, PURINTON AND BOOKHAGEN 10.1029/2021JF006260 21 of 29 and/or imbrication can all cause discrepancies in manual versus digital results, potentially propagating into percentile uncertainties. These effects have been recognized and examined extensively since the first photo-sieving attempts (e.g., Ibbeken & Schleyer, 1986;Kellerhals & Bray, 1971). Graham et al. (2010) provide detailed analysis of this and find that manual digital grain sizing (i.e., hand-clicking) tends to slightly over-estimate the percentiles of interest. ...
Mixed sand- and gravel-bed rivers record erosion, transport, and fining signals in their bedload size distributions. Thus, grain-size data are imperative for studying these processes. However, collecting hundreds to thousands of pebble measurements in steep and dynamic high-mountain river settings remains challenging. Using the recently published digital grain-sizing algorithm PebbleCounts, we were able to survey seven large (>= 1,000 m2) channel cross-sections and measure thousands to tens-of-thousands of grains per survey along a 100-km stretch of the trunk stream of the Toro Basin in Northwest Argentina. The study region traverses a steep topographic and environmental gradient on the eastern margin of the Central Andean Plateau. Careful counting and validation allows us to identify measurement errors and constrain percentile uncertainties using large sample sizes. In the coarse >= 2.5 cm fraction of bedload, only the uppermost size percentiles (>= 95th) vary significantly downstream, whereas the 50th and 84th percentiles show less variability. We note a relation between increases in these upper percentiles and along-channel junctions with large, steep tributaries. This signal is strongly influenced by lithology and geologic structures, and mixed with local hillslope input. In steep catchments like the Toro Basin, we suggest nonlinear relationships between geomorphic metrics and grain size, whereby the steepest parts of the landscape exert primary control on the upper grain-size percentiles. Thus, average or median metrics that do not apply weights or thresholds to steeper topography may be less predictive of grain-size distributions in such settings.
... where C i = (C i + C i+1 )/2 is the mean size in psi units of size fraction i, and b ′ = bln2 with b being the transformation coefficient. Using Kellerhals and Bray's (1971) (see Table 3 for coefficient b), the conversion from area-by-number to a volume-by-weight distribution (from AGS to sieving) can be obtained by letting b = 2. This conversion assumes that the cube model is appropriate, and that the sieve hole sizes correspond to the intermediate axis of the measured particles (Sulaiman et al., 2014). ...
The automated grain sizing technique (AGS) has been widely used to characterise grain size distribution of a channel bed. A handful number of literatures have been made available in portraying the wide range of AGS application for river and coastal studies. However, the accuracy of this technique is subject to further validation and verification. The accuracy of the AGS technique is lessened due to distortions of image, relief, or tilt. This paper discusses the consistency of the AGS technique at different ground sampling distances, and the implementation of correction factors to modify the grain size distribution (GSD) curve of the AGS technique on fine and coarse fractions. Through a discrepancy ratio test, the GSD curve from the AGS technique was compared with those of the conventional sieving and pebble-counting methods. It was observed that relief distortion did not have a significant impact on the GSD curve. However, textural presence in sediment particles led to ‘over-segmentation,’ which complicated the edge detection of an individual particle. The introduction of correction factors, using least square regression equation, was able to correct those errors by reducing and maintaining the discrepancy ratio to 0.688 – 1.283 for fine fractions, and 0.758 – 1.125 for coarse fractions.
... To ensure representative sampling of sediments of various grain sizes, the weight of the biggest clast in a sample must not to exceed a certain percentage of the gross weight of the entire sample. Existing quality criteria range between 0.1% and 5% (Kellerhals and Bray, 1971;Mosley and Tindale, 1985;Church et al., 1987). We used the milder 5% criterion recommended by Church et al. (1987) and Mosley and Tindale (1985) for samples including grains of the gravel and boulder fraction. ...
Sediment dynamics in river catchments are controlled by tectonics, climate, and biota effecting material production on hillslopes and transport of sediment from their source to the catchment outlet. Tectonics create topography and control erodibility of the bedrock material, whereas climate controls the efficiency of weathering and transporting processes. In contrast, the effects of biota (i.e. vegetation) are more ambiguous. Vegetation accelerates bio-chemical weathering and effects mass wasting through uprooting, trapping and stabilizing hillslope material. Furthermore, vegetation exerts a strong control on the grain size distribution of hillslope sediments and thus on sediment rooting through catchments.
In this study we compare grain size distributions of hillslope and channel sediments collected in four headwater catchments located in the Chilean Coastal Cordillera covering a strong bio-climatic gradient. 76 volumetric bulk samples were taken with grain sizes ranging from clay (d = 0.3 μm) to boulders (d > 300 mm).
Results show that the production of fine material is strongly dependent on bio-chemical weathering intensity and hence humidity. The coarse fraction in hillslope material and channel sediments increases from arid to sub-humid conditions, presumably reflecting a higher intensity of mass wasting processes. Channels show varying degrees of armouring reflecting nonselective and thus transport-limited conditions under arid climate that are contrasted by relatively well sorted and armoured channel sediments in the humid and vegetated catchments, indicating size selective and supply-limited transport conditions. Channel grain sizes generally show high similarity to hillslope grain sizes of the same catchment, revealing the strong dependency of channel sediments on hillslope supply, at least in headwater catchments, transferring a bio-climatic control from the hillslopes to the channel system.
