River Suspended Sediment Concentration Research Articles

Enhanced river suspended sediment concentration identification via multimodal video image, optical flow, and water temperature data fusion

Monitoring suspended sediment concentration (SSC) in rivers is pivotal for water quality management and sustainable river ecosystem development. However, achieving continuous and precise SSC monitoring is fraught with challenges, including low automation, lengthy measurement processes, and high cost. This study proposes an innovative approach for SSC identification in rivers using multimodal data fusion. We developed a robust model by harnessing colour features from video images, motion characteristics from the Lucas–Kanade (LK) optical flow method, and temperature data. By integrating ResNet with a mixed density network (MDN), our method fused the image and optical flow fields, and temperature data to enhance accuracy and reliability. Validated at a hydropower station in the Xinjiang Uygur Autonomous Region, China, the results demonstrated that while the image field alone offers a baseline level of SSC identification, it experiences local errors under specific conditions. The incorporation of optical flow and water temperature information enhanced model robustness, particularly when coupling the image and optical flow fields, yielding a Nash–Sutcliffe efficiency (NSE) of 0.91. Further enhancement was observed with the combined use of all three data types, attaining an NSE of 0.93. This integrated approach offers a more accurate SSC identification solution, enabling non-contact, low-cost measurements, facilitating remote online monitoring, and supporting water resource management and river water–sediment element monitoring.

Comparison between Hyperspectral and Multispectral Retrievals of Suspended Sediment Concentration in Rivers

Remote sensing (RS) is often employed to estimate suspended sediment concentration (SSC) in rivers, and the availability of hyperspectral imagery enhances the effectiveness of RS-based water quality monitoring due to its high spectral resolution. Yet, the necessity of hyperspectral imagery for SSC estimation in rivers has not been fully validated. This study thus compares the performance of hyperspectral RS with that of multispectral RS by conducting field-scale experiments in shallow rivers. In the field experiments, we measured radiance from a water body mixed with suspended sediments using a drone-mounted hyperspectral sensor, with the sediment and riverbed types considered as controlling factors. We retrieved the SSC from UAV imagery using an optimal band ratio analysis, which successfully estimated SSC distributions in the sand bed conditions with both multispectral and hyperspectral data. In the vegetated bed conditions, meanwhile, the prediction accuracy decreased significantly due to the temporally varying bottom reflectance associated with the random movement of vegetation caused by near-bed turbulence. This is because temporally inhomogeneous bottom reflectance distorts the relationship between the SSC and total reflectance. Nevertheless, the hyperspectral imaging exhibited better prediction accuracy than the multispectral imaging, effectively extracting optimal spectral bands sensitive to back-scattered reflectance from sediments while constraining the bottom reflectance caused by the vegetation-covered bed.

Improved adaptive neuro-fuzzy inference system with bacterial foraging optimization algorithm for suspended sediment concentration estimation

The accuracy of predicting river-suspended sediment concentration (SSC) is crucial for evaluating the functional lifespan of reservoirs, analyzing river geomorphological evolution, and assessing riverbed stability. In this study, we aim to develop new models for SSC prediction at two hydrological stations near Puerto Rico, USA, by integrating the bacterial foraging optimization algorithm and adaptive neural fuzzy inference network (ANFIS). The models comprise ANFIS with grid partition (ANFIS-GP), ANFIS with subtractive clustering (ANFIS-SC), and ANFIS with fuzzy c-means clustering (ANFIS-FCM). Additionally, we employ an artificial neural network (ANN) and the sediment rating curve (SRC) for predicting daily series data of flow discharge-suspended sediment concentration (SSC). Different scenarios are considered based on varying input and output variables, leading to predictions for four distinct scenarios. At the Rio Valenciano Station, the MRSE values for ANFIS-BFO, ANFIS-FCM, ANFIS-GP, ANFIS-SC, ANN, and SRC are 2.2172, 2.5389, 2.6627, 2.7549, 2.7994, and 3.7882, respectively. For the Quebrada Blanca Station, the MRSE values for ANFIS-BFO, ANFIS-FCM, ANFIS-SC, ANFIS-GP, ANN, and SRC are 0.8295, 0.8664, 0.8964, 0.9110, 0.9684, and 1.6742, respectively. It can be inferred that ANFIS-BFO exhibits superior prediction results compared to all other models. Furthermore, ANFIS-SC and ANFIS-FCM demonstrate slightly better prediction performance than ANFIS-GP. In comparison to ANN, ANFIS-GP, ANFIS-SC, and ANFIS-FCM exhibit slightly superior prediction performance.

