Spatiotemporal trend analysis for fine particulate matter concentrations in China using high-resolution satellite-derived and ground-measured PM2.5 data

Bai, Kaixu; Минг, Ма; Chang, Ni‐Bin; Gao, Wei

doi:10.1016/j.jenvman.2018.12.071

Cited by 68 publications

(32 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, air pollution from regional transport has received considerable attention, but most studies aiming to predict the distribution of PM2.5-related cancers have only focused on local air pollution (Chang et al, 2018;Guo et al, 2016;Han et al, 2017;Wang et al, 2015). In this study, we found the three regions where high concentrations of PM2.5 are mainly located (i.e., BTH, YRD and NEC), which is a finding consistent with previous studies (Bai et al, 2019;He et al, 2019;Hu et al, 2014). Although the constitutes of PM2.5 are different in the BTH, it was found that an average of approximately 34% of PM2.5 in Beijing could be attributed to outside sources-even Hebei province contributed 50-70% of the PM2.5 concentration in Beijing during the period of the eastern and southern winds (Gao et al, 2014;Gao et al, 2018).…”

Section: Discussionsupporting

confidence: 89%

County level mortality and morbidity associated with PM2.5-related cancers in China based on satellite-derived PM2.5 data from 1998 to 2016: a forecasting and spatial analysis

Liao

Zhou

Gao

et al. 2019

Preprint

View full text Add to dashboard Cite

Background: Ambient fine particulate matter (PM2.5) pollution is becoming increasingly serious in China, and is a major risk factor for various cancers. Moreover, recent studies have indicated that there is an impact of PM2.5 from regional transport. We aimed to predict spatial distribution of cancers related to PM2.5, and draw a series of forecasting maps of these cancers. Methods: A 1000-loops simulation was done to choose the optimal forecasting model between five alternative models: ridge regression, partial least square regression, regression tree, model tree, and the combined forecasting model. Then a kriging interpolation method was used to draw the forecasted cancer maps. Results: The trend showed a gradual increase in the mortality and morbidity of breast cancer, pancreatic cancer, and all-cause cancer. We found a significant spatial autocorrelation between cancer incidence and PM2.5. Our results from forecasting showed a constant growth in mortality and morbidity of all cancers, and the kriging method suggested a similar spatial pattern. High morbidity and mortality areas were mainly in central-east and south-east China. We found a similar distribution pattern between PM2.5 concentration and mortality and morbidity associated with PM2.5-related cancers. Conclusions: These findings serve as a valuable reference for the development of effective policies to reduce air pollution emissions, with the efforts from governments in high-risk areas.

show abstract

Section: Discussionsupporting

confidence: 89%

County level mortality and morbidity associated with PM2.5-related cancers in China based on satellite-derived PM2.5 data from 1998 to 2016: a forecasting and spatial analysis

Liao

Zhou

Gao

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…Moreover, the diurnal cycle reconstructed from the neighborhood field in space is more accurate than that using PM 2.5 observations from nearterm days, which is evidenced by smaller correlation values with limited neighboring stations. Such an effect is also in line with our recent results when comparing the beneficial effects of spatial and temporal neighboring terms in advancing gridded PM 2.5 concentration mapping (Bai et al, 2019c). Figure 10 shows the benefits of the DCCEOF method to our retrieved in situ hourly PM 2.5 concentration record at each individual monitoring station in terms of the improvement of the data completeness ratio as well as the reduction of gap frequency.…”

Section: Performance Of the Dcceof Methodssupporting

confidence: 89%

“…Considering a too-compact neighborhood field may be inadequate to reconstruct the local diurnal cycle of PM 2.5 fairly due to limited valid samples (because missingness may also emerge in each candidate 24 h PM 2.5 concentration record); here m was defined as the number of stations within 100 km (spatial window size) of the target station, while n was set to 7 (temporal window size) in our case. The spatial and temporal window sizes used here are based on our recent results in which an optimal window size of 50 km and 3 d was found to attain a good autocorrelation of PM 2.5 concentration in space and time, respectively (Bai et al, 2019c). To have adequate samples for the construction of X m,n p,t , here we enlarged both the window sizes by simply doubling the values found in our previous study.…”

Section: Introductionmentioning

confidence: 99%

Filling the gaps of in situ hourly PM<sub>2.5</sub> concentration data with the aid of empirical orthogonal function analysis constrained by diurnal cycles

Bai

Guo

et al. 2020

Atmos. Meas. Tech.

Self Cite

View full text Add to dashboard Cite

Abstract. Data gaps in surface air quality measurements significantly impair the data quality and the exploration of these valuable data sources. In this study, a novel yet practical method called diurnal-cycle-constrained empirical orthogonal function (DCCEOF) was developed to fill in data gaps present in data records with evident temporal variability. The hourly PM2.5 concentration data retrieved from the national ambient air quality monitoring network in China were used as a demonstration. The DCCEOF method aims to reconstruct the diurnal cycle of PM2.5 concentration from its discrete neighborhood field in space and time firstly and then predict the missing values by calibrating the reconstructed diurnal cycle to the level of valid PM2.5 concentrations observed at adjacent times. The statistical results indicate a high frequency of data gaps in our retrieved hourly PM2.5 concentration record, with PM2.5 concentration measured on about 40 % of the days suffering from data gaps. Further sensitivity analysis results reveal that data gaps in the hourly PM2.5 concentration record may introduce significant bias to its daily averages, especially during clean episodes at which PM2.5 daily averages are observed to be subject to larger uncertainties compared to the polluted days (even in the presence of the same amount of missingness). The cross-validation results indicate that our suggested DCCEOF method has a good prediction accuracy, particularly in predicting daily peaks and/or minima that cannot be restored by conventional interpolation approaches, thus confirming the effectiveness of the consideration of the local diurnal variation pattern in gap filling. By applying the DCCEOF method to the hourly PM2.5 concentration record measured in China from 2014 to 2019, the data completeness ratio was substantially improved while the frequency of days with gapped PM2.5 records reduced from 42.6 % to 5.7 %. In general, our DCCEOF method provides a practical yet effective approach to handle data gaps in time series of geophysical parameters with significant diurnal variability, and this method is also transferable to other data sets with similar barriers because of its self-consistent capability.