... First, the wavelet-based optical Second, optical granulometry methods quantify the size of apparent axes of grains in the 74 image plane, where many grains may be overlapping. If a bulk (i.e. by mass or by volume) 75 sample size distribution is the information required, the Buscombe (2013) or similar method 76 can provide comparable grain size distributions to those derived using sieves or similar 77 methods usually only if the appropriate conversion of area-to mass-by-size is made, which 78 takes the form (Diplas and Sutherland, 1988;Kellerhals and Bray, 1971): 'poorly' gravels, which was determined using known grain-size distributions. Since this study 186 is fully reproducible using software described in section 5.4, the interested reader is 187 encouraged to explore different subjective groupings of the provided 409 sediment images 188 and of their own. ...
I describe a configurable machine-learning framework to estimate a suite of continuous and categorical sedimentological properties from photographic imagery of sediment, and to exemplify how machine learning can be a powerful and flexible tool for automated quantitative and qualitative measurements from remotely sensed imagery. The model is tested on a large dataset consisting of 400 images and associated detailed label data. The data are from a much wider sedimentological spectrum than previous optical granulometry studies, consisting of both well- and poorly sorted sediment, terrigenous, carbonate, and volcaniclastic sands and gravels and their mixtures, and grain sizes spanning over two orders of magnitude. I demonstrate the model framework by configuring it in several ways, to estimate two categories (describing grain shape and population, respectively) and nine numeric grain-size percentiles in pixels from a single input image. Grain size is then recovered using the physical size of a pixel. Finally, I demonstrate that the model can be configured and trained to estimate equivalent sieve diameters directly from image features, without the need for area-to-mass conversion formulas and without even knowing the scale of one pixel. Thus, it is the only optical granulometry method proposed to date that does not necessarily require image scaling. The flexibility of the model framework should facilitate numerous application in the spatio-temporal monitoring of the grain size distribution, shape, mineralogy and other quantities of interest, of sedimentary deposits as they evolve as well as other texture-based proxies extracted from remotely sensed imagery.
The sediment of alluvial riverbeds plays a significant role in river systems both in engineering and natural processes. However, the sediment composition can show great spatial and temporal heterogeneity, even on river reach scale, making it difficult to representatively sample and assess. Indeed, conventional sampling methods in such cases cannot describe well the variability of the bed surface texture due to the amount of energy and time they would require. In this paper, an attempt is made to overcome this issue introducing a novel image-based, Deep Learning algorithm and related field measurement methodology with potential for becoming a complementary technique for bed material samplings and significantly reducing the necessary resources. The algorithm was trained to recognise main sediment classes in videos that were taken underwater in a large river with mixed bed sediments, along cross-sections, using semantic segmentation. The method is fast, i.e., the videos of 300–400 meter long sections can be analysed within minutes, with very dense spatial sampling distribution. The goodness of the trained algorithm is evaluated mathematically and via intercomparison with other direct and indirect methods. Suggestions for performing proper field measurements are also given, furthermore, possibilities for combining the algorithm with other techniques are highlighted, briefly showcasing the multi-purpose of underwater videos for hydromorphological adaptation. The paper is to show the potential of underwater videography and Deep Learning through a case study.
This study reports on the physical properties and geochemistry of aeolian bedform grains along the Curiosity rover traverse in Gale crater from Vera Rubin Ridge to the Sands of Forvie (Sols 1902–2995), and includes comparisons to results made earlier in the mission. Volumetrically, <150 μm grains dominate active aeolian bedforms in the study area, similar to previous findings elsewhere at Gale crater and at other locations on Mars. Coarser grains, up to 2.9 mm long, are present on larger active bedforms. The larger 1–3 mm active grains commonly are reddish or whitish in color and irregular in shape, suggesting erosion of local bedrock as sources. One inactive megaripple had a surface of dust‐covered 2–15 mm grains, with smaller <150 μm grains between and within the bedform interior. Geochemical measurements show element concentrations vary according to position on the bedform, sand activity, and grain contributions from local bedrock. A strong positive correlation between Mg and Ni is identified on active bedform surfaces, with the highest Ni always corresponding to ripple crests where the coarsest gray and clear grains were commonly found. There is also a correlation between Ti and Cr for the majority of active sands, with the finer active sands in ripple troughs and sand patches having the greatest number of red grains and highest Cr concentrations. These results show the smaller scale physical properties and geochemistry of several types of aeolian bedforms on Mars formed under current and ancient environments.
Mass movement deposit grain‐size distributions (GSDs) record initiation, transport, and deposition mechanisms, and contribute to the rate at which sediment is exported from hillslopes to channels. Defining the GSD of a mass movement deposit is a significant challenge because they are often difficult to access, are heterogeneous in planform and with depth, contain grain sizes from clay (<63 μm) to boulders (> 1 m), and require considerable time to calculate accurately. There are numerous methods used to measure mass movement GSDs, but no single method alone can measure the entire range of grain sizes. This paper compares five common methods for determining mass movement deposit GSDs to assess how their accuracy may affect their applicability to different research areas. We applied an automated wavelet analysis (pyDGS), Wolman pebble counts, survey tape counts, manual photo counts and sieving to three different mass movement deposits (two debris flows, one rockslide) in Tredegar, Wales and the Longmen Shan, China. We found that pyDGS and survey tape counts produced comparable GSDs to sieving over a single order of magnitude. PyDGS required calibration to achieve accurate results, limiting its use for many applications. In Tredegar, Wolman pebble counts overestimated grain sizes in the lower 80% of the distribution relative to the other four methods used. We demonstrate that method choice can introduce significant uncertainties, particularly at the edges of the distributions such that D16 values differ by up to a factor of five. These methodological uncertainties limit GSD comparisons across studies particularly where these are used to infer processes within deposits. To minimise these challenges, the methods chosen should both be carefully reported and consistent with the research question.
