Neural network forecasting of air pollutants hourly concentrations using optimised temporal averages of meteorological variables and pollutant concentrations

The new method for the forecasting hourly concentrations of air pollutants is presented in the paper. The method was developed for a site in urban residential area in city of Zagreb, Croatia, for four air pollutants (NO₂, O₃, CO and PM₁₀). Meteorological variables and concentrations of the respective pollutant were taken as predictors. A novel approach, based on families of univariate regression models, was employed in selecting the averaging intervals for input variables. For each variable and each averaging period between 1 and 97 h, a separate model was built. By inspecting values of the coefficient of correlation between measured and modelled concentrations, optimal averaging periods for each variable were selected. A new dataset for building the forecasting model was then calculated as temporal moving averages (running means) of former variables. A multi-layer perceptron type of neural networks is used as the forecasting model. Index of agreement, calculated for the entire dataset including the data for model building, ranged from 0.91 to 0.97 for the respective pollutants. As suggested by the analysis of the relative importance of the input variables, different agreements for different pollutants are likely due to different sources and production mechanisms of investigated pollutants. A comparison of the new method with more traditional method, which takes hourly averages of the forecast hour as input variables, showed similar or better performance. The model was developed for the purpose of public-health-oriented air quality forecasting, aiming to use a numerical weather forecast model for the prediction of the part of input data yet unknown at the forecasting time. It is to expect that longer term averages used as inputs in the proposed method will contribute to smaller input errors and the greater accuracy of the model.     

Application of Synchronous Fluorescence Spectroscopy for Characterization of the Aromatic Hydrocarbons in Sediments of the Middle Adriatic

UV-fluorescence spectroscopy method with synchronous mode of scanning was used to characterize the types of aromatic hydrocarbons in surface sediments. the sampling stations were located on two transverses between Yugoslavian and the Italian coasts in the middle of the Adriatic Sea. the preparation of sediment samples was made according to IOC/UNESCO (1982). Synchronous excitation/emission scanning were done at wavelengths from 236/260 to 516/540 nm. Since the wavelength of maximum emission is a function of fused aromatic rings in a molecule, the fluorescence spectra of each sample were divided into three intervals: 300–340, 340–400, over 400 nm, corresponding to compounds with 2, 3 and 4,5 and more rings respectively.

Large qualitative differences were established between sediment samples. PAH with 5 and more rings are more prevalent near the Italian coast.     

Application of Synchronous Fluorescence Spectroscopy for Characterization of the Aromatic Hydrocarbons in Sediments of the Middle Adriatic

UV-fluorescence spectroscopy method with synchronous mode of scanning was used to characterize the types of aromatic hydrocarbons in surface sediments. the sampling stations were located on two transverses between Yugoslavian and the Italian coasts in the middle of the Adriatic Sea. the preparation of sediment samples was made according to IOC/UNESCO (1982). Synchronous excitation/emission scanning were done at wavelengths from 236/260 to 516/540 nm. Since the wavelength of maximum emission is a function of fused aromatic rings in a molecule, the fluorescence spectra of each sample were divided into three intervals: 300-340, 340-400, over 400 nm, corresponding to compounds with 2, 3 and 4,5 and more rings respectively.

Large qualitative differences were established between sediment samples. PAH with 5 and more rings are more prevalent near the Italian coast.     

Influence of Meteorological Factors NO2, SO2, CO and PM10 on the Concentration of O3 in the Urban Atmosphere of Eastern Croatia

Ozone, NO₂, SO₂, CO, PM₁₀ and meteorological parameters were measured simultaneously during the summer?Cautumn season 2007 in Osijek??the eastern, flat, agricultural part of Croatia. Fourier analysis confirms the existence of variation in ozone volume fractions with periods ranging from the usual semi-daily and daily to 7 and 28 daily cycles. The relationships between O₃ and other variables were modelled in three ways: principal component analysis, multiple linear regression and principal component regression. The results of the principal component analysis detected underlying relationships among ozone concentrations and meteorological variables. An extremely simple meteorological model is suitable for the prediction of ozone levels. The meteorological factors, temperature and cloudiness played a main role in the MLR model (R ²?=?0.83). The application of the principal component regression approach confirmed that the original variables associated with the valid principal components were meteorological variables (R ²?=?0.82).     

Modelling spatial distribution of the Croatian marine benthic habitats

Within the framework of the 3-year project “Mapping the habitats of the Republic of Croatia” the marine benthic habitats of the entire Croatian maritory were mapped. The supralittoral and the mediolittoral were mapped as a function of the coastal lithology and the presumed levels of human impact (both in scale of 1:100,000). The infralittoral was mapped on the basis of spatial modelling (using neural networks as a modelling tool, data about habitats collected by fieldwork as the independent variable for training and testing the model, and the digital bathymetrical model, the distance from coast, the second spectral channel of Landsat ETM+ satellite image and the sea bottom sea temperature, salinity and current magnitude, as dependent variables). The circalittoral and the bathyal were mapped by overlapping and reinterpretation of the existing spatial databases (bathymetry and lithology) within the framework of the raster-GIS.