Search

Accurate PM 2.5 forecasting provides a possibility for establishing an early warning system to notify the public and take precautionary measures to prevent negative effects on ambient air quality and public health. Considering strong seasonal variation in meteorological conditions, in this paper, a seasonal stacked autoencoder model combining seasonal analysis and deep feature learning is proposed for forecasting the hourly PM 2.5 concentration, named DL-SSAE model. The original data are firstly decomposed into four seasonal subseries according to the Chinese calendar, and then the Kendall correlation coefficient method is employed to search inherent relationships between PM 2.5 concentrations and meteorological parameters within 1-h ahead for each seasonal time series. The inherent relationships of each seasonal subseries are finally extracted, learned, and modeled by different deep neural networks (stacked autoencoders for regression), and the hourly PM 2.5 forecasts are yielded. The addressed model is tested by the dataset collected from three environmental monitoring stations in Beijing, China. The results demonstrate that the proposed model outperforms all other considered models with/without seasonality consideration in this paper. [ABSTRACT FROM AUTHOR]

With the rapid development of China's industrialization, the air pollution is becoming more and more serious. It is vital for us to predict the air quality for determining the further prevention measures of avoiding the brought disasters. In this paper, we are going to propose an approach of predicting the air quality based on the multiple data features through fusing the multiple machine learning models. The approach takes the meteorological data and air quality data for the past six days as one batch of input (the whole data set is for 46 days) and employs a multi-model fusion to provide an improved 24-hour prediction of PM2.5 pollutant concentration all over Beijing. During the above process, two focal feature groups are composed. The first focal feature group contains the historical meteorological data, while the second group includes the statistical information, the date information and the polynomial variations. Besides the two groups, we complement one million more data items by employing the time sliding means. Among the supplementary data, we select the most critical 500 features with Light Gradient Boosting Machine (LightGBM) model and send the features as the input to Gradient Boosting Decision Tree (GBDT) and LightGBM models. Meanwhile, we screen the most critical 300 features with eXtreme Gradient Boosting (XGBoost) model and send them as the input to the three prediction models. Referring to each of the models, we respectively gain the optimal parameters through grid search methods and then fuse the models' contribution with the linear weighting. The experiments indicate that the proposed approach based on the weighting fusion is better than that provided by a single modeling scheme, and the loss value is 0.4158 under the SMAPE index. • An approach of predicting air quality with fusing the multiple machine learning models. • The machine learning solutions to model PM2.5 pollutant concentration distribution feature. • The time sliding means for developing one million more data items. • The fusion of multiple models under linear weighting strategy. The experiments indicate that the approach based on the weighting fusion is better than that provided by the single modeling scheme, and the loss value is 0.4158 under the SMAPE index. [ABSTRACT FROM AUTHOR]

Abstract: This paper reports findings of a recent field study that characterized the on-road emissions of individual diesel vehicles in and around Beijing, China during November and December of 2009. We successfully sampled 230 individual trucks on 4 major expressways around the city as well as 57 individual buses in the city using refined mobile chasing techniques and fast response instruments. Emission factors (EF) for carbon monoxide (CO), black carbon (BC) and particulate matter with diameters less than 0.5 μm (PM0.5) are derived from the measurements, which are consistent with the results from laboratory dynamometer tests. The PM0.5 number emission factor distributions demonstrate consistent bimodal modes with peaks around 10 nm and 80 nm, while the mass emission factor distributions demonstrate a unimodal maximum around 110 nm for a majority of the trucks. The BC emissions are shown to be highly correlated with the mass emission of particles with 100–250 nm diameters, which are in good agreement with the results from previous studies. A number of important policy implications are discussed based on the results from this study. First, we identified “heavy emitters” in the on-road fleet we encountered, finding that 5% of diesel trucks in this sample are responsible for 50% of total BC emissions, and 20% of the trucks are responsible for 50% CO and PM0.5 number emissions, 60% PM0.5 mass emissions and over 70% of BC emissions. This suggests that emissions control programs should include identifying and removing heavy emitters from the road or improving their emissions. Second, the BC and PM0.5 number emission factors of trucks registered in regions outside Beijing are significantly higher than those of Beijing-registered trucks, suggesting that improving engine and fuel standards in Beijing alone is not sufficient in reducing the traffic-related air pollution in Beijing. Third, the significantly lower emissions from Euro IV and CNG buses compared to the Euro II and Euro III buses support that introducing clean-burning buses is an effective way to reduce the overall vehicle emissions for the pollutants studied. Fourth, our refined chasing method demonstrates a cost-effective approach to characterize the emissions from a large number of on-road diesel vehicles. Further studies with large sample sizes are critical to constructing more accurate mobile emission inventories to represent the on-road vehicle population. [Copyright &y& Elsevier]

Abstract: Air pollutant emission inventory is an important input parameter for chemical transport models (CTMs). Since great uncertainties exist in the emission inventory, further improvements and refinements are required. In this paper, genetic algorithm (GA), a global search and optimization method, was applied to optimize the emission inventory for the Models-3/Community Multiscale Air Quality (CMAQ) model. An emission optimizing system based on GA was developed and embedded to the CMAQ through the design of several core modules, which implemented the basic functions such as emission adjusting, GA population initializing, CMAQ results evaluating and GA operating. Hypothetical and real-data experiments were respectively performed to examine the validity of GA for emission calibrating. GA showed good performance in both experiments and was always able to find the global minimum. The emission optimizing system was then used to calibrate seasonal PM10 emission inventories of Beijing. Results revealed that PM10 emission in Beijing was underestimated in 2002, an average of 62.74% higher adjustment factor should be imposed on the original emission in target months of different seasons. With the calibrated emission inventories, CMAQ model errors were decreased by 6.46% on average in different seasons. It was concluded that GA was a promising search technique in calibrating emission inputs for CTMs. [ABSTRACT FROM AUTHOR]