We describe the first detailed reach-scale study of an incisional till-bed river. Our analysis focusses on boundary till characteristics, bare till patch features, annual erosion rates, bedform dimensions and spacing, and grain size distributions of bedforms. Results show that till exposures constitute a relatively small portion of the bed and that till erosion rates are relatively high compared to bedrock rivers, although highly variable between till patches and within patches. The bedforms are not well organized in terms of spacing and show high morphologic variability. The sediment forming the bed is poorly sorted, and grain sizes of the bedforms show high variability ranging from fines to large boulders, although gravel contribution to the alluvium is relatively small. We found evidence of some in situ and transport rounding of clasts. As expected, riffles and steps are coarser while glides and pools are finer-grained. Sedimentary stability metrics show that riffles are unstable, while pools and glides are more stable. These results indicate that the bedform morphology and sedimentology till-bedded rivers differ substantially from their alluvial and bedrock counterparts in a variety of ways. Consequently, we recommend that semi-alluvial rivers be differentiated from their alluvial and bedrock counterparts in future channel classifications. Such a practice will be useful for the river research and practitioners' community to gain the appropriate research tools needed for assessment, management, and restoration practices for these rivers.
Process-based classification of rivers is widely used for assessing the effects of past landscape disturbance, current biophysical conditions, and likely responses to future disturbance, including land management and restoration activities. This article reviews the purposes of geomorphic channel classification, the different types of classifications that have been developed, their use, compatibility, and popularity (i.e., why some classifications are used more than others), and concludes with a look at future needs and directions for channel classification.
We present a simple image analysis technique for measuring the sediment composition of the bed surface during flow with bedload transport in flume experiments. The bed was composed by a bimodal mixture of natural sediments and the flow conditions were such that both fractions were mobile. A camera placed 2.5 meters above the flume framed the central part and took pictures every minute, allowing for a high-resolution data acquisition. The two sediment fractions had different colors, thus we developed an image analysis technique based on color detection. This technique provided the areal fraction content of each fraction and the detection of patterns of sorted sediments, namely patches of each grain size, and it was applied in three laboratory experiments conducted to study the influence of lateral width constraint on channel morphology, sediment sorting, and bedload transport. The three experiments were carried out under constant feeding rate conditions, in particular a sediment supply equal to the transport capacity was provided, thereby matching dynamic equilibrium conditions. The image processing was implemented through a code written in Matlab and made available for other users.
A dearth of field measurements exists from sandy gravel-bed rivers. Laboratory experiments suggest that sediment mixtures with a large sand component are particularly efficient at transporting bedload downstream. To evaluate transport processes in these environments, we constructed a sediment monitoring research facility with the ability to continuously monitor sediment fluxes. The Arroyo de los Pinos is a gravel-bed ephemeral channel with a large (>30%) sand component in New Mexico, USA. This field station incorporates direct measurements of bedload flux using Reid-type slot samplers with high quality measurements of suspended sediment, water depth, water velocity, and grain size. Measurements are collected at a stable cross section constructed for long-term, consistent monitoring.
Instantaneous bedload flux is high compared to global averages, as high as 12 kg s⁻¹ m⁻¹. Suspended sediment is also high, peaking at 100,000 mg L⁻¹, with evidence of clockwise hysteresis. The Pinos nondimensional bedload flux rates are similar to those in other ephemeral channels, but the high sand content allows equal mobility at relatively shallow water depths and high rates of bedload transport during hydrograph recession. Bedload flux responded hysteretically, higher at equal shear stresses during stage rise. This is likely caused by an inadequate use of the depth-slope product for calculation of bed shear, steeper friction slopes during the rising limb, or both.
These first monitored sediment transporting flash flood events reveal a channel that is very active during short periods of time. On average, the Arroyo de los Pinos flows 12–24 h during 3–5 events every year. However, in these brief periods, the channel is hyper-efficient at transporting bedload and suspended sediment at rates that are orders of magnitude higher than perennial counterparts.
Azufral volcano is an active composite volcano located in SW Colombia, whose edifice is mostly formed by pyroclastic deposits and lava flows. Recent road works in the region, NE of the volcano, revealed unknown deposits, one of them related to a debris avalanche event. This deposit, herein named San Roque debris avalanche deposit, outcrops along a 13 km section in the road Túquerres - Samaniego. Using this long exposure, the deposit was studied through sedimentological, grain size, microtextural and mineralogical analyses. The deposit is massive, varicolored, poorly sorted, with a predominance of gravel and sand-sized clasts; it contains angular megaclasts some of them with jigsaw cracks, and exhibits injections, faults, and horsts-and-graben structures. Surface microtextures in some clasts evidence collisions and shearing interactions among them, while mineralogy evidences the collapsed material involved in the avalanche. Based on stratigraphic relationships, the collapse is bracketed between 0.58 Ma and ∼18 ka. The San Roque debris avalanche deposit covers an area of 56.01 km², has an estimated volume of 0.51 km³, an approximated runout of 11 km, and an apparent friction coefficient H/L of 0.109; these parameters are indicative of a large and very mobile flow. Its main transport mechanism responded to a dynamic disintegration of the larger clasts and a subsequent interaction by contact or collision (inertial-granular flow model), although some zones moved in a cushioned way remaining coherent over long distances (multi-shear model).
Cluster microforms in bed armoring have important implications for bedload transport. In this study, two sets of flume experiments were carried out to study the evolution of the cluster microforms in the bed armoring and their influences on the bedload transport. Dynamically stable armor layers were formed by applying a staged hydrograph on a mobile sediment bed consisted of poorly sorted sand and gravel, thus enabling the investigation of the cluster microforms associated with different flow characteristics simulated in these experiments. The size and the planimetric positions of the large particles on the bed surface were extracted using photogrammetric technique, allowing analysis of surface grain size distribution and particle clustering pattern of the bed. The results showed a tendency towards a coarsened and clustered bed surface, which is likely responsible for the declining bedload transport rate. The cluster microforms evolved with increasing bed shear stress in the following sequence: flat bed - pebble cluster - line cluster - heap cluster - reticulate structure. We found a bed armoring statistics that can partially explain the declining bedload transport rate under constant discharge. Nonetheless, additional parameters are needed to simulate the fluctuation in the bedload transport rate under changing flow conditions.