Rapid changes to global river suspended sediment flux by humans.

Rivers support indispensable ecological functions and human health and infrastructure. Yet limited river sampling hinders our understanding of consequential changes to river systems. Satellite-based estimates of suspended sediment concentration and flux for 414 major rivers reveal widespread global change that is directly attributable to human activity in the past half-century. Sediment trapping by dams in the global hydrologic north has contributed to global sediment flux declines to 49% of pre-dam conditions. Recently, intensive land-use change in the global hydrologic south has increased erosion, with river suspended sediment concentration on average 41 ± 7% greater than in the 1980s. This north-south divergence has rapidly reconfigured global patterns in sediment flux to the oceans, with the dominant sources of suspended sediment shifting from Asia to South America.

Physical controls and a priori estimation of raising land surface elevation across the southwestern Bangladesh delta using tidal river management

Abstract. The Ganges–Brahmaputra–Meghna delta in Bangladesh is one of the largest and most densely populated deltas in the world and is threatened by relative sea level rise (RSLR). Renewed sediment deposition through tidal river management (TRM), a controlled flooding with dike breach, inside the lowest parts of the delta polders (so-called beels) can potentially counterbalance the RSLR. The potential of TRM application in different beels across southwestern Bangladesh has been estimated previously but requires further exploration. Neither the seasonal and spatial variations in physical drivers nor the non-linear character of physical drivers and several sensitive parameters for sediment deposition have been taken into account so far. We used a 2D morphodynamic model to explore the physical controls of the following five parameters on the total sediment deposition inside the beels during TRM: river tidal range (TR), river suspended sediment concentration (SSC), inundation depth (ID), width of the inlet (IW), and surface area of the beel (BA). Our model results indicate that these five parameters and their interactions are significant for sediment deposition per day (SPD), where SSC and BA have a high impact, TR and ID have a moderate impact, and IW has a low impact on sediment deposition. Non-linear regression models (NLMs) were developed using the results of 2D models to quantify how sediment deposition inside the beels depends on these parameters. The NLMs have an average coefficient of determination of 0.74 to 0.77. Application of the NLMs to 234 beels in southwestern Bangladesh indicates that TRM operation in beels located closer to the sea will retain more sediment as a result of decreasing SSC further inland. Lower average land surface elevation is one of the reasons that the beels in the western part retain more sediment. Smaller beels have a higher potential to raise the land surface elevation due to the non-linear increase in sediment deposition per day (SPD) with beel area. Compartmentalization of larger beels may increase their potential to raise the land surface elevation. Thus, the length of time of the TRM application in a cyclic order will need to vary across the delta from 1 to multiple years to counterbalance RSLR, depending on the current beel land surface elevation and local TRM sediment accumulation rates. We found that operating TRM only during the monsoon season is sufficient to raise the land surface in 96 % and 80 % of all beels by more than 3 and 5 times the yearly RSLR, respectively. Applying TRM only seasonally offers huge advantages as to keeping the land available for agriculture during the rest of the year. The methodology presented here, applying regression models based on 2D morphodynamic modeling, may be used for the low-lying sinking deltas around the world to provide an a priori estimation of sediment deposition from controlled flooding to counterbalance RSLR.

Application of soft-computing techniques in forecasting sediment load and concentration

ABSTRACT This paper investigates conventional and soft-computing methods for the estimation of suspended sediment concentration (SSC) and load (SSL) in rivers. Frequently used methods of sediment rate curve (SRC) and multi-nonlinear regression, and soft-computing methods of multi-layer perceptron, multi-linear regression and adaptive neuro-fuzzy inference system are implemented using various hydrological and hydraulic parameters for the Little Kickapoo Creek Watershed, Illinois, USA. All methods performed equally well in the estimation of SSL, without any noticeable outperformance from any from the methods. However, the application of soft-computing methods decreased SSC estimation errors considerably as compared to the results of SRC. The results are significant in the way they reconcile traditionally used hydrological parameters into the soft-computing methods. Overall, soft-computing methods are recommended for the estimation of SSC in rivers because of their reasonably better performance and ease of implementation.