Abstract: Valuation of health effects of air pollution is becoming a critical component of the performance of cost–benefit analysis of pollution control measures, which provides a basis for setting priorities for action. Beijing has focused on control of transport emission as vehicular emissions have recently become an important source of air pollution, particularly during Olympic games and Post-games. In this paper, we conducted an estimation of health effects and economic cost caused by road transport-related air pollution using an integrated assessment approach which utilizes air quality model, engineering, epidemiology, and economics. The results show that the total economic cost of health impacts due to air pollution contributed from transport in Beijing during 2004–2008 was 272, 297, 310, 323, 298 million US$ (mean value), respectively. The economic costs of road transport accounted for 0.52, 0.57, 0.60, 0.62, and 0.58% of annual Beijing GDP from 2004 to 2008. Average cost per vehicle and per ton of PM10 emission from road transport can also be estimated as 106 US $/number and 3584 US $ t−1, respectively. These findings illustrate that the impact of road transport contributed particulate air pollution on human health could be substantial in Beijing, whether in physical and economic terms. Therefore, some control measures to reduce transport emissions could lead to considerable economic benefit. [Copyright &y& Elsevier]

Abstract: The aggressive traffic interventions and emission control measures implemented to improve air quality during the 2008 Beijing Olympic Games created a valuable case study to evaluate the effectiveness of these measures on mitigating air pollution and protecting public health. In this paper, we report the results from our field campaign in summer 2008 on the on-road emission factors of carbon monoxide, black carbon (BC) and ultrafine particles (UFP) as well as the ambient BC concentrations. The fleet average emission factors for light-duty gasoline vehicles (LDGV) showed considerable reduction in the Olympic year (2008) compared to the pre-Olympic year (2007). Our measurement of Black Carbon (BC), a primary pollutant, at different elevations at the ambient site suggests consistent decrease in BC concentrations as the height increased near the ground level, which indicates that the nearby ground level sources, probably dominated by traffic, contributed to a large portion of BC concentrations in the lower atmospheric layer in Beijing during summertime. These observations indicate that people living in near ground levels experience higher exposures than those living in higher floors in Beijing. The BC diurnal patterns on days when traffic control were in place during the Olympic Games were compared to those on non-traffic-control days in both 2007 and in 2008. These patterns strongly suggest that diesel trucks are a major source of summertime BC in Beijing. The median BC concentration on Olympic days was 3.7 μg m−3, which was dramatically lower than the value on non-traffic-control days, indicating the effectiveness of traffic control regulations in BC reduction in Beijing. [Copyright &y& Elsevier]

In this paper, we report the results and analysis of a recent field campaign in August 2007 investigating the impacts of emissions from transportation on air quality and community concentrations in Beijing, China. We conducted measurements in three different environments, on-road, roadside and ambient. The carbon monoxide, black carbon and ultrafine particle number emission factors for on-road light-duty vehicles are derived to be 95gkg−1-fuel, 0.3gkg−1-fuel and 1.8×1015 particleskg−1-fuel, respectively. The emission factors for on-road heavy-duty vehicles are 50gkg−1-fuel, 1.3gkg−1-fuel and 1.1×1016 particleskg−1-fuel, respectively. The carbon monoxide emission factors from this study agree with those derived from remote sensing and on-board vehicle emission testing systems in China. The on-road black carbon and particle number emission factors for Chinese vehicles are reported for the first time in the literature. Strong traffic impacts can be observed from the concentrations measured in these different environments. Most clear is a reflection of diesel truck traffic activity in black carbon concentrations. The comparison of the particle size distributions measured at the three environments suggests that the traffic is a major source of ultrafine particles. A four-day traffic control experiment conducted by the Beijing Government as a pilot to test the effectiveness of proposed controls was found to be effective in reducing extreme concentrations that occurred at both on-road and ambient environments. [Copyright &y& Elsevier]

In this paper, an integrated MM5–CMAQ modeling approach was employed to investigate the PM10 air pollution issue in Beijing, China, with a focus on assessing pollution contributions from surrounding provinces. A 2-level-nested grid domain with spatial resolutions of 36 and 12km was designed for the study region. Seven monitoring stations across Beijing municipality were selected to provide hourly PM10 measurement data. The months of January, April, July and October in 2002 were taken as target periods for model performance evaluation. Five emission scenarios were designed and run in order to quantitatively assess the trans-boundary PM10 contributions. The results show that, while Beijing needs to take positive steps to reduce its own pollution emissions, much effort should also be placed on demanding more pollution reduction and better environmental performance from surrounding provinces. [Copyright &y& Elsevier]

Abstract: Beijing, the capital of China, is a rapidly developing mega-city with a huge population in the world. Beijing has been experiencing severe air pollution problems since 1980s due to more and more energy consumption. In the case of severe organic aerosol pollution in Beijing, a one-year-long observation on major organic compounds in PM2.5 was performed at an urban site to gain information of their ambient concentration levels and sources in Beijing. The compound classes identified by GC/MS included alkanes, PAHs, fatty acids, dicarboxylic acids, and hopanes, whose seasonal average concentrations during the year were 39–477, 9–283, 260–483, 16–42, and 2.3–16ngm−3, respectively. Since the molecular distributions of particulate organic matter are indicative of different emission sources, their seasonal characteristics and variations were discussed in detail to explore the major contributing sources in Beijing. Conclusively, organic aerosol pollution in Beijing was found to be mostly associated with vehicle exhaust, coal combustion for residential heating, and cooking emissions, and the latter two sources are unique Chinese pollution sources. The dataset in this paper provides useful information for establishing effective control measures to reduce organic aerosol pollution in Beijing. [Copyright &y& Elsevier]