The deposits of volcanic debris avalanches (VDAs) contain diagnostic features that distinguish them from those of other landslides. In this chapter, we summarize the sedimentary characteristics and the different (litho-)facies described over the past four decades, and how findings from individual case studies can be adapted as globally applicable sedimentological tools. A plethora of descriptive terms and partially conflicting definitions emerged in the ever-growing literature on VDA deposits (VDADs). These we summarize and make recommendations for future use. Different facies models that were developed at different volcanoes might point to unique emplacement conditions (e.g. dry versus wet; confined versus unconfined) and, if confirmed, the apparent ‘conflict’ of terminology might help identify the paleo-settings of ancient VDAs. General observations of large unsaturated landslides of different origin show that preservation of source stratigraphy, (mega-)clasts, jigsaw-fractured clasts, and incorporation of runout path material are common features. Their unique composition, grain sizes, and abundance of matrix sets VDADs apart from deposits of large rockslides and debris flows. The latter can be associated with VDAs, and whether they formed syn- or post-VDAD emplacement is reflected in forensic evidence within the depositional sequences. Recent case studies illustrate the advances in analytical techniques and in understanding the processes of debris avalanche transport and deposition forty years after the eruption and lateral collapse of Mount St. Helens volcano.
This article proposes a new methodological approach to measure and map the size of coarse clasts on a land surface from photographs. This method is based on the use of the Mask Regional Convolutional Neural Network (R-CNN) deep learning algorithm, which allows the instance segmentation of objects after an initial training on manually labeled data. The algorithm is capable of identifying and classifying objects present in an image at the pixel scale, without human intervention, in a matter of seconds. This work demonstrates that it is possible to train the model to detect non-overlapping coarse sediments on scaled images, in order to extract their individual size and morphological characteristics with high efficiency (R 2 = 0.98; Root Mean Square Error (RMSE) = 3.9 mm). It is then possible to measure element size profiles over a sedimentary body, as it was done on the pebble beach of Etretat (Normandy, France) in order to monitor the granulometric spatial variability before and after a storm. Applied at a larger scale using Unmanned Aerial Vehicle (UAV) derived ortho-images, the method allows the accurate characterization and high-resolution mapping of the surface coarse sediment size, as it was performed on the two pebble beaches of Etretat (D50 = 5.99 cm) and Hautot-sur-Mer (D50 = 7.44 cm) (Normandy, France). Validation results show a very satisfying overall representativity (R 2 = 0.45 and 0.75; RMSE = 6.8 mm and 9.3 mm at Etretat and Hautot-sur-Mer, respectively), while the method remains fast, easy to apply and low-cost, although the method remains limited by the image resolution (objects need to be longer than 4 cm), and could still be improved in several ways, for instance by adding more manually labeled data to the training dataset, and by considering more accurate methods than the ellipse fitting for measuring the particle sizes.
The vertical position of the streambed‐water boundary fluctuates during the course of sediment transport episodes, due to particle entrainment/deposition and bedform migration, amongst other hydraulic and bedload mechanisms. These vertical oscillations define a topmost stratum of the streambed i.e. the ‘active layer or active depth’, which usually represents the main source of particles entrained during long and high‐magnitude bedload transport episodes. The vertical extent of this layer is hence a capital parameter for the quantification of bedload volumes and a major driver of stream ecology in gravel‐bed rivers. However, knowledge on how the active depth scales to flow strength and the nature of the different controls on the relation between the flow strength and the active depth is still scarce. In this paper we present a meta‐analysis over active depth data coming from ~130 transport episodes extracted from a series of published field studies. We also incorporate our own field data for the rivers Ebro and Muga (unpublished), both in the Iberian Peninsula. We explore the database searching for the influence of flow strength, grain‐size, streambed mobility, and channel morphology on the vertical extent of the active layer. A multivariate statistical analysis (stepwise multiple regression) confirms that the set of selected variables explains a significant amount of variance in the compiled variables. The analysis shows a positive scaling between active depth and flow strength. We have also identified some links between the active depth and particle travel distances. However, these relations are also largely modulated by other fluvial drivers, such as the grain‐size of the bed surface and the dominant channel macro‐bedforms, with remarkable differences between plane‐bed, step‐pool and riffle‐pool channels.
The structure of a mobile bed in a laboratory channel composed of sand and gravel (D50=8 mm) was characterized over a series of experiments with steady flows from a low flow, where only the finer fractions of the bed material were in motion to flows in which most bed material grain sizes were in motion. In each experiment, sediment transport rates were observed to initially be greater than the long-term mean rates, and fluctuations in transport rate decreased in period as bed shear stresses were increased. The bed surface median grain size increased with bed shear stress, while the sand fraction of the bed material organized into longitudinally extended corridors, which persisted as flow and transport rates were increased. The presence of the sand corridors was reflected by changes in the probability density function of the bed-surface elevation standard deviation evaluated at the grain-scale. The formation and organization of these corridors may have a strong influence on sand and gravel transport in channels with mixed sand and gravel bed material.