Estimating Suspended Sediment Concentrations from River Discharge Data for Reconstructing Gaps of Information of Long-Term Variability Studies

Suspended sediment rating-curves are low cost and reliable tools used all around the world to estimate river suspended sediment concentrations (SSC) based on either linear or non-linear regression with a second variable, such as the river discharge. The aim of this paper is to undertake an evaluation of four different suspended sediment rating-curves for three turbid large river tributaries flowing into the largest choked coastal lagoon of the world, a very turbid system. Statistical parameters such as Nash–Sutcliffe efficiency coefficient (NSE), percent of bias (PBIAS) and a standardized root-mean-square error (RMSE), referred to as RSR (RMSE-observations standard deviation ratio) were used to calibrate and validate the suspended sediment rating-curves. Results indicated that for all tributaries, the non-linear approach yielded the best correlations and proved to be an effective tool to estimate the SSC from river flow data. The tested curves show low bias and high accuracy for monthly resolution. However, for higher temporal resolution, and therefore variability, an ad hoc data acquisition to capture extreme rating-curve values is required to reliably fill gaps of information for both performing modeling approaches and setting monitoring efforts for long-term variability studies.

Sediment transport in a restored, river-influenced Pacific Northwest estuary

Predicting the success of future investments in coastal and estuarine ecosystem restorations is limited by scarce data quantifying sediment budgets and transport processes of prior restorations. This study provides detailed analyses of the hydrodynamics and sediment fluxes of a recently restored U.S. Pacific Northwest estuary, a 61 ha former agricultural area near the mouth of the Stillaguamish River in Washington, USA. Water level, flow velocity, and suspended-sediment concentration (SSC) were measured between 21 March 2014 and 1 June 2015 at breaches excavated in the former flood-protection levee to determine transport patterns and the net sediment budget of the restoration area. SSC within the restoration area was primarily controlled by SSC variability of the nearby main stem Stillaguamish, but coastal processes also played a major role in sediment delivery. Fluvial sediment loading was dominated by runoff events associated with rainfall that lasted hours to a few days. Additionally, the 22 March 2014 SR 530 (Oso) landslide elevated sediment supply to the restoration area and coastal region for several weeks, indicating the importance of distal geomorphic events to coastal sediment budgets in small mountainous river systems. Sediment fluxes were controlled by river SSC and tidal dynamics, which set the quantity of water transported into the restoration area. Peak water discharge at the restoration area was about 12% of the river discharge, and peak sediment flux at the restoration area was about 5% of the river sediment discharge, although net sediment import was <1% of the total river load. Although sediment was imported to the restoration area, and inferred rates of accretion appear sufficient to keep pace with present rates of local sea-level rise, full recovery is challenged by significant lost grade from historical subsidence and will likely take decades to centuries. These results have implications for estuary restoration planning globally and indicate the importance of understanding coupled fluvial–coastal processes.

Estimating Impacts of Dam Development and Landscape Changes on Suspended Sediment Concentrations in the Mekong River Basin’s 3S Tributaries

The Mekong River Basin (MRB) is undergoing rapid dam development, which is altering the river suspended sediment concentration (SSC). In this study, we used satellite remote sensing records spanning 31 years to detect SSC changes (SSC prediction r2=0.78, RMSE=21.2 mg/L) due to dam development. We focused on the 3S basin of the MRB. We also used satellite data on nighttime lights, which reflect human settlement patterns, and land cover to explain SSC patterns. Our technique allowed for quantification of SSC changes due to dam construction (e.g., +120 mg/L near basin outlet), reservoir sediment trapping (e.g., −108 mg/L), deforestation, and human settlement (e.g., +117 mg/L near impacts). Our technique also demonstrated how the SSC of the 3S rivers compared to that of the Mekong mainstem over time (e.g., from ∼13% to 100% greater). Our comprehensive analyses of SSC records with dam development indicate that SSC changes will continue with ongoing dam and landscape development in the MRB. From a hydrologic perspective, SSC monitoring will be imperative for effective sediment and water management. Our satellite-based approach answers critical sediment needs of improved monitoring and adaptive management throughout the MRB and other global locations for practitioners who are engaged in real-world management of their river basins.

