<trans-title>Estimating Ground-Level PM<sub>2.5</sub> Concentrations Across China Using Geographically Neural Network Weighted Regression</trans-title>

Zhenhong DU; Sensen WU; Zhongyi WANG; Yuanyuan WANG; Feng ZHANG; Renyi LIU

doi:10.12082/dqxxkx.2020.190533

Journal of Geo-information Science, Volume. 22, Issue 1, 122(2020)

Estimating Ground-Level PM_2.5 Concentrations Across China Using Geographically Neural Network Weighted Regression

Zhenhong DU^1,1,2,2、*, Sensen WU^1,1,2,2, Zhongyi WANG^1,1, Yuanyuan WANG^1,1,2,2, Feng ZHANG^1,1,2,2, and Renyi LIU^1,1,2,2

Author Affiliations

¹School of Earth Sciences, Zhejiang University, Hangzhou 310027, China

¹浙江大学地球科学学院,地理与空间信息研究所,杭州 310027

²Zhejiang Provincial Key Laboratory of Geographic Information Science, Hangzhou 310028, China

²浙江省资源与环境信息系统重点实验室,杭州 310028

show less

Abstract Get PDF(in Chinese)

Figures & Tables(15)

Fig. 1. Spatial distribution of the PM_2.5 monitoring stations of China in 2017 and the spatial partitions of the training, validation, and testing datasets

Download full size

View in Article

Fig. 2. Pre-processing of the experimental dataset

Download full size

View in Article

Fig. 3. Definition of the GNNWR model for PM_2.5 estimation

Download full size

View in Article

Fig. 4. Structure design of the spatial weighted neural network

Download full size

View in Article

Fig. 5. Training and validation procedures of the GNNWR model

Download full size

View in Article

Fig. 6. Performance variations for the training and validation datasets of the GNNWR model

Download full size

View in Article

Fig. 7. Estimates of the annual mean PM_2.5 across China in 2017

Download full size

View in Article

Fig. 8. Details comparison of the annual mean PM_2.5 estimates between the GWR and GNNWR models

Download full size

View in Article

Fig. 9. Spatial distribution of absolute estimation errors of the annual mean PM_2.5 of China in 2017 for the OLR, GWR, and GNNWR models

Download full size

View in Article

Table 1. Data sources and description

View table

View in Article

Table 1. Data sources and description

数据类型	数据名称	变量名称	时间分辨率	空间分辨率	数据来源
站点	PM_2.5监测站点	PM_2.5	h	-	中国国家气象局
遥感	气溶胶	AOD	d	3 km、10 km	LAADS
气象	2 m温度	TEMP	h	0.5°	ERA5 hourly data
	降水量	TP	h	0.5°	ERA5 hourly data
	10 m风速	WS	h	0.5°	ERA5 hourly data
	10 m风向	WD	h	0.5°	ERA5 hourly data
地理	地形	DEM	-	1弧分	NOAA

Table 2. Settings of GWR models
View table
View in Article
Table 2. Settings of GWR models
模型名称带宽优化准则核函数
类型结构
GWR-AFG AICc 固定型 Gaussian
GWR-AAB AICc 适应型 Bi-square

Table 3. Settings of architectures and hyper-parameters for the GNNWR model
View table
View in Article
Table 3. Settings of architectures and hyper-parameters for the GNNWR model
输入层隐含层1 隐含层2 隐含层3 输出层
1004 512 256 128 7
学习率 epoch最大值批处理大小 Dropout
0.2 20 000 64 0.8

Table 4. Exploratory analysis and descriptive statistics of the experimental dataset across China in 2017

View table

View in Article

Table 4. Exploratory analysis and descriptive statistics of the experimental dataset across China in 2017