The aim of this paper is to achieve a quantitative assessment of rockfall protection forest efficiency at regional scale, considering site specific forest, morphological and lithological parameters. At first, a semi-automatic GIS-based method, integrated with a multi-scenario 3D-rockfall model realized by using the simulation code HY-STONE, was used to map protection forests of Regione Lombardia (central Italian Alps). For each different forest type, a rockfall protective efficiency was assessed by using empirical (energy line angle) and modelling (HY-STONE) approaches. The empirical approach shows an increase of the energy line angle value from about 36° for the bare slopes to over 40° in forested slopes, with a value ranging from 37° to 44° for different forests types. The modelling approach is based on a new efficiency index EEI ranging from 0 (minimum efficiency, equal to no forest condition) to 1 (maximum efficiency): the efficiency of different forest types ranges from 0.08 to 0.98 by using average values of the controlling parameters. To modulate the efficiency in each single forest at regional scale, a set of parametric simulations was performed to evaluate the effects of controlling parameters. The parametric simulations show that block volume, slope gradient, DBH (diameter at breast height) and forest density are the most important parameters at controlling the efficiency. These parameters were used within a multiple linear regression function to associate a protection efficiency to each specific protection forest in the regional map. This allows to discriminate quantitatively the individual forests according to their actual efficiency. Most of the protection forest area (46%) shows an efficiency greater than 0.50 and only the 24% of the total covered area shows a value lower than 0.25.
The surface of the streambed in gravel-bed rivers is commonly coarser than the underlying bed material. This surface coarsening, or ‘armouring’, is usually described by means of the ratio between surface and subsurface grain-size metrics (the ‘armour ratio’). Such surface coarsening is typical of river reaches that are degrading due to a deficit in sediment supply (e.g. gravel-bed reaches below dams or lakes), but non-degrading gravel-bed streams may also exhibit some degree of armouring in relation to specific hydrological patterns. For instance, selective transport during the recession limbs of long lasting floods may coarsen the bed more significantly than flash floods. Consequently, regional differences in bed coarsening should exist, reflecting in turn the variability in sediment and water regimes. In this paper, we explore the trends linking armour ratios to sediment supply, taking into account the differences in hydrological context. We based our analysis on a large data set of bedload and grain size measurements from 49 natural gravel-bed streams and four flume experiments compiled from the scientific literature. Our main outcome documents how the balances between sediment yields and transport capacities have a quantifiable reflection on the armour ratios measured in the field: we report statistically significant correlations between bedload fluxes and surface grain-size, and an asymptotic rise in armour ratios with the decline of sediment supply. Hydrological controls are also observed. Additionally, the trends observed in the field data are comparable to those previously documented in flume experiments with varying sediment feed. In this regard, different kinds of bedforms and particle arrangements have been commonly described with progressive reductions in sediment inputs and the subsequent coarsening of the streambed. Hence, armour ratios serve as a proxy for the general organization of the streambed of gravel-bed streams, and our results quantify this streambed adjustment to the dominant sediment regime.
Most grain size monitoring is still being conducted by manual sampling in the field, which is time consuming and has low spatial representation. Due to new remote sensing methods, some limitations have been partly overcome, but methodological progress is still needed for large rivers as well as in underwater conditions. In this paper, we tested the reliability of two methods along the Old Rhine River (France/Germany) to estimate the grain size distribution (GSD) in above-water conditions: (i) a low-cost terrestrial photosieving method based on an automatic procedure using Digital Grain Size (DGS) software and (ii) an airborne LiDAR topo-bathymetric survey. We also tested the ability of terrestrial photosieving to estimate the GSD in underwater conditions. Field pebble counts were performed to compare and calibrate both methods. The results showed that the automatic procedure of terrestrial photosieving is a reliable method to estimate the GSD of sediment patches in both above-water and underwater conditions with clean substrates. Sensitivity analyses showed that environmental conditions, including solar lighting conditions and petrographic variability, significantly influence the GSD from the automatic procedure in above-water conditions. The presence of biofilm in underwater conditions significantly altered the GSD estimation using the automatic procedure, but the proposed manual procedure overcame this problem. The airborne LiDAR topographic survey is an accurate method to estimate the GSD of above-water bedforms and is able to generate grain size maps. The combination of terrestrial photosieving and airborne topographic LiDAR methods is adapted to assess the GSD over several kilometers long reaches of large rivers.
The volume and grain‐size of sediment supplied from catchments fundamentally control basin stratigraphy. Despite their importance, few studies have constrained sediment budgets and grain‐size exported into an active rift at the basin scale. Here, we used the Corinth Rift as a natural laboratory to quantify the controls on sediment export within an active rift. In the field, we measured the hydraulic geometries, surface grain‐sizes of channel bars and full‐weighted grain‐size distributions of river sediment at the mouths of 47 catchments draining the rift (constituting 83% of the areal extent). Results show that the sediment grain‐size increases westward along the southern coast of the Gulf of Corinth, with the coarse‐fraction grain‐sizes (84th percentile of weighted grain‐size distribution) ranging from 20 to 110 mm. We find that the median and coarse‐fraction of the sieved grain‐size distribution are primarily controlled by bedrock lithology, with late Quaternary uplift rates exerting a secondary control. Our results indicate that grain‐size export is primarily controlled by the input grain‐size within the catchment and subsequent abrasion during fluvial transport, both quantities that are sensitive to catchment lithology. We also demonstrate that the median and coarse‐fraction of the grain‐size distribution are predominantly transported in bedload; however, typical sand‐grade particles are transported as suspended load at bankfull conditions, suggesting disparate source‐to‐sink transit timescales for sand and gravel. Finally, we derive both a full Holocene sediment budget and a grain‐size specific bedload discharged into the Gulf of Corinth using the grain‐size measurements and previously‐published estimates of sediment fluxes and volumes. Results show that the bedload sediment budget is primarily comprised (~79%) of pebble to cobble grade (0.475–16 cm). Our results suggest that the grain‐size of sediment export at the rift scale is particularly sensitive to catchment lithology and fluvial mophodynamics, which complicates our ability to make direct inferences of tectonic and palaeoenvironmental forcing from local stratigraphic characteristics.
