Estimation of suspended sediment concentration in rivers using acoustic methods

Acoustic Doppler current meters (ADV, ADCP, and ADP) are widely used in water systems to measure flow velocities and velocity profiles. Although these meters are designed for flow velocity measurements, they can also provide information defining the quantity of particulate matter in the water, after appropriate calibration. When an acoustic instrument is calibrated for a water system, no additional sensor is needed to measure suspended sediment concentration (SSC). This provides the simultaneous measurements of velocity and concentration required for most sediment transport studies. The performance of acoustic Doppler current meters for measuring SSC was investigated in different studies where signal-to-noise ratio (SNR) and suspended sediment concentration were related using different formulations. However, these studies were each limited to a single study site where neither the effect of particle size nor the effect of temperature was investigated. In this study, different parameters that affect the performance of an ADV for the prediction of SSC are investigated. In order to investigate the reliability of an ADV for SSC measurements in different environments, flow and SSC measurements were made in different streams located in the Aegean region of Turkey having different soil types. Soil samples were collected from all measuring stations and particle size analysis was conducted by mechanical means. Multivariate analysis was utilized to investigate the effect of soil type and water temperature on the measurements. Statistical analysis indicates that SNR readings ob tained from the ADV are affected by water temperature and particle size distribution of the soil, as expected, and a prediction model is presented relating SNR readings to SSC mea surements where both water temperature and sediment characteristics type are incorporated into the model. The coefficients of the suggested model were obtained using the multivariate anal ysis. Effect of high turbidity conditions on ADV performance was also investigated during and after rain events.     

Suspended particulate matter estimates using optical and acoustic sensors: application in Nestos River plume (Thracian Sea,North Aegean Sea)

The present study investigates the use of combined methods of optical and acoustic sensors, in collaboration with direct in situ measurements, for the calibration and validation of a model transforming acoustic backscatter intensity series into suspended particulate matter (SPM) concentration datasets. The model follows previously elaborated techniques, placing particular attention to the parameterization of the acoustic absorption index as a function of water physical properties. Results were obtained from the annual deployment (during 2007–2008) of an upward-facing acoustic Doppler current profiler (ADCP) (307 kHz), equipped with a Wave Array, and an optical backscatter sensor (OBS), at the bottom of Thassos Passage near Nestos River plume (Thracian Sea, Northern Greece). The OBS was calibrated through linear regression, using 2007 and 2012 field sampling data, exhibiting an error of 13–14 % due to chlorophyll presence. The ADCP signal was calibrated through simultaneous measurements of backscatter intensity and turbidity profiles. Harmonic analysis on the model-produced SPM concentrations explained the tidal influence on their variability, especially during the summer. Empirical orthogonal functions analysis revealed the impact of waves and wave-induced currents on SPM variability. Finally, Nestos River sediment load was found uncorrelated to the SPM change in Thassos Passage, due to the dispersal and sediment deposition near the river mouth.     

Multilayer perceptron neural network for flow prediction

Artificial neural networks (ANNs) have proven to be a tool for characterizing, modeling and predicting many of the non-linear hydrological processes such as rainfall-runoff, groundwater evaluation or simulation of water quality. After proper training they are able to generate satisfactory predictive results for many of these processes. In this paper they have been used to predict 1 or 2 days ahead the average and maximum daily flow of a river in a small forest headwaters in northwestern Spain. The inputs used were the flow and climate data (precipitation, temperature, relative humidity, solar radiation and wind speed) as recorded in the basin between 2003 and 2008. Climatic data have been utilized in a disaggregated form by considering each one as an input variable in ANN(1), or in an aggregated form by its use in the calculation of evapotranspiration and using this as input variable in ANN(2). Both ANN(1) and ANN(2), after being trained with the data for the period 2003-2007, have provided a good fit between estimated and observed data, with R(2) values exceeding 0.95. Subsequently, its operation has been verified making use of the data for the year 2008. The correlation coefficients obtained between the data estimated by ANNs and those observed were in all cases superior to 0.85, confirming the capacity of ANNs as a model for predicting average and maximum daily flow 1 or 2 days in advance.     

