Performance Comparison of Artificial Neural Network and Kriging in Spatial Estimation of PM2.5 Concentration
-
摘要: 针对人工神经网络与克里格插值在PM2.5浓度空间估算中精度随样本点数量与耦合因素不同差异较大的问题, 基于相关分析与径向基函数(radical basis function, RBF)筛选PM2.5空间变异关键影响因素, 对比不同比例训练样本下普通克里格插值(ordinary Kriging, OK), 仅考虑地理坐标RBF神经网络, 耦合关键因素的协同克里格插值(CoKriging, CK)及RBF神经网络(CoRBF)的效果差异, 并基于最优方法开展PM2.5浓度空间制图。结果表明:4种方法均能有效实现PM2.5浓度空间估算, 且精度随训练样本比例增大而波动上升。考虑关键因素人口密度的CoRBF最能表现数据变化趋势, 而CK在误差指标上更优越。基于CK与CoRBF的PM2.5浓度空间估算结果较好展示了污染的分异特征, 前者较后者更平滑。Abstract: Performance of artificial neural network modeling and Kriging interpolation in PM2.5 concentration estimation varies with sample sizes and predictor variables change. This paper analyzes the performance of ordinary Kriging (OK), radical basis function (RBF) networks based on geographic coordinates, CoKriging and RBF with the key factor(s) (CK and CoRBF) selected by correlation analysis and RBF network, using different training sets with various sizes. The spatial distribution of PM2.5 concentration is then estimated by the best performed method. Results show that RBF, CoRBF, OK, and CK can all be used to estimate PM2.5 concentration efficiently, and their accuracies improved unstably as the number of training sites increase. CoRBF with the key factor of population illustrates the largest variation of PM2.5 concentration, while CK has the highest coefficient of determination (R2) and index of agreement (IOA) and the lowest mean square error (MSE), mean absolute error (MAE), and relative error (RE). Correspondingly, the spatial pattern of CK estimated PM2.5 concentration is smoother than CoRBF estimated PM2.5 concentration, while they both are very similar to site measurements and reveal detailed information.
-
-
![]()
图 2 检验样本的PM2.5浓度观测值与各方法估算值
Figure 2. Observations and Estimations of PM2.5 in Validation Sets Based on Four Methods
![]()
图 3 基于各种方法的PM2.5浓度估算值与观测值间的精度对比
Figure 3. Accuracy Comparison of Between Observations and Estimations of PM2.5 Based on Four Methods
表 1 基于RBF网络的关键影响因素筛选
Table 1 Selection of Key Factors Based on RBF Network
关键影响因素 MSE/(μg·m-3) MAE/(μg·m-3) RE/% R2 X、Y 0.91 0.54 6.11 0.67 X、Y、人口密度 0.42 0.52 4.14 0.90 X、Y、降水 0.95 0.78 7.55 0.76 X、Y、道路总长度 1.26 0.89 7.98 0.55 X、Y、人口密度、降水 1.08 0.86 8.05 0.63 X、Y、人口密度、道路总长度 1.11 0.84 8.50 0.66 X、Y、降水、道路总长度 0.85 0.73 6.12 0.86 X、Y、人口密度、降水、道路总长度 0.87 0.85 8.25 0.82 
下载: 导出CSV
-
[1] Cohen A J, Brauer M, Burnett R, et al. Estimates and 25-Year Trends of the Global Burden of Disease Attributable to Ambient Air Pollution: An Analysis of Data from the Global Burden of Diseases Study 2015[J]. The Lancet, 2017, 389(10 082):1 907-1 918 https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(17)30505-6/fulltext
[2] Di Q, Wang Y, Zanobetti A, et al. Air Pollution and Mortality in the Medicare Population[J]. New England Journal of Medicine, 2017, 376(26): 2 513-2 522 doi: 10.1056/NEJMoa1702747
