Ferruh Ertürk

A neural network-based approach for the prediction of urban SO2 concentrations in the Istanbul metropolitan area

A three-layer Artificial Neural Network (ANN) model was developed to forecast air pollution levels. The subsequent SO2 concentration (24-hour averaged) being the output parameter of this study was estimated by seven input parameters such as preceding SO2 concentrations (24-hour averaged), average daily temperature, sea-level pressure, relative humidity, cloudiness, average daily wind speed and daily dominant wind direction. After Backpropagation training combined with Principal Component Analysis (PCA), the proposed model predicted subsequent SO2 values based on measured data. ANN testing outputs were proven to be satisfactory with correlation coefficients of about 0.770, 0.744 and 0.751 for the winter, summer and overall data, respectively.

Evaluation of air pollution trends in Istanbul

Previously, SO2 and PM10 (particulate matter less than 10 Bm in size) concentration distributions have been investigated in order to assess air pollution in Istanbul during the winter season (November–March) in which the concentration of these pollutants had reached formidably high levels due to the consumption of low-quality fuels (mainly coal) for residential heating. In this study, the effect of the increased share of natural gas and high-quality coal consumption in residential areas on air pollution levels was investigated. Modelling employing the method of kriging by spherical interpolation was used to obtain the concentration distribution of these pollutants, and spatial distributions of concentrations were generated. The pollution map obtained by this method indicated that increased usage of natural gas and high-quality coal in residential areas significantly improved air quality.

Prediction of Ground Level SO2 Concentration using Artificial Neural Networks

Future (24 h later) daily ground level SO2 concentration in Istanbul was modeled and predicted using a new and powerful technique, Artificial Neural Networks (ANN) in the case of meteorological parameters as input variables. Results show that the trend of SO2 from higher values in winter to lower values in spring and summer, and again to higher values towards winter can be correctly represented by the neural networks. The model better predicted the lower SO2 values in spring and summer seasons when compared to higher values in winter season because of the pattern distribution in training data sets. Beside the amount of the database, the more the variation of the values of the parameters in their own ranges, the more the network learns the database. As a result of this study, considerably successful results were obtained when considering the complex and nonlineer structure of the atmosphere, which is the source of the database.

Effect of diurnal changes in VOC source strengths on performances of receptor models

IntroductionThe effect of diurnal changes in strengths of volatile organic compound (VOC) sources on the performances of positive matrix factorization (PMF) and principal component analysis (PCA) was investigated using ambient measurement results that were taken during daytime and nighttime hours between March 24 and May 14, 2011, within Davutpasa Campus of Yildiz Technical University (Istanbul, Turkey).MethodsForty-five VOC species, ranging from C5 to C11 in volatility, were measured in the samples, 40 of which are included in the analyses. Ambient samples were grouped as daytime, nighttime, and all day datasets, and both PMF and PCA were applied to each dataset. A total of six source groups were extracted from each dataset: solvent use, general industrial paint use, gasoline and diesel vehicle exhausts, and biogenic as well as evaporative emissions. Estimated source contributions showed great diurnal variations.ResultsThe results suggested that extraction of possible sources by PCA depends greatly on the number of samples and the strength of the sources, while PMF produced stable results regardless of number of samples and source strengths.ConclusionAlthough PMF was unable to resolve gasoline vehicle and evaporative emissions, it was found to be successful in explaining diurnal fluctuations in source strengths, while the performance of PCA depends on the strength of emission source.

APPLICATION OF INDUCTIVE LEARNING: AIR POLLUTION FORECAST IN ISTANBUL, TURKEY

Abstract In this study, Istanbul city was taken as the study area. A new and powerful technique, Artificial Intelligence (AI), an Inductive Learning Algorithm (RULES-3), was used in predicting the future (next 24 hours) sulfur dioxide (SOZ) air pollutant on the basis of vazious meteorological parameters. The goal of this study is to forecast the 24-h average SOZ concentration levels in the urban atmosphere using AL As a result of this study, it was seen that AI is a powerful tool in estimating air pollution levels considering the complex and nonlinear structure of the atmospheric parameters, which is the source of the database.

Study of the Decrease of Air Pollution Concentration Levels with the Meteorological Factors in Istanbul

The aim of this study is twofold: First, to gain perspective for the assessment of the spatial distribution of one of the air pollutants, sulfur dioxide, in the region by the use of a statistical modelling scheme, kriging; and second, to show the decrease of sulfur dioxide concentrations over Istanbul by elaborating on the reasons of this decrease in terms of the meteorological factors.

Atmospheric lead concentrations near roadways in a suburban part of Istanbul

In this study, daily aerosol samples of fine (PM2.5) and coarse (PM2.5−10) particles were collected at Buyukcekmece (41°2′35″N, 28°35′25″E), a sub-urban part of Istanbul, Turkey. Sampling area is under the impact of two main transport arteries, namely the Trans European Motorway (TEM) and E5 motorways. Atmospheric lead levels measured in this study were compared with those reported previously from various locations in Turkey and other parts of the world. These values lie midway between those typical of remote and urban sites. The data were statistically evaluated to determine seasonal changes. Then, wind sector analysis was carried out. The analysis showed that the major proportion of the episodes, about 54%, was generated from the direction of TEM motorway, located to the north, while the second major lead source, with a contribution of 20%, was found to be the direction of E-5 motorway, located to the south.

Analysis of Air Pollutants in Istanbul: A Preliminary Study

In last ten years, Istanbul faced severe air pollution problems. The occurrences of high concentrations of air pollution in the city have reached to the levels of danger for the habitants (Erturk, 1986). The main cause of Istanbul ’s air pollution comes from the burning of low quality lignite and fossil fuels containing high percent of sulfur used mainly for heating purposes (Erturk, 1986; Erturk et al., 1995). It is the purpose of this article to gain perspective for the assessment of the spatial distribution of one of the air pollutants, SO2, for the Istanbul municipality region by the usage of a spatial prediction scheme, Kriging (Matheron, 1971).The city ’s population increased exponentially at the beginning of 1980 ’s. l/5th of the population and 70% of the country ’s industry is located in Istanbul. The urbanization of the city is badly organized and unplanned.