This study uses the 2015 ‘Storm Frank’ flood on the River Dee, Aberdeenshire, to assess the impact of extreme events on river dynamics. The Storm Frank flood (>200 year recurrence interval) caused significant local morphological change that was concentrated in the middle portion of the 140 km long river and overall net degradation that primarily occurred through lateral adjustment processes. Although the flood did not cause widespread change in channel planform, morphological change at the reach scale (<1 km) was significant. Bank scour resulted in channel expansion and lateral migration as well as widespread aggradation on existing gravel beds. The HEC-RAS and CAESAR–Lisflood models were used to determine the impact of morphological changes from the Storm Frank flood on a series of future hypothetical floods. The results show that inundation is highly influenced by the degree of morphological change for moderate floods, but not for high magnitude ones. In-channel scour and bank erosion can lead to an increase in channel capacity, thereby decreasing inundation. Conversely, where conveyance capacity is decreased by aggradation, flood risk inherently increases. The impact of these changes was great for a five-year return period flood, but minimal for a magnitude flood comparable to that of Storm Frank. Our modelling results also reveal that the inundation model is sensitive to the grain size and channel bed roughness input parameters, as these parameters impact flow discharge and flood hydraulics. Accurate determination of sediment parameters and degree of morphological change is therefore critical in flooding modelling and flood hazard management.
Supplementary material : Peak discharge and rainfall during the 2015 Storm Frank storm, parameters used in the hydrological model CAESAR–Lisflood and sediment budget statistics of each DEM of difference threshold are available at: https://doi.org/10.6084/m9.figshare.c.4847946
Thematic collection: This article is part of the Early Career Research collection available at: https://www.lyellcollection.org/cc/SJG-early-career-research
Global climatic changes and local disturbances may alter composition and distribution of spring invertebrates in mountains. In this study, we compared the effects of elevation, landscape and local characteristics on spring biodiversity. At 16 springs (from 570 to 1650 m a.s.l.) in The Chartreuse Range (French Alps) benthic, vegetation, and interstitial habitats were sampled in summer for aquatic invertebrate assemblage structure (abundances, richness, reproduction and dispersal traits, functional feeding groups). Assemblages were related to geographic location (elevation), landscape characteristics and local variables. Elevation was the major driver of the fauna: taxonomic richness and the percentage of scrapers decreased with elevation, while the proportion of predators and species with an asexual reproduction increased with elevation. The landscape characteristics around the spring had a weak influence on the benthic taxonomic richness which increased with the percentage of forest and decreased with the proportion of grassland. Finally, the habitat characteristics had no effect on taxonomic richness, but partially control the abundances of benthic assemblages and sediment-feeder organisms that both decreased with increasing sediment grain size. Current and future changes in the temperature patterns would affect alpine spring fauna, but disturbance of the local characteristics of springs must not be neglected.
I describe a configurable machine‐learning framework to estimate a suite of continuous and categorical sedimentological properties from photographic imagery of sediment, and to exemplify how machine learning can be a powerful and flexible tool for automated quantitative and qualitative measurements from remotely sensed imagery. The model is tested on a dataset consisting of 409 images and associated detailed label data. The data are from a much wider sedimentological spectrum than previous optical granulometry studies, consisting of both well‐ and poorly sorted sediment, terrigenous, carbonate, and volcaniclastic sands and gravels and their mixtures, and grain sizes spanning over two orders of magnitude. I demonstrate the model framework by configuring it in several ways, to estimate two categories (describing grain shape and population, respectively) and nine numeric grain‐size percentiles in pixels from a single input image. Grain size is then recovered using the physical size of a pixel. Finally, I demonstrate that the model can be configured and trained to estimate equivalent sieve diameters directly from image features, without the need for area‐to‐mass conversion formulas and without even knowing the scale of one pixel. Thus, it is the only optical granulometry method proposed to date that does not necessarily require image scaling. The flexibility of the model framework should facilitate numerous application in the spatio‐temporal monitoring of the grain size distribution, shape, mineralogy and other quantities of interest, of sedimentary deposits as they evolve as well as other texture‐based proxies extracted from remotely sensed imagery.
Laboratory flume experiments were conducted in order to identify the temporal evolution of bed surface grain size properties by unsteady flows under zero sediment feeding conditions. Purposely designed sediment beds were composed by well-sorted gravel and sand, for each case, detailed data of the bed surface composition were collected. It is found that for all cases, the final sediment bed in the whole reach always shows a coarsening feature, and the coarsening process is much more developed near the inlet boundary. Further, the time variation of the percents of gravel and sand are observed increase or decrease alternately during the experiment and finally close to their initial percents. Besides, the experimental results show that there is no direct correlation between the flow unsteadiness and the coarsening degree of the final bed.
