OPEN ACCESS

Articles online

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

Category: Technical Note

Volume: 15 | Issue: 3 | Pages: 1130-1136
DOI: 10.4209/aaqr.2014.09.0229
PDF | RIS | BibTeX

Asha B. Chelani

  • Air Pollution Control Board, National Environmental Engineering Research Institute (CSIR-NEERI), Nehru Marg, Nagpur – 440020, India

Highlights

Comparison of various functions of nearest neighbours for PM10 forecasting.
Kernel regression of nearest neighbours outperforms the other individual models.
Due to linear and nonlinear patterns in data, combination forecasting is suggested.
Outperformance of combination forecasting over individual forecast models.
Study is useful when the data on predictor variables is not available.


Abstract

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 PM10 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. The study is useful when the data on predictor variables that influence the air pollutant concentrations is not available. The assumption on the underlying distribution of the data is also not required for the approach.

Keywords

Time series forecasting Nearest neighbours PM10 concentration Combination forecasting


Related Article

Simulating Long Range Transport of Radioactive Aerosols Using a Global Aerosol Transport Model

Tanmay Sarkar, Srinivasan Anand , Kapil Deo Singh, Raj Mangal Tripathi, Pradeepkumar Kunhiraman Sarojini, Daisuke Goto, Teruyuki Nakajima
Accepted Manuscripts
DOI: 10.4209/aaqr.2017.01.0049
PDF

Deciphering Effects of Surface Charge on Particle Removal by TiO2 Polyacrylonitrile Nanofibers

Qian Zhang, Fang Liu, Tasi-Yu Yang, Xv Lv Si , Gong Ren Hu, Chang-Tang Chang
Volume: 17 | Issue: 7 | Pages: 1909-1916
DOI: 10.4209/aaqr.2016.07.0313
PDF

Enhanced Antimicrobial Activity on Non-Conducting and Conducting Air Filters by Using Air Ions and Grapefruit Seed Extract

Chang Gyu Woo, Hak-Joon Kim, Yong-Jin Kim, Bangwoo Han
Volume: 17 | Issue: 7 | Pages: 1917-1924
DOI: 10.4209/aaqr.2016.10.0466
PDF

A Low Cost System for Detecting Fog Events and Triggering an Active Fog Water Collector

Peter Weiss-Penzias , Daniel Fernandez, Robert Moranville, Chad Saltikov
Article In Press
DOI: 10.4209/aaqr.2016.11.0508
PDF
;