Gokmen Tayfur

Artificial neural networks for sheet sediment transport

Abstract Sheet sediment transport was modelled by artificial neural networks (ANNs). A three-layer feed-forward artificial neural network structure was constructed and a back-propagation algorithm was used for the training of ANNs. Event-based, runoff-driven experimental sediment data were used for the training and testing of the ANNs. In training, data on slope and rainfall intensity were fed into the network as inputs and data on sediment discharge were used as target outputs. The performance of the ANNs was tested against that of the most commonly used physically-based models, whose transport capacity was based on one of the dominant variables—flow velocity (V), shear stress (SS), stream power (SP), and unit stream power (USP). The comparison results revealed that the ANNs performed as well as the physically-based models for simulating nonsteady-state sediment loads from different slopes. The performances of the ANNs and the physically-based models were also quantitatively investigated to estimate mean sediment discharges from experimental runs. The investigation results indicated that better estimations were obtained for V over mild and steep slopes, under low rainfall intensity; for USP over mild and steep slopes, under high rainfall intensity; for SP and SS over very steep slopes, under high rainfall intensity; and for ANNs over steep and very steep slopes, under very high rainfall intensities.

The use of GA-ANNs in the modelling of compressive strength of cement mortar

In this paper, results of a project aimed at modelling the compressive strength of cement mortar under standard curing conditions are reported. Plant data were collected for 6 months for the chemical and physical properties of the cement that were used in model construction and testing. The training and testing data were separated from the complete original data set by the use of genetic algorithms (GAs). A GA– artificial neural network (ANN) model based on the training data of the cement strength was created. Testing of the model was also done within low average error levels (2.24%). The model was subjected to sensitivity analysis to predict the response of the system to different values of the factors affecting the strength. The plots obtained after sensitivity analysis indicated that increasing the amount of C3S, SO3 and surface area led to increased strength within the limits of the model. C2S decreased the strength whereas C3A decreased or increased the strength depending on the SO3 level. Because of the limited data range used for training, the prediction results were good only within the same range. The utility of the model is in the potential ability to control processing parameters to yield the desired strength levels and in providing information regarding the most favourable experimental conditions to obtain maximum compressive strength. D 2003 Elsevier Science Ltd. All rights reserved.

PHYSICAL AND MATHEMATICAL MODELLING OF ANAEROBIC DIGESTION OF ORGANIC WASTES

Anaerobic digestion of the organic food fraction of municipal solid waste (OFMSW), on its own or co-digested with primary sewage sludge (PSS), produces high quality biogas, suitable as renewable energy. We report the results from one such bench scale laboratory experiment, on the co-digestion of OFMSW and PSS. The experiment used a continuously stirred tank reactor and operated at 36°C for 115 days. Prior to the experiments, activity tests verified that the inoculum sludges were suitable for the biodegradation of the volatile fatty acid substrate and so producing biogas. The experimental data were used to develop and validate a two-stage mathematical model of acidogenesis and methanogenesis. In simulating the behavior of the anaerobic digestion process, including ammonia inhibition, the mathematical model successfully predicts the performance of methane production. Simulations of the pH and ammonia in the MSW anaerobic reactor were also satisfactory. Sensitivity analysis on the 18 model parameters indicated that eight of these parameters were in the most sensitive and highly sensitive range, while the remainder were in the moderate to least sensitive range.

Groundwater contamination and its effect on health in Turkey

The sources of groundwater pollution in Turkey are identified, and pathways of contaminants to groundwater are first described. Then, the effects of groundwater quality on health in Turkey are evaluated. In general, sources of groundwater contamination fall into two main categories: natural and anthropogenic sources. Important sources of natural groundwater pollution in Turkey include geological formations, seawater intrusion, and geothermal fluid(s). The major sources of anthropogenic groundwater contamination are agricultural activities, mining waste, industrial waste, on-site septic tank systems, and pollution from imperfect well constructions. The analysis results revealed that natural contamination due to salt and gypsum are mostly found in Central and Mediterranean regions and arsenic in Aegean region. Geothermal fluids which contain fluoride poses a danger for skeleton, dental, and bone problems, especially in the areas of Denizli, Isparta, and Aydın. Discharges from surface water bodies contaminate groundwater by infiltration. Evidence of such contamination is found in Upper Kızılırmak basin, Gediz basin, and Büyük Melen river basin and some drinking water reservoirs in İstanbul. Additionally, seawater intrusion causes groundwater quality problems in coastal regions, especially in the Aegean coast. Industrial wastes are also polluting surface and groundwater in industrialized regions of Turkey. Deterioration of water quality as a result of fertilizers and pesticides is another major problem especially in the regions of Mediterranean, Aegean, Central Anatolia, and Marmara. Abandoned mercury mines in the western regions of Turkey, especially in Çanakkale, İzmir, Muğla, Kütahya, and Balıkesir, cause serious groundwater quality problems.

