Estimating Regional PM<sub>2.5</sub> Concentrations in China Using a Global-Local Regression Model Considering Global Spatial Autocorrelation and Local Spatial Heterogeneity
Linear regression models are commonly used for estimating ground PM<sub>2.5</sub> concentrations, but the global spatial autocorrelation and local spatial heterogeneity of PM<sub>2.5</sub> distribution are either ignored or only partially considered in commonly used models fo...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-09-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/18/4545 |
| _version_ | 1797482783115837440 |
|---|---|
| author | Heng Su Yumin Chen Huangyuan Tan Annan Zhou Guodong Chen Yuejun Chen |
| author_facet | Heng Su Yumin Chen Huangyuan Tan Annan Zhou Guodong Chen Yuejun Chen |
| author_sort | Heng Su |
| collection | DOAJ |
| description | Linear regression models are commonly used for estimating ground PM<sub>2.5</sub> concentrations, but the global spatial autocorrelation and local spatial heterogeneity of PM<sub>2.5</sub> distribution are either ignored or only partially considered in commonly used models for estimating PM<sub>2.5</sub> concentrations. Therefore, taking both global spatial autocorrelation and local spatial heterogeneity into consideration, a global-local regression (GLR) model is proposed for estimating ground PM<sub>2.5</sub> concentrations in the Yangtze River Delta (YRD) and in the Beijing, Tianjin, Hebei (BTH) regions of China based on the aerosol optical depth data, meteorological data, remote sensing data, and pollution source data. Considering the global spatial autocorrelation, the GLR model extracts global factors by the eigenvector spatial filtering (ESF) method, and combines the fraction of them that passes further filtering with the geographically weighted regression (GWR) method to address the local spatial heterogeneity. Comprehensive results show that the GLR model outperforms the ordinary GWR and ESF models, and the GLR model has the best performance at the monthly, seasonal, and annual levels. The average adjusted R<sup>2</sup> of the monthly GLR model in the YRD region (the BTH region) is 0.620 (0.853), which is 8.0% and 7.4% (6.8% and 7.0%) higher than that of the monthly ESF and GWR models, respectively. The average cross-validation root mean square error of the monthly GLR model is 7.024 μg/m<sup>3</sup> in the YRD region, and 9.499 μg/m<sup>3</sup> in the BTH region, which is lower than that of the ESF and GWR models. The GLR model can effectively address the spatial autocorrelation and spatial heterogeneity, and overcome the shortcoming of the ordinary GWR model that overfocuses on local features and the disadvantage of the poor local performance of the ordinary ESF model. Overall, the GLR model with good spatial and temporal applicability is a promising method for estimating PM<sub>2.5</sub> concentrations. |
| first_indexed | 2024-03-09T22:38:25Z |
| format | Article |
| id | doaj.art-1c887178d1eb46e9aa27557dcd36decb |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T22:38:25Z |
| publishDate | 2022-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-1c887178d1eb46e9aa27557dcd36decb2023-11-23T18:44:24ZengMDPI AGRemote Sensing2072-42922022-09-011418454510.3390/rs14184545Estimating Regional PM<sub>2.5</sub> Concentrations in China Using a Global-Local Regression Model Considering Global Spatial Autocorrelation and Local Spatial HeterogeneityHeng Su0Yumin Chen1Huangyuan Tan2Annan Zhou3Guodong Chen4Yuejun Chen5School of Resource and Environment Science, Wuhan University, Wuhan 430079, ChinaSchool of Resource and Environment Science, Wuhan University, Wuhan 430079, ChinaSchool of Resource and Environment Science, Wuhan University, Wuhan 430079, ChinaSchool of Resource and Environment Science, Wuhan University, Wuhan 430079, ChinaSchool of Resource and Environment Science, Wuhan University, Wuhan 430079, ChinaSchool of Resource and Environment Science, Wuhan University, Wuhan 430079, ChinaLinear regression models are commonly used for estimating ground PM<sub>2.5</sub> concentrations, but the global spatial autocorrelation and local spatial heterogeneity of PM<sub>2.5</sub> distribution are either ignored or only partially considered in commonly used models for estimating PM<sub>2.5</sub> concentrations. Therefore, taking both global spatial autocorrelation and local spatial heterogeneity into consideration, a global-local regression (GLR) model is proposed