Short CommunicationForecasting of daily air quality index in Delhi
Highlights
► AQI of criteria pollutants based on the observed values has been estimated. ► Air quality index has been forecasted using three statistical techniques. ► PCA has been applied to reduce the number of predictive variable. ► Autoregressive and PCR models are combined to improve the forecast. ► Daily forecasting of AQI is useful for the general public to protect their health.
Introduction
Air pollution related problems have resulted in an increased public awareness of the air quality in both developing and developed countries (Kurt and Oktay, 2010). There are many air pollutants adversely affecting human health in the polluted air such as carbon monoxide (CO), RSPM, SO2, NO2, SPM, ozone (O3), etc. The high concentration of these pollutants can be life threatening, causing breathing difficulty, headache and dizziness. They may even result in heart attacks (Kunzli et al., 2000). Thus, the authorities advise the monitoring and forecasting of criteria pollutants in the air. The forecasting of air pollutants can be made through models. The Gaussian dispersion models are, generally, used for air quality prediction in most of the air pollution studies. Although, the dispersion models have some physical basis, detailed information about the source of the pollutants and other parameters are not generally known (Chelani et al., 2002). In order to overcome these limitations, the statistical models are used, which facilitate the forecasting of pollutant concentrations (Finzi and Tebaldi, 1982, Ziomass et al., 1995, Polydoras et al., 1998).
AQI is an important task for general public to understand easily how bad or good the air quality is for their health and to assist in data interpretation for decision making processes related to pollution mitigation measures and air quality management. Basically, the AQI is defined as an index or rating scale for reporting daily combined effect of ambient air pollutants recorded in the monitoring sites. Recently, Van den Elshout et al. (2008) gave the review of existing air quality indices. A regression model was also used by Cogliani (2001) for air pollution forecast in cities by an air pollution index highly correlated with meteorological variables. A study of Goyal et al. (2006) is made for daily air quality forecasting of air pollutants in Delhi through ARIMA and multiple linear regression (MLR) models. The most of the air quality forecasting studies present in the literature have been made for individual air pollutants.
The main objective of the present study is to develop forecasting models for predicting the daily air quality indices, which can provide the timely information to the public to take precautionary measures to protect their health.
Section snippets
Study area
Delhi is the capital city of India spread over 1483 km2 with 16.9 million inhabitants in 2007–08. Due to the presence of large number of industries and migration of people from neighboring states, nearly 56.27 lakh vehicles are plying on Delhi roads (Economic Survey of Delhi, 2008–2009). Delhi has one of the highest road densities as 1749 km of road length per 100 km2 in India. Its high population growth rate, coupled with high economic growth rate, has resulted in ever-increasing demand for
Methodology
There are primarily two steps involved for forecasting of daily AQI.
- i)
The estimation of AQI through USEPA method using daily observed concentration of air pollutants. In this formulation, the formation of sub-indices of each pollutant and the breakpoints aggregation of sub indices are made according to the Indian National Ambient Air Quality Standard. The results of epidemiological studies are indicating the risk of adverse health effects of specific pollutants. In order to assess the status of
Results and discussion
The daily AQI has been estimated using monitored concentrations of criteria pollutants in all the four seasons over the period from 2000 to 2006. The percentage of very poor and severe descriptors of AQI is found to be 81.52%, 81.51%, 69.54% and 38.04% during summer, winter, post-monsoon and monsoon seasons respectively. As one can see, summer and winter seasons have very poor and severe descriptors of AQI. It can be expected in winter due to worst meteorological scenario. However, the same
Conclusions
The present study focuses on the forecasting of daily air quality in terms of AQI in urban city Delhi. The AQI, based on air quality of air pollutants, is a simple number, easily understandable by general public to know how bad or good air quality is? The methodology of forecasting the AQI through different statistical models has been discussed in details. The three statistical models namely ARIMA, PCR and combination of ARIMA and PCR have been used for forecasting. The performance of all the
References (19)
Air pollution forecast in cities by an air pollution index highly correlated with metrological variables
Atmos Environ
(2001)- et al.
A mathematical model for air pollution forecast and alarm in an urban area
Atmos Environ
(1982) - et al.
Statistical models for the prediction of respirable suspended particulate matter in urban cities
Atmos Environ
(2006) - et al.
Public-health impact of outdoor and traffic-related air pollution: a European assessment
Lancet
(2000) - et al.
Forecasting air pollutant indicator levels with geographic models 3 days in advance using neural networks
Expert Syst Appl
(2010) - et al.
Assessment of air quality near traffic intersections in Bangalore city using air quality index
Transp Res D
(2007) - et al.
Air quality predictions: dispersion model vs Box–Jenkins stochastic models. An implementation and comparison for Athens, Greece
Appl Therm Eng
(1998) - et al.
Multiple linear regression and artificial neural networks based on principal components to predict ozone concentrations
Environ Modell Softw
(2007) - et al.
Comparing urban air quality in Europe in real time: a review of existing air quality indices and the proposal of a common alternative
Environ Int
(2008)
Cited by (131)
A forecasting framework on fusion of spatiotemporal features for multi-station PM<inf>2.5</inf>
2024, Expert Systems with ApplicationsAir Quality Index prediction using machine learning for Ahmedabad city
2023, Digital Chemical EngineeringA forecasting system for deterministic and uncertain prediction of air pollution data
2022, Expert Systems with ApplicationsReview of Intelligent Detection, Identification and Warning Methods for Airborne Particulate Matter Based on Machine Perception and Learning
2024, Beijing Gongye Daxue Xuebao/Journal of Beijing University of Technology