Remote Sensing Monitoring on Spatial Differentiation of Suspended Sediment Concentration in a River-Lake System Based on Sentinel-2 MSI Imaging:A Case for Shengjin Lake and Connected Yangtze River Section in Anhui Province

Carrying out monitoring of suspended sediment concentration in river and lake systems is of great significance for understanding the laws of sediment transport in water and formulating policies on water environmental control. Taking Shengjin Lake and the connected Yangtze river section in Anhui province as the study area, band reflectance of a Sentinel-2 MSI sensor is simulated according to field spectral datasets, and the retrieval model is established by statistical regression from the synchronized suspended sediment concentration measurements. Then, the retrieved results from 28 scene MSI images during 2017-2019 are used to analyze the spatiotemporal variation of suspended sediment concentration in rivers and lakes, and the influence of water level variation on their spatial differentiation is also discussed. The results show that:① The retrieval model established by the ratio of the sixth band to the third band of the MSI sensor is suitable for high-turbidity water type, with high accuracy (R2=0.863, RMSE=22.211 mg·L-1). ② Spatially, the suspended sediment concentration near the lake entrances, northwestern parts of the upper and middle lake areas, and the lower lake is relatively higher, and that of Shengjin Lake is lower than that of the Yangtze River overall except for in summer. Temporally, the suspended sediment concentration in Shengjin Lake is relatively lower in summer and higher in other seasons, while the connected Yangtze River section exhibits the opposite intra-annual variation. ③ The water level, which is caused by the connectivity of rivers and lakes under the influence of the sluice, is the key factor affecting the spatial differentiation of suspended sediment concentration in the river and lake system. The suspended sediment concentration in Shengjin Lake contributes to the Yangtze River in dry and normal water periods, and that in the normal water period is more significant. In contrast, during the flood period, the correlation between suspended sediment concentration in the Yangtze River and that in Shengjin Lake is not obvious.

Applying ADCPs for Long‐Term Monitoring of SSC in Rivers

AbstractThe flow rate and the suspended sediment concentration (SSC) of two different rivers draining into the Adriatic Sea basin, the Secchia in Italy and the Devoll in Albania, were analyzed by processing the data collected in a 245‐day period at monitoring stations equipped with side‐looking acoustic Doppler current profilers (H‐ADCPs). SSC was determined as a by‐product of the echo profiles along the horizontal aligned acoustic beams emitted by H‐ADCPs. For the first time, the effect of organic matter other than a change of inorganic particles size distribution was evaluated as a possible reason for backscatter and attenuation variations during hydrological events. This reduced average deviations between acoustically inferred and sampled concentrations from 1 order of magnitude to 20% of actual values. The improvement required the measuring of attenuation to backscatter ratio (ABR) in addition to sound attenuation. The validation interval covered 3 orders of magnitude from to g L . In this paper the potential of the ABR method using H‐ADCP to continuously monitor suspended sediment fluxes is tested in two different rivers, enabling single peak flood analysis and reliable assessment of sediment budget across a river cross section. The advantages and disadvantages of this method are presented and discussed.