Grain-size distributions are a key geomorphic metric of gravel-bed rivers. Traditional measurement methods include manual counting or photo sieving, but these are achievable only at the 1–10 m² scale. With the advent of drones and increasingly high-resolution cameras, we can now generate orthoimagery over hectares at millimeter to centimeter resolution. These scales, along with the complexity of high-mountain rivers, necessitate different approaches for photo sieving. As opposed to other image segmentation methods that use a watershed approach, our open-source algorithm, PebbleCounts, relies on k-means clustering in the spatial and spectral domain and rapid manual selection of well-delineated grains. This improves grain-size estimates for complex riverbed imagery, without post-processing. We also develop a fully automated method, PebbleCountsAuto, that relies on edge detection and filtering suspect grains, without the k-means clustering or manual selection steps. The algorithms are tested in controlled indoor conditions on three arrays of pebbles and then applied to 12 × 1 m² orthomosaic clips of high-energy mountain rivers collected with a camera-on-mast setup (akin to a low-flying drone). A 20-pixel b-axis length lower truncation is necessary for attaining accurate grain-size distributions. For the k-means PebbleCounts approach, average percentile bias and precision are 0.03 and 0.09 ψ, respectively, for ∼1.16 mm pixel⁻¹ images, and 0.07 and 0.05 ψ for one 0.32 mm pixel⁻¹ image. The automatic approach has higher bias and precision of 0.13 and 0.15 ψ, respectively, for ∼1.16 mm pixel⁻¹ images, but similar values of −0.06 and 0.05 ψ for one 0.32 mm pixel⁻¹ image. For the automatic approach, only at best 70 % of the grains are correct identifications, and typically around 50 %. PebbleCounts operates most effectively at the 1 m² patch scale, where it can be applied in ∼5–10 min on many patches to acquire accurate grain-size data over 10–100 m² areas. These data can be used to validate PebbleCountsAuto, which may be applied at the scale of entire survey sites (10²–10⁴ m²). We synthesize results and recommend best practices for image collection, orthomosaic generation, and grain-size measurement using both algorithms.
The study of grain size distribution of gravels is an important and challenging issue in stratigraphy and morphology, especially in the field of automated measurement. It largely reduces many manual processes and time consumption. Using digital image processing method to extract effective information from digital images is a more advanced method. In this study, a digital image method using automated multiple superpixel segmentation and merging is proposed for extraction of the grain size data. It adds grain size estimation and improved merging rules and repeated execution rules to the superpixel segmentation and merging. It has made great improvements in many respects, especially in the accuracy of edge segmentation and measurement. Compared with manual measurement and other image processing ways, the method proposed in this paper is an efficient approach for precisely measuring the grain size distribution of riverbed material.
The
grain-size distribution of ancient alluvial systems is commonly determined
from surface samples of vertically exposed sections of gravel deposits. This
method relies on the hypothesis that the grain-size distribution obtained
from a vertical cross section is equivalent to that of the riverbed. Such an
hypothesis implies first that the sediments are uniform in size in the river
bed, and second that the sampling method implemented on a vertical section
leads to a grain-size distribution equivalent to the bulk one. Here, we
report a field test of this hypothesis on granulometric samples collected in
an active, gravel-bedded, braided stream: the Urumqi River in China. We
compare data from volumetric samples of a trench excavated in an active
thread and from surface counts performed on the trench vertical faces. Based
on this data set, we show that the grain-size distributions obtained from all
the samples are similar and that the deposit is uniform at the scale of the
river active layer, a layer extending from the surface to a depth of
approximately 10 times the size of the largest clasts. As a consequence, the
grid-by-number method implemented vertically leads to a grain-size
distribution equivalent to the one obtained by a bulk volumetric sampling.
This study thus brings support to the hypothesis that vertical surface counts
provide an accurate characterization of the grain-size distribution of
paleo-braided rivers.
Among the most urgent challenges in future climate change scenarios is accurately predicting the magnitude to which precipitation extremes will intensify. Analogous changes have been reported for an episode of millennial-scale 5 °C warming, termed the Palaeocene-Eocene Thermal Maximum (PETM; 56 Ma), providing independent constraints on hydrological response to global warming. However, quantifying hydrologic extremes during geologic global warming analogs has proven difficult. Here we show that water discharge increased by at least 1.35 and potentially up to 14 times during the early phase of the PETM in northern Spain. We base these estimates on analyses of channel dimensions, sediment grain size, and palaeochannel gradients across the early PETM, which is regionally marked by an abrupt transition from overbank palaeosol deposits to conglomeratic fluvial sequences. We infer that extreme floods and channel mobility quickly denuded surrounding soil-mantled landscapes, plausibly enhanced by regional vegetation decline, and exported enormous quantities of terrigenous material towards the ocean. These results support hypotheses that extreme rainfall events and associated risks of flooding increase with global warming at similar, but potentially at much higher, magnitudes than currently predicted.
Scour and fill patterns at East St. Louis Creek, Colorado, were investigated via repeat, detailed surveys of the channel bed at 11 cross sections during the 1995 snowmelt season. Spatial variability was remarkably high, with significant differences in cross section scour and fill patterns over distances as short as 0.5 m. Most sites had small net changes in bed elevation, both daily and over the entire runoff season. The data and observations indicate the presence of small pulses of fine material that are temporarily deposited on top of the channel pavement in wider areas of the channel and near woody debris complexes. Scour and fill are primarily limited to the finer material of such pulses. ANOVA analysis indicates that although discharge was important in predicting changes in bed elevation, the relationship between discharge and bed mobility is complicated by the effects of local channel morphology and a slight hysteresis. Regression analysis shows that variations in channel width determine where finer sediments are deposited, and therefore the locations of greater change in bed elevation. The proximity of morainal ridges and boulders to the channel edge locally influence the channel width and also the distribution of woody debris complexes. Results of this study suggest that the channel morphology and sediment transport along some reaches of small, high-gradient streams in watersheds with a glacial history may not respond as substantially to changes in discharge characteristic as do other types of alluvial channels.
