Statistical trend analysis and forecast modeling of air pollutants

Statistical trend analysis and forecast modeling of air pollutants

The study provides a statistical trend analysis of different air pollutants using Mann-Kendall and Sen's slope estimator approach on past pollutants statistics from air quality index station of Varanasi, India. Further, using autoregressive integrated moving average model, future values of air...

Full description

Saved in:
Bibliographic Details
Published in:Global journal of environmental science and management Vol. 4; no. 4; pp. 427 - 438
Main Authors: Jaiswal, A, Samuel, C, Kadabgaon, V M
Format: Journal Article
Language:English
Published: Tehran Solid Waste Engineering and Management Association, Faculty of Environment, University of Tehran 01-10-2018
GJESM Publisher
Subjects:
Air pollutants
Air pollution
Air quality
Autoregressive integrated moving average (ARIMA)
Autoregressive models
Carbon monoxide
Emission measurements
Forecast
Goodness of fit
Mann-Kendall
Modelling
Nitrogen dioxide
Outdoor air quality
Particulate emissions
Particulate matter
Pollutants
Pollution control
Sen’s slope estimator
Statistical analysis
Statistical tests
Statistics
Sulfur
Sulfur dioxide
Trend analysis
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The study provides a statistical trend analysis of different air pollutants using Mann-Kendall and Sen's slope estimator approach on past pollutants statistics from air quality index station of Varanasi, India. Further, using autoregressive integrated moving average model, future values of air pollutant levels are predicted. Carbon monoxide, nitrogen dioxide, sulphur dioxide, particulate matter particles as PM2.5 and PM10 are the pollutants on which the study focuses. Mann-Kendall and Sen's slope estimator tests are used on summer (February-May), monsoon (June-September) and winter (October-January) seasonal data from year 2013 to 2016 and trend results and power of the slopes are estimated. For predictive analysis, different autoregressive integrated moving average models are compared with goodness of fit statistics, and the observed results stated autoregressive integrated moving average (1,1,1) as the best-suited for forecast modeling of different pollutants in Varanasi. Autoregressive integrated moving average model (1,1,1) is also used on the annual concentration levels to predict forthcoming year's annual pollutants value. Study reveals that PM 10 shows a rising trend with predicted approximate annual concentration of 273 µg/m[sup]3[/sup] and PM2.5, carbon monoxide, nitrogen dioxide and sulphur dioxide show a reducing trend with approximate annual concentration of 139 µg/m[sup]3[/sup], 1.37 mg/ m[sup]3[/sup], 38 µg/m[sup]3[/sup] and 17 µg/m[sup]3[/sup], respectively, by the year 2030. The study predicted carbon monoxide, nitrogen dioxide and sulphur dioxide concentrations are lower and PM10 and PM2.5 concentrations are much higher to the standard permissible limits in future years also, and specific measures are required to control emissions of these pollutants in Varanasi.
ISSN:2383-3572
2383-3866
DOI:10.22034/gjesm.2018.04.004