Using wavelet-feedforward neural networks to improve air pollution forecasting in urban environments.

Publisher: Springer Country of Publication: Netherlands NLM ID: 8508350 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1573-2959 (Electronic) Linking ISSN: 01676369 NLM ISO Abbreviation: Environ Monit Assess Subsets: MEDLINE

Publication: 1998- : Dordrecht : Springer
Original Publication: Dordrecht, Holland ; Boston : D. Reidel Pub. Co., c1981-

Air Pollution* ; Neural Networks, Computer*; Nitrogen Dioxide/*analysis ; Ozone/*analysis ; Particulate Matter/*analysis; Air Pollutants/analysis ; Cities ; Environmental Monitoring ; Forecasting/methods ; Models, Theoretical ; Romania ; Sensitivity and Specificity

The paper presents the screening of various feedforward neural networks (FANN) and wavelet-feedforward neural networks (WFANN) applied to time series of ground-level ozone (O3), nitrogen dioxide (NO2), and particulate matter (PM10 and PM2.5 fractions) recorded at four monitoring stations located in various urban areas of Romania, to identify common configurations with optimal generalization performance. Two distinct model runs were performed as follows: data processing using hourly-recorded time series of airborne pollutants during cold months (O3, NO2, and PM10), when residential heating increases the local emissions, and data processing using 24-h daily averaged concentrations (PM2.5) recorded between 2009 and 2012. Dataset variability was assessed using statistical analysis. Time series were passed through various FANNs. Each time series was decomposed in four time-scale components using three-level wavelets, which have been passed also through FANN, and recomposed into a single time series. The agreement between observed and modelled output was evaluated based on the statistical significance (r coefficient and correlation between errors and data). Daubechies db3 wavelet-Rprop FANN (6-4-1) utilization gave positive results for O3 time series optimizing the exclusive use of the FANN for hourly-recorded time series. NO2 was difficult to model due to time series specificity, but wavelet integration improved FANN performances. Daubechies db3 wavelet did not improve the FANN outputs for PM10 time series. Both models (FANN/WFANN) overestimated PM2.5 forecasted values in the last quarter of time series. A potential improvement of the forecasted values could be the integration of a smoothing algorithm to adjust the PM2.5 model outputs.

Environ Monit Assess. 2004 Nov;98(1-3):275-94. (PMID: 15473541)
Environ Monit Assess. 2008 Aug;143(1-3):131-46. (PMID: 17929183)
Environ Monit Assess. 2012 Apr;184(4):2151-9. (PMID: 21584759)
Environ Health. 2013 Nov 04;12:93. (PMID: 24188173)
Int J Public Health. 2012 Oct;57(5):757-68. (PMID: 22592907)
Environ Monit Assess. 2003 Sep;87(3):235-54. (PMID: 12952354)

0 (Air Pollutants)
0 (Particulate Matter)
66H7ZZK23N (Ozone)
S7G510RUBH (Nitrogen Dioxide)

Date Created: 20150702 Date Completed: 20160204 Latest Revision: 20191210

20240829

10.1007/s10661-015-4697-x

26130243