This paper examines the potential to reverse bed degradation trends in the Drôme River (southeast France), a prealpine gravel-bed stream, by assessing coarse sediment supply in tributary subbasins. Two approaches are developed. First, a Principal Component Analysis was performed on 24 subwatersheds to characterize the most productive ones. Second, fieldwork and aerial photographs provide detailed assessment for two mountain streams, one characterized by abundant bedload (the Esconavette Creek), the other characterized by 30 yr of inactivity (the Barnavette Creek). Active tributaries (with developed active gravel bars) have been discriminated according to variables describing watershed sediment storage and production potential. Variables describing human intervention on erosion and vegetation cover appear to exert secondary influences on geomorphic activity in a context of general geomorphic stabilization. Declining activity associates features such as bed incision, channel narrowing, and vegetation development in riparian buffer strips in the two mountain streams studied. They are more frequent on the Barnavette Creek, the inactivity of which is partly explained by smaller lateral inputs from valley-filling.
Laboratory flume experiments were done to investigate bed load sediment transport by both steady and unsteady flows in a degrading channel. The bed, respectively composed of uniform sand, uniform gravel, or sand-gravel mixtures, always undergoes bulk degradation. It is found that both uniform and non-uniform bed load transport is enhanced greatly by unsteady flows as compared to their volume-equivalent steady flows. This enhancement effect is evaluated by means of an enhancement factor, which is shown to be larger with a coarser bed and lower discharges. Also, the fractional transport rates of gravel and sand in non-uniform sand-gravel mixtures are compared with their uniform counterparts under both steady and unsteady flows. The sand is found to be able to greatly promote the transport of gravel, whilst the gravel considerably hinders the transport of sand. Particularly, the promoting and hindering impacts are more pronounced at lower discharges and tend to be weakened by flow unsteadiness.
This paper presents basic tests to develop an airborne photogrammetric methodology that derives grain size characteristics of gravel bed rivers. The data acquisition was done using a lightweight action cam and a hand-held digital single lens reflex camera. Image processing comprised the structure from motion technique and multiview-stereo algorithms to obtain digital elevation models of non-cohesive gravel beds. Laboratory results indicate that the method accuracy is about four to six times lower than laser-scan data when based on action cam data. The accuracy of digital elevation models computed via photos taken by the reflex camera is almost of the same range as the laser data. Field experiments were done to test the performance of image based gravelometry against manual surface sampling. For this application the action cam was mounted to a low-cost quadrocopter, while the reflex camera was operated by hand. Results indicate that this combination has a high potential to generate data from which characteristic grain size parameters can be estimated.
Predictions of river bed substrate cover are required for various purposes including delineating management zones, linking with ecological status and assessing river rehabilitation options. Three contrasting methods were tested for predicting the proportion of river bed covered by seven different substrate categories: generalised linear models (GLMs), machine learning regression models (random forest), and a summed normal distribution model (SND) which incorporates distribution of predictors and substrate covers throughout the modelling framework. Various predictors representing climate, geomorphology, land cover and geology were derived from existing environmental databases to generate predictive models. Model performance was assessed through a cross-validated comparison with substrate samples collected from 229 river sites distributed across New Zealand. Model performance for 10-fold cross-validated predictions showed that the SND model performed best in predicting the proportions of riverbed covered by bedrock, boulder, cobble and fine gravel categories. Random forest models performed best in predicting coarse gravel, sand and mud plus vegetation proportions. Therefore, combined random forest and SND methods were used for estimating substrate cover proportions at unsampled sites across New Zealand. Texture analysis of predicted substrate cover consistently showed downstream fining of sediment size. The national predictions of substrate cover proportions are key descriptors that can be linked with a wide range of national scale applications for ecological assessment of New Zealand Rivers. The techniques developed and tested are applicable to other locations but it is notable that relatively poor performance in regional cross-validation tests shows that transferability of substrate models to locations with no calibration data is challenging.
The formation of a large boulder delta complex found in the Båldakatj area, Swedish Lapland, is related to the catastrophic drainage of ice-dammed lakes. The interpretation of the sedimentological characteristics of these boulder deposits has resulted in the recognition of at least two phases in their development. An initial deposition of material by debris flows was subsequently followed by erosion and deposition by fluvial processes. The resulting reworking of the material was not uniform throughout the area. Two major lobes of boulder deposits can be recognized. These lobes have been exposed to different processes, as interpreted by differences in morphological and sedimentological characteristics, such as gradient, distal slope, channel pattern, scour marks, and the size, shape, roundness, imbrication and sorting of particles. Scour marks around large boulders occur frequently in the study area. They proved to be excellent flow direction indicators. The literature on similar boulder deposits, formed by flood flows or debris flows, in other regions or environments, has been reviewed and the results compiled for comparative purposes.
In mountainous rivers, large relatively immobile grains partly control the local and reach-averaged flow hydraulics and sediment fluxes. When the flow depth is similar to the size of these grains (low relative submergence), heterogeneous flow structures and plunging flow cause spatial distributions of bed surface elevations, textures, and sedimentation rates. To explore how the bed surface responds to these flow variations we conducted a set of experiments in which we varied the relative submergence of staggered hemispheres (simulated large boulders) between runs. All experiments had the same average sediment transport capacity, upstream sediment supply, and initial bed thickness and grain size distribution. We combined our laboratory measurements with a 3D flow model to obtain the detailed flow structure around the hemispheres. The local bed shear stress field displayed substantial variability and controlled the bedload transport rates and direction in which sediment moved. The divergence in bed shear stress caused by the hemispheres promoted size-selective bedload deposition, which formed patches of coarse sediment upstream of the hemisphere. Sediment deposition caused a decrease in local bed shear stress, which combined with the coarser grain size, enhanced the stability of this patch. The region downstream of the hemispheres was largely controlled by a recirculation zone and had little to no change in grain size, bed elevation, and bed shear stress. The formation, development and stability of sediment patches in mountain streams is controlled by the bed shear stress divergence and magnitude and direction of the local bed shear stress field.
ResearchGate has not been able to resolve any references for this publication.