for estimating ground PM<sub>2.5</sub> concentrations in the Yangtze River Delta (YRD) and in the Beijing, Tianjin, Hebei (BTH) regions of China based on the aerosol optical depth data, meteorological data, remote sensing data, and pollution source data. Considering the global spatial autocorrelation, the GLR model extracts global factors by the eigenvector spatial filtering (ESF) method, and combines the fraction of them that passes further filtering with the geographically weighted regression (GWR) method to address the local spatial heterogeneity. Comprehensive results show that the GLR model outperforms the ordinary GWR and ESF models, and the GLR model has the best performance at the monthly, seasonal, and annual levels. The average adjusted R<sup>2</sup> of the monthly GLR model in the YRD region (the BTH region) is 0.620 (0.853), which is 8.0% and 7.4% (6.8% and 7.0%) higher than that of the monthly ESF and GWR models, respectively. The average cross-validation root mean square error of the monthly GLR model is 7.024 μg/m<sup>3</sup> in the YRD region, and 9.499 μg/m<sup>3</sup> in the BTH region, which is lower than that of the ESF and GWR models. The GLR model can effectively address the spatial autocorrelation and spatial heterogeneity, and overcome the shortcoming of the ordinary GWR model that overfocuses on local features and the disadvantage of the poor local performance of the ordinary ESF model. Overall, the GLR model with good spatial and temporal applicability is a promising method for estimating PM<sub>2.5</sub> concentrations.https://www.mdpi.com/2072-4292/14/18/4545PM<sub>2.5</sub>global-local regressiongeographically weighted regressioneigenvector spatial filteringYRD regionBTH region |
| spellingShingle | Heng Su Yumin Chen Huangyuan Tan Annan Zhou Guodong Chen Yuejun Chen Estimating Regional PM<sub>2.5</sub> Concentrations in China Using a Global-Local Regression Model Considering Global Spatial Autocorrelation and Local Spatial Heterogeneity Remote Sensing PM<sub>2.5</sub> global-local regression geographically weighted regression eigenvector spatial filtering YRD region BTH region |
| title | Estimating Regional PM<sub>2.5</sub> Concentrations in China Using a Global-Local Regression Model Considering Global Spatial Autocorrelation and Local Spatial Heterogeneity |
| title_full | Estimating Regional PM<sub>2.5</sub> Concentrations in China Using a Global-Local Regression Model Considering Global Spatial Autocorrelation and Local Spatial Heterogeneity |
| title_fullStr | Estimating Regional PM<sub>2.5</sub> Concentrations in China Using a Global-Local Regression Model Considering Global Spatial Autocorrelation and Local Spatial Heterogeneity |
| title_full_unstemmed | Estimating Regional PM<sub>2.5</sub> Concentrations in China Using a Global-Local Regression Model Considering Global Spatial Autocorrelation and Local Spatial Heterogeneity |
| title_short | Estimating Regional PM<sub>2.5</sub> Concentrations in China Using a Global-Local Regression Model Considering Global Spatial Autocorrelation and Local Spatial Heterogeneity |
| title_sort | estimating regional pm sub 2 5 sub concentrations in china using a global local regression model considering global spatial autocorrelation and local spatial heterogeneity |
| topic | PM<sub>2.5</sub> global-local regression geographically weighted regression eigenvector spatial filtering YRD region BTH region |
| url | https://www.mdpi.com/2072-4292/14/18/4545 |
| work_keys_str_mv | AT hengsu estimatingregionalpmsub25subconcentrationsinchinausingagloballocalregressionmodelconsideringglobalspatialautocorrelationandlocalspatialheterogeneity AT yuminchen estimatingregionalpmsub25subconcentrationsinchinausingagloballocalregressionmodelconsideringglobalspatialautocorrelationandlocalspatialheterogeneity AT huangyuantan estimatingregionalpmsub25subconcentrationsinchinausingagloballocalregressionmodelconsideringglobalspatialautocorrelationandlocalspatialheterogeneity AT annanzhou estimatingregionalpmsub25subconcentrationsinchinausingagloballocalregressionmodelconsideringglobalspatialautocorrelationandlocalspatialheterogeneity AT guodongchen estimatingregionalpmsub25subconcentrationsinchinausingagloballocalregressionmodelconsideringglobalspatialautocorrelationandlocalspatialheterogeneity AT yuejunchen estimatingregionalpmsub25subconcentrationsinchinausingagloballocalregressionmodelconsideringglobalspatialautocorrelationandlocalspatialheterogeneity |