Areally-averagei overland flow equations at hillslope scale

Abstract Microscale-averaged inter-rill area sheet flow and rill flow equations (Tayfur & Kavvas, 1994) are averaged along the inter-rill area length and rill length to obtain local areally-averaged inter-rill area sheet flow and rill low equations (local-scale areai averaging). In this averaging, the local areally-averaged flow depths are related to the microscale-averaged flow depths at the outlet sections (downstream ends) of a rill and an inter-rill area by the assumption that the flow in these sections has the profile of a sine function. The resulting local areally-averaged flow equations become time dependent only. To minimize computational efforts and economize on the number of model parameters, local areally-averaged flow equations are then averaged over a whole hillslope section (hillslope-scale areal averaging). The expectations of the terms containing more than one variable are obtained by the method of regular perturbation. Comparison of model results with observed data is satisfactory. The co...

A simplified model for two-dimensional overland flows

Abstract Numerical models of two-dimensional overland flow equations are often prohibitively expensive due to the highly nonlinear nature of the flow equations and the dense mesh required for accurate solutions. These models are frequently under-utilized due to lack of sufficiently detailed data at the grid scale. Observed results of the outflow hydrographs show fluctuations due to variability in the surface topography and precision limitations in the measuring instruments. A new solution methodology is presented in this paper using an eigenfunction expansion which is then combined with the kinematic wave approximation. The computational effort required by this new method is negligible when compared to the usual numerical methods. The results from this method are compared with the more expensive numerical results and experimentally observed results. These comparisons suggested that the semi-analytical solution methodology is an attractive modeling tool for practical two-dimensional overland flow computations.

Predicting Suspended Sediment Loads and Missing Data for Gediz River, Turkey

Prediction of suspended sediment load (SSL) is important for water resources quantity and quality studies. The SSL of a stream is generally determined by direct measurement of the suspended sediment concentration or by employing sediment rating curve method. Although direct measurement is the most reliable method, it is very expensive, time consuming, and, in many instances, problematic for inaccessible sections, especially during floods. On the other hand, measuring precipitation and flow discharge is relatively easier and hence, there are more rain and flow gauging stations than SSL gauging stations in Turkey. Furthermore, due to its cost, measurements of SSL are carried out in longer periods compared to precipitation and flow measurements. Although daily precipitation and flow measurements are available for most of the Turkish river basins, at best semimonthly measurements are available for SSL. As such, it is essential to predict SSL from precipitation and flow data and to fill the gap for the missing...

Groundwater quality and hydrogeochemical properties of Torbali Region, Izmir, Turkey.

The large demand for drinking, irrigation and industrial water in the region of Torbalı (Izmir, Turkey) is supplied from groundwater sources. Almost every factory and farm has private wells that are drilled without permission. These cause the depletion of groundwater and limiting the usage of groundwater. This study investigates spatial and temporal change in groundwater quality, relationships between quality parameters, and sources of contamination in Torbalı region. For this purpose, samples were collected from 10 different sampling points chosen according to their geological and hydrogeological properties and location relative to factories, between October 2001 and July 2002. Various physical (pH, temperature, EC), chemical (calcium, magnesium, potassium, sodium, chloride, alkalinity, copper, chromium, cadmium, lead, zinc) and organic (nitrate, nitrite, ammonia, COD and cyanide) parameters were monitored. It was observed that the groundwater has bicarbonate alkalinity. Agricultural contamination was determined in the region, especially during the summer. Nitrite and ammonia concentrations were found to be above drinking water standard. Organic matter contamination was also investigated in the study area. COD concentrations were higher than the permissible limits during the summer months of the monitoring period.

Supervised Intelligent Committee Machine Method for Hydraulic Conductivity Estimation

Hydraulic conductivity is the essential parameter for groundwater modeling and management. Yet estimation of hydraulic conductivity in a heterogeneous aquifer is expensive and time consuming. In this study; artificial intelligence (AI) models of Sugeno Fuzzy Logic (SFL), Mamdani Fuzzy Logic (MFL), Multilayer Perceptron Neural Network associated with Levenberg–Marquardt (ANN), and Neuro-Fuzzy (NF) were applied to estimate hydraulic conductivity using hydrogeological and geoelectrical survey data obtained from Tasuj Plain Aquifer, Northwest of Iran. The results revealed that SFL and NF produced acceptable performance while ANN and MFL had poor prediciton. A supervised intelligent committee machine (SICM), which combines the results of individual AI models using a supervised artificial neural network, was developed for better prediction of the hydraulic conductivity in Tasuj plain. The performance of SICM was also compared to those of the simple averaging and weighted averaging intelligent committee machine (ICM) methods. The SICM model produced reliable estimates of hydraulic conductivity in heterogeneous aquifers.

Principle Component Analysis in Conjuction with Data Driven Methods for Sediment Load Prediction

This study investigates sediment load prediction and generalization from laboratory scale to field scale using principle component analysis (PCA) in conjunction with data driven methods of artificial neural networks (ANNs) and genetic algorithms (GAs). Five main dimensionless parameters for total load are identified by using PCA. These parameters are used in the input vector of ANN for predicting total sediment loads. In addition, nonlinear equations are constructed, based upon the same identified dimensionless parameters. The optimal values of exponents and constants of the equations are obtained by the GA method. The performance of the so-developed ANN and GA based methods is evaluated using laboratory and field data. Results show that the expert methods (ANN and GA), calibrated with laboratory data, are capable of predicting total sediment load in field, thus showing their transferability. In addition, this study shows that the expert methods are not transferable for suspended load, perhaps due to insufficient laboratory data. Yet, these methods are able to predict suspended load in field, when trained with respective field data.