show abstract

“…Such an endeavour helps users to acquire hourly air quality data more efficiently. The retrieved hourly mass concentration record, taking PM2.5 for instance, has been widely used as a critical data source in many studies related to haze pollutions, because of its good accuracy and high temporal resolution as well as its national-scale coverage (Gao et al, 2018;Miao et al, 2018;Bai et al, 2019aBai et al, , 2019bZhang et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

“…resolution (e.g., Bai et al, 2019b;Zhang et al, 2019). On the other hand, many previous studies preferred to exclude records of days with a certain degree of missing values (e.g., no more than 6 missing values within 24-h) from their analysis (e.g., van Donkelaar et al, 2016;Li et al, 2017;Huang et al, 2018;Manning et al, 2018;Shen et al, 2018;Bai et al, 2019a;Zhang et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

Filling the gaps of in-situ hourly PM<sub>2.5</sub> concentration data with the aid of empirical orthogonal function constrained by diurnal cycles

Bai¹,

Li²,

Guo³

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

Abstract. Data gaps are frequently observed in the hourly PM2.5 mass concentration records measured from the China national air quality monitoring network. In this study, we proposed a novel gap filling method called the diurnal cycle constrained empirical orthogonal function (DCCEOF) to fill in data gaps present in hourly PM2.5 concentration records. This method mainly calibrates the diurnal cycle of PM2.5 that is reconstructed from discrete PM2.5 neighborhood fields in space and time to the level of valid PM2.5 data values observed at adjacent times. Prior to gap filling, possible impacts of varied number of data gaps in the time series of hourly PM2.5 concentration on PM2.5 daily averages were examined via sensitivity experiments. The results showed that PM2.5 data suffered from the gaps on about 40% of days, indicating a high frequency of missing data in the hourly PM2.5 records. These gaps could introduce significant bias to daily-averaged PM2.5. Particularly, given the same number of gaps, larger biases would be introduced to daily-averaged PM2.5 during clean days than polluted days. The cross-validation results indicate that the predicted missing values from the DCCEOF method with the consideration of the local diurnal phases of PM2.5 are more accurate and reasonable than those from the conventional spline interpolation approach, especially for the reconstruction of daily peaks and/or minima that cannot be restored by the latter method. To fill the gaps in the hourly PM2.5 records across China during 2014 to 2019, as a practical application, the DCCEOF method can be able to reduce the averaged frequency of missingness from 42.6 % to 5.7 %. In general, the present work implies that the DCCEOF method is realistic and robust to be able to handle the missingness issues in time series of geophysical parameters with significant diurnal variability and can be expectably applied in other data sets with similar barriers because of its self-consistent capability.

show abstract

Spatiotemporal trend analysis for fine particulate matter concentrations in China using high-resolution satellite-derived and ground-measured PM2.5 data

Cited by 68 publications

References 37 publications

County level mortality and morbidity associated with PM2.5-related cancers in China based on satellite-derived PM2.5 data from 1998 to 2016: a forecasting and spatial analysis

County level mortality and morbidity associated with PM2.5-related cancers in China based on satellite-derived PM2.5 data from 1998 to 2016: a forecasting and spatial analysis

Filling the gaps of in situ hourly PM<sub>2.5</sub> concentration data with the aid of empirical orthogonal function analysis constrained by diurnal cycles

Filling the gaps of in-situ hourly PM<sub>2.5</sub> concentration data with the aid of empirical orthogonal function constrained by diurnal cycles

Contact Info

Product

Resources

About

Spatiotemporal trend analysis for fine particulate matter concentrations in China using high-resolution satellite-derived and ground-measured PM2.5 data

Cited by 68 publications

References 37 publications

County level mortality and morbidity associated with PM2.5-related cancers in China based on satellite-derived PM2.5 data from 1998 to 2016: a forecasting and spatial analysis

County level mortality and morbidity associated with PM2.5-related cancers in China based on satellite-derived PM2.5 data from 1998 to 2016: a forecasting and spatial analysis

Filling the gaps of in situ hourly PM&lt;sub&gt;2.5&lt;/sub&gt; concentration data with the aid of empirical orthogonal function analysis constrained by diurnal cycles

Filling the gaps of in-situ hourly PM&lt;sub&gt;2.5&lt;/sub&gt; concentration data with the aid of empirical orthogonal function constrained by diurnal cycles

Contact Info

Product

Resources

About

Filling the gaps of in situ hourly PM<sub>2.5</sub> concentration data with the aid of empirical orthogonal function analysis constrained by diurnal cycles

Filling the gaps of in-situ hourly PM<sub>2.5</sub> concentration data with the aid of empirical orthogonal function constrained by diurnal cycles