[3] Apte J S, Messier K P, Gani S, et al. High-Resolution Air Pollution Mapping with Google Street View Cars: Exploiting Big Data[J]. Environmental Science & Technology, 2017, 51(12): 6 999-7 008 doi: 10.1021/acs.est.7b00891
[4] 李沈鑫, 邹滨, 刘兴权, 等. 2013-2015年中国PM2.5污染状况时空变化[J].环境科学研究, 2017, 30(5): 678-687 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=hjkxyj201705005 Li Shenxin, Zou Bin, Liu Xingquan, et al. Pollution Status and Spatial-Temporal Variations of PM2.5 in China During 2013-2015[J]. Research of Environmental Sciences, 2017, 30(5):678-687 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=hjkxyj201705005
[5] He J, Zhang Y, Wang K, et al. Multi-Year Application of Wrf-Cam5 over East Asia-Part Ⅰ: Comprehensive Evaluation and Formation Regimes of O3 and PM2.5 [J]. Atmospheric Environment, 2017, 165: 122-142 doi: 10.1016/j.atmosenv.2017.06.015
[6] Fang X, Zou B, Liu X, et al. Satellite-Based Ground PM2.5 Estimation Using Timely Structure Adaptive Modeling[J]. Remote Sensing of Environment, 2016, 186: 152-163 doi: 10.1016/j.rse.2016.08.027
[7] Zou B, Pu Q, Bilal M, et al. High-Resolution Satellite Mapping of Fine Particulates Based on Geographically Weighted Regression[J]. IEEE Geoscience & Remote Sensing Letters, 2017, 13(4): 495-499 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=b64b214ee611a8372ed0e871154af967
[8] 焦利民, 许刚, 赵素丽, 等.基于LUR的武汉市PM2.5浓度空间分布模拟[J].武汉大学学报∙信息科学版, 2015, 40(8):1 088-1 094 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=whchkjdxxb201508016 Jiao Limin, Xu Gang, Zhao Suli, et al. LUR-Based Simulation of the Spatial Distribution of PM2.5 of Wuhan[J]. Geomatics and Information Science of Wuhan University, 2015, 40(8):1 088-1 094 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=whchkjdxxb201508016
[9] Zou B, Wang M, Wan N, et al. Spatial Modeling of PM2.5 Concentrations with a Multifactoral Radial Basis Function Neural Network[J]. Environmental Science & Pollution Research, 2015, 22(14): 1-10 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=9e31aa75e7c015257bc0b82c703a73c2
[10] Xiao L, Lang Y, Christakos G. High-Resolution Spatiotemporal Mapping of PM2.5 Concentrations at Mainland China Using a Combined BME-GWR Technique[J]. Atmospheric Environment, 2018, 173: 295-305 doi: 10.1016/j.atmosenv.2017.10.062
[11] Liu Y, Cao G, Zhao N, et al. Improve Ground-Level PM2.5 Concentration Mapping Using A Random Forests-Based Geostatistical Approach[J]. Environmental Pollution, 2017, 235: 272-282 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=f8c786a15a6f1773b2e43de904c53dbd
[12] Mishra D, Goyal P, Upadhyay A. Artificial Intelligence Based Approach to Forecast PM2.5 During Haze Episodes: A Case Study of Delhi, India[J]. Atmospheric Environment, 2015, 102:239-248 doi: 10.1016/j.atmosenv.2014.11.050
[13] Adams M D, Kanaroglou P S. Mapping Real-Time Air Pollution Health Risk for Environmental Management: Combining Mobile and Stationary Air Pollution Monitoring with Neural Network Models[J]. Journal of Environmental Management, 2016, 168:133-141 http://www.ncbi.nlm.nih.gov/pubmed/26706225