Artificial neural network modeling of dissolved oxygen in reservoir

The water quality of reservoirs is one of the key factors in the operation and water quality management of reservoirs. Dissolved oxygen (DO) in water column is essential for microorganisms and a significant indicator of the state of aquatic ecosystems. In this study, two artificial neural network (ANN) models including back propagation neural network (BPNN) and adaptive neural-based fuzzy inference system (ANFIS) approaches and multilinear regression (MLR) model were developed to estimate the DO concentration in the Feitsui Reservoir of northern Taiwan. The input variables of the neural network are determined as water temperature, pH, conductivity, turbidity, suspended solids, total hardness, total alkalinity, and ammonium nitrogen. The performance of the ANN models and MLR model was assessed through the mean absolute error, root mean square error, and correlation coefficient computed from the measured and model-simulated DO values. The results reveal that ANN estimation performances were superior to those of MLR. Comparing to the BPNN and ANFIS models through the performance criteria, the ANFIS model is better than the BPNN model for predicting the DO values. Study results show that the neural network particularly using ANFIS model is able to predict the DO concentrations with reasonable accuracy, suggesting that the neural network is a valuable tool for reservoir management in Taiwan.     

An observational study of the carrying capacity of suspended sediment during a storm event

We measured the in situ suspended sediment concentration (SSC) and hydrodynamics (waves, currents, and sea level) concurrently during a storm event using self-recording instruments at offshore of the Shandong Peninsula in the Yellow Sea. We analyzed the temporal variation in suspended sediment carrying capacity and its correlation with wave, current, and water-level conditions. There was about 40% increase in SSC during the storm event. A 3-4-h lag was observed between the peak of wave height and SSC. The SSC increased in a fluctuating pattern up to the peak then decreased rapidly. This process was positively correlated with wave height and duration. The maximum SSC was 524.3 mg/l, which is about 10 times of that under normal weather conditions. This peak was observed after of a series of 1.8 m waves. The increased suspended sediments were the fine particles resuspended by the storm waves from seabed near the observation spot and transported by northward ebb currents from Shidao Bay.     

Modeling of non-point source pollution in a Mediterranean drainage basin

SWAT ver. 2000 was used to predict hydrographs, and sediment, nitrate and total phosphorus loadings from a 1349 km² mountainous/agricultural watershed in Northern Greece. The model was calibrated and verified using continuous meteorological data from eight stations within the drainage area, and runoff, sediment and nutrient concentrations measured at nine stations located within the main tributaries of the watershed, for the time period from May 1st, 1998 to January 31st, 2000. Model validation methodology and resulting input parameters appropriate for Mediterranean drainage basins are presented. Predicted by the model hydrographs, sedimentographs and pollutographs are plotted against observed values and show good agreement. Model performance is evaluated using the root mean square error computation and scattergrams of predicted versus observed data. The validated model is also used to test the effectiveness of three alternative cropping scenarios in reducing nutrient loadings from the agricultural part of the watershed. The study showed that this model, if properly validated, can be used effectively in testing management scenarios in Mediterranean drainage basins.     

Modeling of Hydrodynamics and Cohesive Sediment Processes in an Estuarine System: Study Case in Danshui River

The Danshui River estuarine system is the largest estuarine system in northern Taiwan and is formed by the confluence of Tahan Stream, Hsintien Stream, and Keelung River. A comprehensive one-dimensional (1-D) model was used to model the hydrodynamics and cohesive sediment transport in this branched river estuarine system. The applied unsteady model uses advection/dispersion equation to model the cohesive sediment transport. The erosion and deposition processes are modeled as source/sink terms. The equations are solved numerically using an implicit finite difference scheme. Water surface elevation and longitudinal velocity time series were used to calibrate and verify the hydrodynamics of the system. To calibrate and verify the mixing process, the salinity time series was used and the dispersion coefficient of the advection/dispersion equation was determined. The cohesive sediment module was calibrated by comparing the simulated and field measured sediment concentration data and the erosion coefficient of the system was determined. A minimum mean absolute error of 4.22 mg/L was obtained and the snapshots of model results and field measurements showed a reasonable agreement. Our modeling showed that a 1-D model is capable of simulating the hydrodynamics and sediment processes in this estuary and the sediment concentration has a local maximum at the limit of salinity intrusion. Furthermore, it was indicated that for Q ₅₀ (the flow which is equaled or exceeded 50% times), the turbidity maximum location during neap tide is about 1 km closer to the mouth compared to that during spring tide. It was found that deposition is the dominant sediment transport process in the river during spring–neap periods. It was shown that, while sediment concentration at the upstream depends on the river discharge, the concentration in the downstream is not a function of river discharge.     