The potential of hybrid evolutionary fuzzy intelligence model for suspended sediment concentration prediction

Providing a robust and reliable prediction model for suspended sediment concentration (SSC) is an essential task for several environmental and geomorphology prospective including water quality, river bed engineering sustainability, and aquatic habitats. In this research, a novel hybrid intelligence approach based on evolutionary fuzzy (EF) approach is developed to predict river suspended sediment concentration. To demonstrate the modeling application, one of the highly affected rivers located in the north-western part of California is selected as a case study (i.e., Eel River). Eel River is considered as one of the most polluted river due to the streamside land sliding, owing to the highly stochastic water river discharge. Thus, the predictive model is constructed using discharge information as it is the main trigger for the SSC amount. The prediction conducted on different locations of the stream (i.e., up-stream and down-stream stations). Three different well-established integrative fuzzy models are developed for the validation purpose including adaptive neuro-fuzzy inference system coupled with subtractive clustering (ANFIS-SC), grid partition (ANFIS-GP), and fuzzy c-means (ANFIS-FCM) models. The predictive models evaluated based on several numerical indicators and two-dimension graphical diagram (i.e., Taylor diagram) that vividly exhibits the observed and predicted values. The attained results evidenced the predictability of the EF model for the SSC over the other models. The discharge information provided an excellent input attributes for the predictive models. In summary, the discovered model showed an outstanding data-intelligence model for the environmental perspective and particularly for Eel River. The methodology is highly qualified to be implemented as a real-time prediction model that can provide a brilliant approach for the river engineering sustainability.

What factors affect the suspended sediment concentrations in rivers? A study of the upper Warta River (Central Poland)

AbstractThe suspended sediment concentration (SSC) of the Warta River was analysed using data collected at the Sieradz gauging station (Central Poland) during the period of 1961–1980. The aim of the study was to characterize the trend in the suspended sediment transported over this multi‐annual period and search for possible correlations between the suspended sediment values and the discharge and thermal seasonality factors. This study also investigated whether the SSC is affected by anthropogenic factors. The SSC in the river water increased over the analysed 20‐year period. It was caused mainly by the training works in the Warta valley upstream of Sieradz and the opening of the Bełchatów lignite mine. No direct relationship between the SSC and fluctuations in the discharge was noted. A connection existed between the SSC and ice phenomena on the river. The river ice breakup was often coincident with increases in SSC; however, the SSC increased rapidly during the freezing of the river. There was also a positive correlation between the SSC and the temperature of the water during the summer half‐year, which resulted from the growth of phytoplankton. This process was supported by the general warming trend observed in the river water and by an increase in the quantity of nutrients noted since the early 1970s. It is worth emphasizing that the relationships established between the SSC and the studied factors are not always unequivocal and repetitive. It follows that, as a rule, the SSC is influenced by more factors than are actually considered in this study.

Organic sediment pulses impact rivers across multiple levels of ecological organization

AbstractSedimentation is a pervasive environmental pressure affecting rivers globally. Headwaters draining catchments rich in organic soils (i.e., peat) are particularly vulnerable to enhanced sedimentation caused by land management and environmental change, yet many of the ecological consequences of peat deposition are poorly understood. We conducted a before‐after‐control‐impact experiment in two rivers draining blanket peatland in Northern England to test the effect of sediment inputs on water quality, macroinvertebrate drift, macroinvertebrate community structure, and ecosystem metabolism. Sediment addition increased concentrations of dissolved organic carbon, total oxidised nitrogen and suspended sediment concentration in rivers, and intensified the total drift of macroinvertebrates particularly at night. By contrast, the abundance and richness of benthic macroinvertebrates were unaffected, except for declines in Coleoptera abundance in one river. The gross primary production of both rivers was strongly suppressed as the benthos was smothered by sediment. Community respiration also declined, albeit by different extents in the two rivers. Our experiment revealed that short‐term pulses of organic sediment in rivers can have broad effects on water quality and biota, from influences on the dispersal of individual organisms to the modification of ecosystem processes. Organic sediments therefore warrant further examination, to include longer observation periods and more sites. It is particularly important to clarify the extent to which impacts extend from peatland streams into larger rivers downstream. Such studies are necessary to inform global management efforts to restore the integrity of river ecosystems under a range of water and biodiversity policy mechanisms.