变量	PM_2.5/ $(μg / m^{3})$	AOD	DEM $/ m$	温度/ $K$	降水量/ $m$	风速/ $(m / s)$	风向/°
相关系数	-	0.564	-0.345	0.135	-0.234	-0.373	-0.461
显著性水平	-	0.001	0.001	0.001	0.001	0.001	0.001
方差膨胀因子	-	3.005	1.636	4.639	2.107	2.357	4.184
平均值	46.07	0.540	393.460	288.000	8.95E-05	7.080	147.990
标准差	15.56	0.180	658.460	5.170	4.82E-05	0.940	43.150
最小值	8.34	0.070	-5.250	271.810	1.06E-06	4.430	80.110
最大值	103.89	1.140	4539.960	297.720	2.34E-04	11.500	236.040

Table 5. Fitting and prediction performances of the PM_2.5 estimates for the OLR, GWR, and GNNWR models

View table

View in Article

Table 5. Fitting and prediction performances of the PM_2.5 estimates for the OLR, GWR, and GNNWR models

模型	训练集（拟合精度）							测试集（预测精度）
模型	R²	RMSE	MAE	MAPE/%	AICc	F₁	p-value	R²	RMSE	MAE	MAPE/%
OLR	0.517	10.578	8.132	20.2	7579.202	-	-	0.477	12.023	8.838	19.7
GWR-AFG	0.759	7.478	5.797	14.5	6999.491	0.538	0.010	0.683	9.359	6.963	16.1
GWR-AAB	0.898	4.860	3.478	8.7	6673.537	0.297	0.010	0.675	9.484	5.423	12.6
GNNWR	0.914	4.452	3.235	7.9	5833.002	0.137	0.010	0.831	6.837	4.688	11.0

Table 6. Spatial nonstationarity diagnosis of each variable in the GNNWR model

View table

View in Article

Table 6. Spatial nonstationarity diagnosis of each variable in the GNNWR model

变量	GWR-AFG		GWR-AAB		GNNWR
变量	F₂	p-value	F₂	p-value	F₂	p-value
常量项	2407.9	0.001	1269.1	0.001	590.5	0.001
AOD	1379.7	0.001	1199.4	0.001	243.7	0.001
DEM	1983.9	0.001	1272.2	0.001	146.7	0.001
温度	3783.7	0.001	1229.6	0.001	127.7	0.001
降水量	1237.4	0.001	1170.1	0.001	214.7	0.001
风速	1626.0	0.001	1016.2	0.001	159.0	0.001
风向	2379.8	0.001	1081.6	0.001	73.6	0.001

Tools

Get Citation

Copy Citation Text

Zhenhong DU, Sensen WU, Zhongyi WANG, Yuanyuan WANG, Feng ZHANG, Renyi LIU. Estimating Ground-Level PM_2.5 Concentrations Across China Using Geographically Neural Network Weighted Regression[J]. Journal of Geo-information Science, 2020, 22(1): 122

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Received: Sep. 22, 2019

Accepted: --

Published Online: Sep. 16, 2020

The Author Email: Zhenhong DU (duzhenhong@zju.edu.cn)

DOI:10.12082/dqxxkx.2020.190533

Topics

Table 1. Data sources and description

Table 1. Data sources and description

Table 2. Settings of GWR models

Table 2. Settings of GWR models

Table 3. Settings of architectures and hyper-parameters for the GNNWR model

Table 3. Settings of architectures and hyper-parameters for the GNNWR model

Table 4. Exploratory analysis and descriptive statistics of the experimental dataset across China in 2017

Table 4. Exploratory analysis and descriptive statistics of the experimental dataset across China in 2017

Table 5. Fitting and prediction performances of the PM2.5 estimates for the OLR, GWR, and GNNWR models

Table 5. Fitting and prediction performances of the PM2.5 estimates for the OLR, GWR, and GNNWR models

Table 6. Spatial nonstationarity diagnosis of each variable in the GNNWR model

Table 6. Spatial nonstationarity diagnosis of each variable in the GNNWR model

Table 5. Fitting and prediction performances of the PM_2.5 estimates for the OLR, GWR, and GNNWR models

Table 5. Fitting and prediction performances of the PM_2.5 estimates for the OLR, GWR, and GNNWR models