Artificial neural network modeling of dissolved oxygen in the Heihe River, Northwestern China

Identification and quantification of dissolved oxygen (DO) profiles of river is one of the primary concerns for water resources managers. In this research, an artificial neural network (ANN) was developed to simulate the DO concentrations in the Heihe River, Northwestern China. A three-layer back-propagation ANN was used with the Bayesian regularization training algorithm. The input variables of the neural network were pH, electrical conductivity, chloride (Cl^?), calcium (Ca²⁺), total alkalinity, total hardness, nitrate nitrogen (NO₃-N), and ammonical nitrogen (NH₄-N). The ANN structure with 14 hidden neurons obtained the best selection. By making comparison between the results of the ANN model and the measured data on the basis of correlation coefficient (r) and root mean square error (RMSE), a good model-fitting DO values indicated the effectiveness of neural network model. It is found that the coefficient of correlation (r) values for the training, validation, and test sets were 0.9654, 0.9841, and 0.9680, respectively, and the respective values of RMSE for the training, validation, and test sets were 0.4272, 0.3667, and 0.4570, respectively. Sensitivity analysis was used to determine the influence of input variables on the dependent variable. The most effective inputs were determined as pH, NO₃-N, NH₄-N, and Ca²⁺. Cl^? was found to be least effective variables on the proposed model. The identified ANN model can be used to simulate the water quality parameters.     

Spatial and temporal variation in suspended sediment, organic matter, and turbidity in a Minnesota prairie river: implications for TMDLs

The Minnesota River Basin (MRB), situated in the prairie pothole region of the Upper Midwest, contributes excessive sediment and nutrient loads to the Upper Mississippi River. Over 330 stream channels in the MRB are listed as impaired by the Minnesota Pollution Control Agency, with turbidity levels exceeding water quality standards in much of the basin. Addressing turbidity impairment requires an understanding of pollutant sources that drive turbidity, which was the focus of this study. Suspended volatile solids (SVS), total suspended solids (TSS), and turbidity were measured over two sampling seasons at ten monitoring stations in Elm Creek, a turbidity impaired tributary in the MRB. Turbidity levels exceeded the Minnesota standard of 25 nephelometric units in 73% of Elm Creek samples. Turbidity and TSS were correlated (r ²?=?0.76), yet they varied with discharge and season. High levels of turbidity occurred during periods of high stream flow (May–June) because of excessive suspended inorganic sediment from watershed runoff, stream bank, and channel contributions. Both turbidity and TSS increased exponentially downstream with increasing stream power, bank height, and bluff erosion. However, organic matter discharged from wetlands and eutrophic lakes elevated SVS levels and stream turbidity in late summer when flows were low. SVS concentrations reached maxima at lake outlets (50 mg/l) in August. Relying on turbidity measurements alone fails to identify the cause of water quality impairment whether from suspended inorganic sediment or organic matter. Therefore, developing mitigation measures requires monitoring of both TSS and SVS from upstream to downstream reaches.     

Distribution of heavy metals in water and sediments of selected sites of Yamuna river (India)

The concentration of Cd, Co, Cr, Cu, Fe, Mn, Ni, Pb, and Zn metals in water and sediments of Yamuna river were determined by atomic absorption spectrophotometry in the year 1981. The data showed that there was considerable variation in the concentration of elements from one sampling station to the other which may be due to the variation in the quality of industrial and sewage wasters being added to the river at different sampling stations. The sediment samples collected from different sampling stations were also analysed for calcium carbonate, organic matter, potassium, and phosphorus.     