Relationship between turbidity and suspended sediment concentration from a small hydrographic basin in Santa Maria (Rio Grande do Sul, Brazil)

ABSTRACTSediment production has been the focus of many studies due to the problems caused by the alterations of the natural environment which are driven by the urbanization process. In Santa Maria, Brazil, urban area expansion and population growth have caused changes in the environment, such as in the structure and natural form of the land, in drainage channels, quality and quantity of water resources, soil sealing and intensive agricultural cultivation. Consequently, studies are necessary to quantify the sediment in rivers for a better understanding and also to develop tools in order to minimize impacts to the environment. This study assesses the relationship between turbidity and suspended sediment concentration (SSC) from a small hydrographic basin located in Santa Maria (Brazil). The methodology was based on monitoring the suspended sediment transport using automated turbidity registrations and suspended sediment samples, which were collected by automatic ISCO sampler. To do so, four field measurements were carried out from 2 September to 6 October 2014. The research has resulted in a good correlation between SSCs and turbidity (T), r = 0.8602, (P < .001), when the SSC data were analysed as a function of T it was observed that follow a model SSC=0.833T1.2266 (P < .001), with a coefficient of determination R2 = 0.740. Therefore, it may be a good alternative to generate continuous records of SSCs in rivers as a tool for water management in a hydrographic basin.

Conversion to drip irrigated agriculture may offset historic anthropogenic and wildfire contributions to sediment production

This study is an investigation into the roles of wildfire and changing agricultural practices in controlling the inter-decadal scale trends of suspended sediment production from semi-arid mountainous rivers. In the test case, a decreasing trend in suspended sediment concentrations was found in the lower Salinas River, California between 1967 and 2011. Event to decadal scale patterns in sediment production in the Salinas River have been found to be largely controlled by antecedent hydrologic conditions. Decreasing suspended sediment concentrations over the last 15years of the record departed from those expected from climatic/hydrologic forcing. Sediment production from the mountainous headwaters of the central California Coast Ranges is known to be dominated by the interaction of wildfire and large rainfall/runoff events, including the Arroyo Seco, an ~700km2 subbasin of the Salinas River. However, the decreasing trend in Salinas River suspended sediment concentrations run contrary to increases in the watershed's effective burn area over time. The sediment source area of the Salinas River is an order of magnitude larger than that of the Arroyo Seco, and includes a more complicated mosaic of land cover and land use. The departure from hydrologic forcings on suspended sediment concentration patterns was found to coincide with a rapid conversion of irrigation practices from sprinkler and furrow to subsurface drip irrigation. Changes in agricultural operations appear to have decreased sediment supply to the Salinas River over the late 20th to early 21st centuries, obscuring the influence of wildfire on suspended sediment production.

Evaluation of data driven models for river suspended sediment concentration modeling

Using eight-year data series from hydrometric stations located in Arkansas, Delaware and Idaho (USA), the ability of artificial neural network (ANN) and support vector regression (SVR) models to forecast/estimate daily suspended sediment concentrations ([SS]d) was evaluated and compared to that of traditional multiple linear regression (MLR) and sediment rating curve (SRC) models. Three different ANN model algorithms were tested [gradient descent, conjugate gradient and Broyden–Fletcher–Goldfarb–Shanno (BFGS)], along with four different SVR model kernels [linear, polynomial, sigmoid and Radial Basis Function (RBF)]. The reliability of the applied models was evaluated based on the statistical performance criteria of root mean square error (RMSE), Pearson’s correlation coefficient (PCC) and Nash–Sutcliffe model efficiency coefficient (NSE). Based on RMSE values, and averaged across the three hydrometric stations, the ANN and SVR models showed, respectively, 23% and 18% improvements in forecasting and 18% and 15% improvements in estimation over traditional models. The use of the BFGS training algorithm for ANN, and the RBF kernel function for SVR models are recommended as useful options for simulating hydrological phenomena.

3. Quantifying Sediment Deposition Patterns of Lake Underflows Using a Novel Underflow Sediment Trap

Lake underflow deposition is an important limnological process which greatly affects sediment deposition patterns and lake varve formation. Currently no feasible, cost‐effective device or method has been regularly utilized to quantify sediment deposition patterns. This study utilizes a novel underflow trap which was deployed at two locations at the bottom of a High Arctic lake subject to seasonal river inflow. It was found that a peak in lake bottom temperature departures, lake bottom turbidity, and river suspended sediment concentration are strongly associated with peak underflow deposition events. Furthermore, evidence shows that deposition amounts are greatly reduced as underflow distance increases. One year was also found to show a clear lag in deposition patterns between two distant stations. This method of quantifying underflow deposition is useful for determining deposition patterns over time and space. This knowledge is useful in monitoring the changes in the lake bottom waters, and for aiding in the reconstruction of past sediment deposition patterns.

