The forecasting of air pollution is an important and popular topic in environmental engineering. Due to health impacts caused by unacceptable particulate matter (PM) levels, it has become one of the great est concerns in metropolitan cities like Karaj City in Iran. In this study, the concentration of PM 2.5 was predicted by applying a multilayer percepteron (MLP) neural network, a radial basis function (RBF) neural network and a Markov chain model. Two months of hourly data including temperature, NO, NO 2 , NO x , CO, SO 2 and PM 10 were used as inputs to the artifi cial neural networks. From 1,488 data, 1,300 of data was used to train the models and the rest of the data were applied to test the models. The results of using artificial neural networks indicated that the models performed well in predicting PM 2.5 concen trations. The application of a Markov chain described the probable occurrences of unhealthy hours. The MLP neural network with two hidden layers including 19 neurons in the first layer and 16 neurons in the second layer provided the best results. The coeffi cient of determination (R 2 ), Index of Agreement (IA) and Efficiency (E) between the observed and the predicted data using an MLP neural network were 0.92, 0.93 and 0.981, respectively. In the MLP neu ral network, the MBE was 0.0546 which indicates the adequacy of the model. In the RBF neural net work, increasing the number of neurons to 1,488 caused the RMSE to decline from 7.88 to 0.00 and caused R 2 to reach 0.93. In the Markov chain model the absolute error was 0.014 which indicated an ac ceptable accuracy and precision. We concluded the probability of occurrence state duration and transi tion of PM 2.5 pollution is predictable using a Markov chain method.

Coagulation and flocculation are two main processes used to integrate colloidal particles into larger particles and are two main stages of primary water treatment. Coagulation and flocculation processes are only needed when colloidal particles are a significant part of the total suspended solid fraction. Our objective was to predict turbidity of water after the coagulation and flocculation process while other parameters such as types and concentrations of coagulants, pH, and influent turbidity of raw water were known. We used a multilayer perceptron (MLP), a radial basis function (RBF) of artificial neural networks (ANNs) and various kinds of fuzzy regression analysis to predict turbidity after the coagulation and flocculation processes. The coagulant used in the pilot plant, which was located in water treatment plant, was poly aluminum chloride. We used existing data, including the type and concentrations of coagulant, pH and influent turbidity, of the raw water because these types of data were available from the pilot plant for simulation and data was collected by the Tehran water authority. The results indicated that ANNs had more ability in simulating the coagulation and flocculation process and predicting turbidity removal with different experimental data than did the fuzzy regression analysis, and may have the ability to reduce the number of jar tests, which are time-consuming and expensive. The MLP neural network proved to be the best network compared to the RBF neural network and fuzzy regression analysis in this study. The MLP neural network can predict the effluent turbidity of the coagulation and the flocculation process with a coefficient of determination (R) of 0.96 and root mean square error of 0.0106.

We applied multilayer perceptron (MLP) and radial basis function (RBF) neural networks using data from two water quality monitoring stations at the Karaj Dam in Iran. Input data were calcium ions (Ca2+), magnesium ions (Mg2+), sodium ions (Na+), chloride ions (Cl−), sulfate ( SO 42−), and pH, and the output data were total dissolved solids (TDS). An MLP with one hidden layer containing eight neurons was selected for the upstream water quality station using normalized input data. We developed a second MLP neural network for the downstream station with one hidden layer containing 10 neurons in the hidden layer using normalized input data. Considering applying normalized input data and one hidden layer, the coefficient of determination (R2) and index of agreement (IA) between the observed and the predicted data for the upstream and downstream monitoring stations using the MLP neural networks were 0.985, 0.84, 0.99, and 0.92, respectively. The RBF neural network with 100 neurons in its hidden layer reached the minimum errors between the observed and the predicted results in upstream and downstream stations. The R2 between observed and predicted data for upstream and downstream monitoring stations for the RBF was 0.999 and 0.998, respectively. Data normalization improved the performance of the MLP neural networks. Sensitivity analysis indicated that magnesium is the most effective water quality parameter for predicting TDS, and sulfate is the second most effective water quality parameter affecting TDS prediction at the Karaj Dam.