Chemometric Analysis of Surface Water Quality Data: Case Study of the Gorganrud River Basin, Iran

Canonical correlation analysis (CCA), principal component analysis (PCA), and principal factor analysis (PFA) have been adopted to provide ease of understanding: interpretation of a large complex data set in the Gorganrud River monitoring networks, evaluation of the temporal and spatial variations of water quality, and finally identification of monitoring stations and parameters which are most important in assessing annual variations of water quality in the river. In accomplishing the research, 11 surface water quality data related to both of physical and chemical parameters have been collected from seven monitoring stations from 1996 to 2002. In general, our results from CCA method indicated strong relationship between physical and chemical parameters in the Gorganrud River. In addition, analyzing data through the PCA and PFA techniques revealed that all monitoring stations are important in explaining the annual variation of data set. From the point of view of the degree of importance of parameters contributing to water quality variations, further investigations by running two scenarios (rotated factor correlation coefficient value equal to 0.95 and 0.90 for the first and second scenarios, respectively) showed that the important parameters in one season may not be important for another season. For example, unlike in summer, water temperature, total suspended solids, total phosphorous, and nitrate parameters were important, electrical conductivity, and turbidity parameters had been realized as important parameters in spring through the first scenario.     

Transition index maps for urban growth simulation: application of artificial neural networks,weight of evidence and fuzzy multi-criteria evaluation

Transition index maps (TIMs) are key products in urban growth simulation models. However, their operationalization is still conflicting. Our aim was to compare the prediction accuracy of three TIM-based spatially explicit land cover change (LCC) models in the mega city of Mumbai, India. These LCC models include two data-driven approaches, namely artificial neural networks (ANNs) and weight of evidence (WOE), and one knowledge-based approach which integrates an analytical hierarchical process with fuzzy membership functions (FAHP). Using the relative operating characteristics (ROC), the performance of these three LCC models were evaluated. The results showed 85%, 75%, and 73% accuracy for the ANN, FAHP, and WOE. The ANN was clearly superior compared to the other LCC models when simulating urban growth for the year 2010; hence, ANN was used to predict urban growth for 2020 and 2030. Projected urban growth maps were assessed using statistical measures, including figure of merit, average spatial distance deviation, producer accuracy, and overall accuracy. Based on our findings, we recomend ANNs as an and accurate method for simulating future patterns of urban growth.     

Application of ANN and ANFIS models for reconstructing missing flow data

Hydrological yearbooks, especially in developing countries, are full of gaps in flow data series. Filling missing records is needed to make feasibility studies, potential assessment, and real-time decision making. In this research project, it was tried to predict the missing data of gauging stations using data from neighboring sites and a relevant architecture of artificial neural networks (ANN) as well as adaptive neuro-fuzzy inference system (ANFIS). To be able to evaluate the results produced by these new techniques, two traditionally used methods including the normal ratio method and the correlation method were also employed. According to the results, although in some cases all four methods presented acceptable predictions, the ANFIS technique presented a superior ability to predict missing flow data especially in arid land stations with variable and heterogeneous data. Comparing the results, ANN was also found as an efficient method to predict the missing data in comparison to the traditional approaches.     

Estimating soil temperature using neighboring station data via multi-nonlinear regression and artificial neural network models

The aim of this study is to estimate the soil temperatures of a target station using only the soil temperatures of neighboring stations without any consideration of the other variables or parameters related to soil properties. For this aim, the soil temperatures were measured at depths of 5, 10, 20, 50, and 100 cm below the earth surface at eight measuring stations in Turkey. Firstly, the multiple nonlinear regression analysis was performed with the “Enter” method to determine the relationship between the values of target station and neighboring stations. Then, the stepwise regression analysis was applied to determine the best independent variables. Finally, an artificial neural network (ANN) model was developed to estimate the soil temperature of a target station. According to the derived results for the training data set, the mean absolute percentage error and correlation coefficient ranged from 1.45% to 3.11% and from 0.9979 to 0.9986, respectively, while corresponding ranges of 1.685–3.65% and 0.9988–0.9991, respectively, were obtained based on the testing data set. The obtained results show that the developed ANN model provides a simple and accurate prediction to determine the soil temperature. In addition, the missing data at the target station could be determined within a high degree of accuracy.     

