Air quality forecasting using a hybrid autoregressive and nonlinear model

Air quality forecasting using nearest neighbour technique provides an alternative to statistical and neural network models, which needs the information on predictor variables and understanding of underlying patterns in the data. k-nearest neighbour method of forecasting that does not assume any linear or nonlinear form of the data is used in this study to obtain the next step forecast of PM 10 concentrations. Various function approximation techniques such as mean, median, linear combination and kernel regression of nearest neighbours are evaluated. It is observed that kernel regression of nearest neighbours outperforms the other individual models including bench mark persistence model for obtaining the next step forecasts. As the data may involve both linear and nonlinear patterns and any individual model cannot capture both types of patterns, combination forecasting is suggested as an alternative. The forecast error showed the outperformance of combination forecasting over individual forecast, which is quite obvious as it assigns more weightage to the model with minimum error.

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Nearest Neighbour Based Forecast Model for PM10 Forecasting: Individual and Combination Forecasting

Chelani

2015

“…5(a) and 5(b) for Nagpur and Chandrapur. The error statistics such as mean absolute percentage error (MAPE), relative error (RE) and Nash-Sutcliffe coefficient of efficiency (CE) is used to evaluate the model performance (Chelani and Devotta, 2006; http://en.wikipedia.org/wiki/Nash%E2%80 %93Sutcliffe_model_efficiency_coefficient). For a perfect fit, MAPE and RE should be close to 0 whereas CE should be close to 1.…”

Section: Predicting Number Of Observations Between Two Exceedancesmentioning

confidence: 99%

Exceedance Analysis of PM10 Concentration in Central Indian City: Predicting Gap between Two Exceedances

Chelani

2015

In this study the gap between the two exceedances is analyzed using time series analysis. The time series of PM 10 (particulate matter of size less than 10 micron) observed during 2005-2013 in two cities; Nagpur and Chandrapur in central India is considered. Higher PM 10 concentration is observed in Chandrapur as compared to Nagpur. Exponential relationship is observed between the average time between the two exceedances and annual average PM 10 concentration. This information along with the PM 10 concentration prediction model is utilized to predict the average number of observations between the two exceedances for the following year. k-nearest neighbor approach is used for forecasting PM 10 concentration which enabled estimating the average number of observations between two exceedances using exponential relationship. The approach can be used for estimating the average number of observations between the two exceedances over a year, which can further be utilized to make appropriate decision to control and manage high particulate matter pollution in an area.

“…The model captured the trends in ozone concentrations and showed a good agreement with the measured data. Chelani and Devotta (2006) used a hybrid autoregressive and nonlinear model to predict nitrogen dioxide concentration at a site in Delhi. Their results indicate hybrid model outperforms the individual linear and nonlinear models.…”

Section: Introductionmentioning

confidence: 99%

An Integrated SSA-ARIMA Approach to Make Multiple Day Ahead Forecasts for the Daily Maximum Ambient O3 Concentration

Kumar

2015

An integrated approach that combines a signal extraction method SSA (Singular Spectrum Analysis) with ARIMA (Autoregressive Integrated Moving Average) has been introduced to make multiple day ahead forecasts for daily maximum ambient O 3 concentrations. The results are compared with the forecasts obtained by more commonly used signal extraction method FFT (Fast Fourier Transforms) in combination with ARIMA. The data from six different European AIRBASE stations have been analyzed to test their skills in one-day ahead to five-day ahead forecasts. In SSA-ARIMA, the SSA has been used to extract the different structural components of the time series. The first few principal components have been selected based on their dominance in the eigenvalues spectrum. Each component has been modelled separately and their sum is called SSA-component. The residuals of the "SSA-component subtracted from the actual data" exhibited stationarity and have been modeled using ARIMA process. In the second approach called by FFT-ARIMA, FFT has been used to model the periodicities exhibited by the time series. The first few predominant frequencies have been selected based on the power-spectrum characteristics of the time series. The residuals of the "FFT-component subtracted from the actual data" were modeled using ARIMA process. The performances of both the SSA-ARIMA and FFT-ARIMA were evaluated for 30 out of sample days for one to five day ahead forecasts. The statistical indicators RMSE (root mean square error) and MAE (mean absolute error) reveal that the SSA-ARIMA consistently outperforms FFT-ARIMA. Moreover, SSA-ARIMA has also been found to perform better especially for multiple day ahead forecasts. The results clearly indicate that the SSA-ARIMA approach can be a potential alternative to make short term air quality forecasting.