There are many studies discussing the correlation between air pollution and human health hazards. Yet, in Tehran there is not a survey using time series methodology. Thus, we conducted a study based on time series data on the topic in Tehran, Iran. Mean levels of NO, NO2, NOX, CO, 03, SO2 and PM10 (particulate matters smaller than 10micrometer in diameter) were measured in one station of Tehran's Air Quality Control Corporation and were used as main independent variables. Mean temperature, mean humidity, day of the week, month and season were considered as potential confounders and deaths in people older than 64 years in Tehran was the dependent variable. All the variables were measured during Mar. 1998 to Dec. 1999. Concentrations of air pollutants were different between seasons and so were the means of daily deaths. Out of main independent variables, SO2, CO and PM10 showed statistically significant relations with the dependent variable (P<0.05). After controling for confounders, there was 3.4%, 2.6% and 3.36% increase in death rates, respectively, for each interquartile ascending (increase from 25th centile to 75th centile) in association to the mentioned pollutant centile concentration. No autocorrelation between residuals was observed (r= -0.059). The study showed that meteorological variables can confound the relation between air pollution and rate of deaths per day.  

In this paper we suggest a procedure for testing reversibility of time series. Our approach is based on a necessary and sufficient condition for time reversibility of linear models. An attractive feature of the procedure is that in converse with other approaches it doesn’t require important assumptions, especially existence of moments of order higher than two. Our simulation confirms the procedure and some empirical examples are given.

In this paper, we propose a new residual analysis method using Fourier series transform into fuzzy time series model for improving the forecast- ing performance. This hybrid model takes advantage of the high predictable power of fuzzy time series model and Fourier series transform to t the esti- mated residuals into frequency spectra, select the low-frequency terms, lter out high-frequency terms, and then have well forecasting performance. Two numerical examples are given to show that our proposed model can be applied with the best forecasting performance than the other models.