Observation of saltwater intrusion and ETM dynamics in a stably stratified estuary: the Yangtze Estuary,China

Spatial and temporal measurement data describing spring–neap variations of velocity, salinity, and suspended sediment concentration (SSC) in the North Passage Deepwater Navigational Channel (DNC) of the Yangtze Estuary, China, were obtained in the wet season of 2012. These data were collected in the middle of the DNC and apparently document the formation of a rather stable density stratification interface and salt wedge, especially during neap tides and slack waters. The convergent zone of residual currents, salinity transport, and sediment transport during neap and spring tides oscillates in the middle and lower reach of the DNC. It encourages the formation of a near-bed high-SSC layer, which favours siltation in the dredged channel. Both the near-bed gradient Richardson number and the bulk/layer Richardson number vary dramatically from around zero to several hundred from spring to neap tides. Stratification and turbulence damping effects near the estuarine turbidity maximum (ETM) area induce the upper half (near water surface) of the water body to be ebb-dominant and the lower part (near-bed) to be flood-dominant, which is a previously undocumented phenomenon in this region. These data reveal that the residual pattern of currents, salt flux and sediment flux are of critical differences in a stratified estuary, and that the salinity-induced baroclinic pressure gradient is a major factor controlling the vertical velocity structure. In addition, field observations indicate that the salinity and sediment transport of residuals generated by internal tidal asymmetry plays a dominant role in maintaining a stable density stratification interface near the estuarine front.     

Sediment fluxes and the littoral drift along northeast Andhra Pradesh Coast, India: estimation by remote sensing

The littoral drift regime along the northeastern coast of India was investigated by analyzing coastal drift indicators and shoreline changes based on multitemporal satellite images. The study of offshore turbidity patterns and quantitative estimation of suspended sediments was undertaken to understand the magnitude and direction of movement of sediment fluxes. The study revealed that: (1) the character of coastal landforms and sedimentation processes indicate that the sediment transport is bidirectional and monsoon dependent; (2) multidate, multitemporal analysis of satellite images helps to show the nature of sediment transport along the coast. The dominant net sediment transport is in a NE direction along the eastern coast of India. Finally, this assessment demonstrates the potential of remote sensing technology in understanding the coastal morphometric changes, long-term sediment transport, shoreline changes, and offshore turbidity distribution pattern and the implications for the transport of sediment-associated pollutants.     

Selection of appropriate sampling stations in a lake through mapping

Much valuable information is obtained from water quality measurements and monitoring of lakes around the world. A powerful tool is the use of mapping techniques, as it offers potential use in water quality research. Both remote sensing techniques and traditional water quality monitoring are required to collect data at sampling stations. This study suggests another approach to determine the most appropriate distribution of sampling stations in water reservoirs that will be mapped for water quality parameters. Tests were conducted for the proposed approach for Secchi disc depth (SDD), chlorophyll-a, turbidity and suspended solids parameters in Lake Beysehir, Turkey. Results of analysis are available for a total of 30 sampling stations in August 2006. Ten sampling stations were used to model Lake Beysehir while the others were used for validation of the model. Sampling stations that offered the best representation of the lake for each parameter were determined. Then, the best representative sampling stations for all parameters in the study were determined. Moreover, in order to confirm the accuracy of these re-determined sampling stations, modelling was performed on the results of the analysis of June 2006, and it was found that the values obtained from the re-determined sampling stations were acceptable.     

Retrieval of total suspended matter (TSM) and chlorophyll-a (Chl-a) concentration from remote-sensing data for drinking water resources

