Explicit Modeling of Meteorological Explanatory Variables in Short-Term Forecasting of Maximum Ozone Concentrations via a Multiple Regression Time Series Framework

Abstract

Statistical time series forecasting is a useful tool for predicting air pollutant concentrations in urban areas, especially in emerging economies, where the capacity to implement comprehensive air quality models is limited. In this study, a general multiple regression with seasonal autoregressive moving average errors model was estimated and implemented to forecast maximum ozone concentrations with a short time resolution: overnight, morning, afternoon and evening. In contrast to a number of short-term air quality time series forecasting applications, the model was designed to explicitly include the effects of meteorological variables on the ozone level as exogenous variables. As the application location, the model was constructed with data from five monitoring stations in the Monterrey Metropolitan Area of Mexico. The results show that, together with structural stochastic components, meteorological parameters have a significant contribution for obtaining reliable forecasts. The resulting model is an interpretable, useful and easily implementable model for forecasting ozone maxima. Moreover, it proved to be consistent with the general dynamics of ozone formation and provides a suitable platform for forecasting, showing similar or better performance compared to models in other existing studies.

Forecasting Daily of Surface Ozone Concentration in the Grand Casablanca Region Using Parametric and Nonparametric Statistical Models

“…The final model (Section 3.4) that considers only five of the most significant predictors such as (DRI NSQ, PRESTN 06 , Vx 06 , Vy 06 , and O 3veilleJahid ) was selected. In fact, these meteorological factors are coherent with the results obtained in the literature [3,7,21]. This final selected model has the best predictive ability (RMSEP = 12.55) compared to the other models.…”

“…Indeed, O 3 concentration is influenced positively by sunshine duration (+1.41), horizontal wind direction at 6 h (+1.34), and the previous day's O 3 (+23.51) and negatively by pressure at 6 h (−1.17) and vertical wind direction at 6 h (−1.33). These results are similar to those obtained in studies conducted by [3,21,48].…”

“…• parametric linear models (Multiple Linear Model and its alternative models): Brazil [31], Canada [32], China [33], Croatia [34], Greece [35], Italy [36], Portugal [14], Spain [37], Malaysia [38], the United States [39], Mexico [21], etc. • nonparametric models (CART Tree, Bagging, and Random Forests): the United States [40], France [41], Spain [42], China [17], Sweden [43], etc.…”

“…During the last decade, many researchers have studied the problem of forecasting tropospheric ozone using multivariate statistical methods.These methods can be classified into two main categories: parametric and nonparametric methods [13][14][15]. Recent studies, such as [16][17][18][19][20][21], have compared the performance of these forecasting methods in selecting the most appropriate one. The results have shown that Multiple Linear Regression (MLR), regression tree, and Random Forests give good results.…”

Analysis of the Main Anthropogenic Sources’ Contribution to Pollutant Emissions in the Lazio Region, Italy

“…In this context, the prediction of the air pollutant level has received great attention by recent research [30][31][32][33]. Numerical prediction methods are widely used to predict the pollutant level in cities [34][35][36]. Cordano and Frieze [37] used a deterministic model, Tian and Chen [38] used an empirical black-box model, Russell et al [39] use a statistical model while Suleiman et al [40] adopt a machine learning model.…”