Large-scale dam removal on the Elwha River, Washington, USA: Source-to-sink sediment budget and synthesis

Understanding landscape responses to sediment supply changes constitutes a fundamental part of many problems in geomorphology, but opportunities to study such processes at field scales are rare. The phased removal of two large dams on the Elwha River, Washington, exposed 21±3millionm3, or ~30milliontonnes (t), of sediment that had been deposited in the two former reservoirs, allowing a comprehensive investigation of watershed and coastal responses to a substantial increase in sediment supply. Here we provide a source-to-sink sediment budget of this sediment release during the first two years of the project (September 2011–September 2013) and synthesize the geomorphic changes that occurred to downstream fluvial and coastal landforms. Owing to the phased removal of each dam, the release of sediment to the river was a function of the amount of dam structure removed, the progradation of reservoir delta sediments, exposure of more cohesive lakebed sediment, and the hydrologic conditions of the river. The greatest downstream geomorphic effects were observed after water bodies of both reservoirs were fully drained and fine (silt and clay) and coarse (sand and gravel) sediments were spilling past the former dam sites. After both dams were spilling fine and coarse sediments, river suspended-sediment concentrations were commonly several thousand mg/L with ~50% sand during moderate and high river flow. At the same time, a sand and gravel sediment wave dispersed down the river channel, filling channel pools and floodplain channels, aggrading much of the river channel by ~1m, reducing river channel sediment grain sizes by ~16-fold, and depositing ~2.2millionm3 of sand and gravel on the seafloor offshore of the river mouth. The total sediment budget during the first two years revealed that the vast majority (~90%) of the sediment released from the former reservoirs to the river passed through the fluvial system and was discharged to the coastal waters, where slightly less than half of the sediment was deposited in the river-mouth delta. Although most of the measured fluvial and coastal deposition was sand-sized and coarser (>0.063mm), significant mud deposition was observed in and around the mainstem river channel and on the seafloor. Woody debris, ranging from millimeter-size particles to old-growth trees and stumps, was also introduced to fluvial and coastal landforms during the dam removals. At the end of our two-year study, Elwha Dam was completely removed, Glines Canyon Dam had been 75% removed (full removal was completed 2014), and ~65% of the combined reservoir sediment masses—including ~8Mt of fine-grained and ~12Mt of coarse-grained sediment—remained within the former reservoirs. Reservoir sediment will continue to be released to the Elwha River following our two-year study owing to a ~16m base level drop during the final removal of Glines Canyon Dam and to erosion from floods with larger magnitudes than occurred during our study. Comparisons with a geomorphic synthesis of small dam removals suggest that the rate of sediment erosion as a percent of storage was greater in the Elwha River during the first two years of the project than in the other systems. Comparisons with other Pacific Northwest dam removals suggest that these steep, high-energy rivers have enough stream power to export volumes of sediment deposited over several decades in only months to a few years. These results should assist with predicting and characterizing landscape responses to future dam removals and other perturbations to fluvial and coastal sediment budgets.

Neural networks approached for modelling river suspended sediment concentration due to tropical storms

Artificial neural networks are one of the advanced technologies employed in hydrology modelling. This paper investigates the potential of two algorithm networks, the feed forward backpropagation (BP) and generalized regression neural network (GRNN) in comparison with the classical regression for modelling the event-based suspended sediment concentration at Jiasian diversion weir in Southern Taiwan. For this study, the hourly time series data comprised of water discharge, turbidity and suspended sediment concentration during the storm events in the year of 2002 are taken into account in the models. The statistical performances comparison showed that both BP and GRNN are superior to the classical regression in the weir sediment modelling. Additionally, the turbidity was found to be a dominant input variable over the water discharge for suspended sediment concentration estimation. Statistically, both neural network models can be successfully applied for the event-based suspended sediment concentration modelling in the weir studied herein when few data are available.