The concentrations of chlorophyll-a (Chl-a) and total suspended matter (TSM) are major water quality parameters that can be retrieved using remotely sensed data. Water sampling works were conducted on 15 July 2007 and 13 September 2008 concurrent with the Indian Remote-Sensing Satellite (IRS-P6) overpass of the Shitoukoumen Reservoir. Both empirical regression and back-propagation artificial neural network (ANN) models were established to estimate Chl-a and TSM concentration with both in situ and satellite-received radiances signals. It was found that empirical models performed well on the TSM concentration estimation with better accuracy (R ² = 0.94, 0.91) than their performance on Chl-a concentration (R ² = 0.62, 0.75) with IRS-P6 imagery data, and the models accuracy marginally improved with in situ spectra data. Our results indicated that the ANN model performed better for both Chl-a (R ² = 0.91, 0.82) and TSM (R ² = 0.98, 0.94) concentration estimation through in situ collected spectra; the same trend followed for IRS-P6 imagery data (R ² = 0.75 and 0.90 for Chl-a; R ² = 0.97 and 0.95 for TSM). The relative root mean square errors (RMSEs) from the empirical model for TSM (Chl-a) were less than 15% (respectively 27.2%) with both in situ and IRS-P6 imagery data, while the RMSEs were less than 7.5% (respectively 18.4%) from the ANN model. Future work still needs to be undertaken to derive the dynamic characteristic of Shitoukoumen Reservoir water quality with remotely sensed IRS-P6 or Landsat-TM data. The algorithms developed in this study will also need to be tested and refined with more imagery data acquisitions combined with in situ spectra data.     

Natural and anthropogenic contamination of the Fratta-Gorzone river (Veneto, Italy)

Stream-bed sediment samples were collected in 2001 and 2004 along the Fratta-Gorzone River (Italy) to assess the level of heavy metal contamination. The river stretch most affected by discharges of tannery effluent showed total and pseudo-total Cr levels (up to 2,860 mg/kg) that greatly exceed national and international chemical sediment quality standards. The most contaminated section of the river bed is located downstream of the main industrial discharge. However, a large fraction of the Cr present in the sediment appears to be of lithogenic origin. At most sites, more than 50% of Cr is not soluble in aqua-regia and thus unlikely to be very mobile or easily bio-available. A negligible risk to the benthic community can be inferred for Pb, Zn, Cd, Cu and Ni. This work highlights the limitation of using existing chemical sediment quality standards alone for risk assessment. The collection and analysis of suspended solids, the determination of river discharge and of dissolved Cr at 10 field stations allowed to estimate the particulate and dissolved Cr load and to locate the river stretch that was the likely source of contaminated sediment. The pumping of dilution water from the Adige River into the Fratta-Gorzone River did not produce the expected contaminant dilution effect due to re-suspension of contaminated solid particles and the release of Cr in solution.     

Restoration of the Kissimmee River, Florida: Water Quality Impacts from Canal Backfilling

The planned restoration of the Kissimmee River ecocystem will backfill approximately 35 km of flood control canal (C-38) that cuts through the meandering river channel, re-establish natural flow patterns, and restore the river/floodplain ecosystem. Water quality monitoring, including nutrients, total suspended solids (TSS), turbidity, dissolved oxygen (DO), and mercury, was conducted during a pilot `test fill' project to determine if soil disturbance during canal backfilling would negatively impact these water quality constituents. Surface water nutrient concentrations varied little between sites. Generally, highest concentrations occurred prior to construction, with lowest concentrations occurring during and after construction. During construction, TSS concentrations increased at sites immediately upstream, downstream, and adjacent to the construction area. Increased turbidity was generally restricted to areas immediately upstream and downstream of the test plug, with maximum levels occurring during the initial construction phase. Some downstream increases in turbidity were observed; however, impacts were short-term, lasting less than 24 h. Depresssed DO levels (<2 mg/l) were observed upstream of the test plug following completion of the initial plug across C-38. Dissolved oxygen levels remained low for approximately 6 weeks, with no apparent ecological impacts. Total mercury (HgT) within canal sediment ranged from 9.2–180 ng/g and methylmercury concentrations ranged from 0.037–0.708 ng/g. Concentration of total mercury and total methylmercury (MeHgT) in the backfill material were much lower than concentrations in the canal sediment. No significant change in aqueous HgT concentrations occurred over the sampling period, although construction-induced turbidity could have temporarily caused a slightly elevated concentration immediately downstream of the construction site. Methylmercury concentrations in the water column ranged from 0.033–0.518 ng/l. No significant differences in mean MeHgT concentrations occured between sites or between sampling dates, except at one downstream site where MeHgT declined significantly over the sampling period.