[14] Ordieres J B, VergarA E P, Capuz R S, et al. Neural Network Prediction Model for Fine Particulate Matter (PM2.5) on the US-Mexico Border in El Paso (Texas) and Ciudad Juárez (Chihuahua) [J]. Environmental Modelling & Software, 2005, 20(5): 547-559 http://www.sciencedirect.com/science/article/pii/S1364815204000830
[15] Park J, Sandberg I W. Universal Approximation Using Radial-Basis-Function Networks[J]. Neural Computation, 1991, 3(2): 246-257 doi: 10.1162/neco.1991.3.2.246
[16] Stein M L. Interpolation of Spatial Data: Some Theory for Kriging[M]. New York: Springer-Verlag, 1999: 23-27
[17] Myers D E. Matrix Formulation of Co-Kriging[J]. Journal of the International Association for Mathematical Geology, 1982, 14(3): 249-257 doi: 10.1007/BF01032887
[18] Wang J F, Li X H, Christakos G, et al. Geographical Detectors-Based Health Risk Assessment and Its Application in the Neural Tube Defects Study of the Heshun Region, China[J]. International Journal of Geographical Information Science, 2010, 24(1): 107-127 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=f637928d711402f020a016140497a2c7
[19] 邹滨, 许珊, 张静.土地利用视角空气污染空间分异的地理分析[J].武汉大学学报∙信息科学版, 2017, 42(2):216-222 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=whchkjdxxb201702012 Zou Bin, Xu Shan, Zhang Jing. Spatial Variation Analysis of Urban Air Pollution Using GIS: A Land Use Perspective[J]. Geomatics and Information Science of Wuhan University, 2017, 42(2):216-222 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=whchkjdxxb201702012
[20] Guo J, Xia F, Zhang Y, et al. Impact of Diurnal Variability and Meteorological Factors on the PM2.5-Aod Relationship: Implications for PM2.5 Remote Sensing[J]. Environmental Pollution, 2017, 221:94-104 doi: 10.1016/j.envpol.2016.11.043
[21] Caselli M, Trizio L, Gennaro G D, et al. A Simple Feedforward Neural Network for the PM10 Forecasting: Comparison with a Radial Basis Function Network and a Multivariate Linear Regression Model[J]. Water Air & Soil Pollution, 2009, 201(1-4):365-377 doi: 10.1007/s11270-008-9950-2
[22] Asadollahfardi G, Zangooei H, Aria S H. Predicting PM2.5 Concentrations Using Artificial Neural Networks and Markov Chain: A Case Study Karaj City[J]. Asian Journal of Atmospheric Environment, 2016, 10(2):67-79 doi: 10.5572/ajae.2016.10.2.067
[23] Yao L, Lu N. Spatiotemporal Distribution and Short-Term Trends of Particulate Matter Concentration over China, 2006-2010[J]. Environmental Science & Pollution Research International, 2014, 21(16):9 665-9 675 doi: 10.1007/s11356-014-2996-3
-
期刊类型引用(10)
1. 曾广泉，马韬，张孟希，戴妍，陈凯文，丁继辉，俞双恩，王中文. 基于无人机多光谱影像的不同施氮量水稻LAI反演方法研究. 江苏农业科学. 2024(20): 41-48 . 
百度学术
2. 高钰琪，许桂玲，冯跃华，王晓珂，任红军，由晓璇，韩志丽，李家乐. 基于冠层高光谱植被指数的水稻产量预测模型研究. 中国稻米. 2023(05): 38-44 . 百度学术
3. 彭晓伟，张爱军，王楠，赵丽，杨晓楠. 高光谱技术在土壤及适种作物的研究进展. 遥感信息. 2022(01): 32-39 . 百度学术
4. 王晓珂，刘婷婷，许桂玲，冯跃华，彭金凤，李杰，罗强鑫，韩志丽，卢苇，PHONENASAY Somsana. 基于冠层高光谱遥感的杂交水稻植被指数氮素营养诊断模型. 中国稻米. 2021(03): 21-29 . 百度学术
5. 王浩淼，宋苗语，李翔，扈朝阳，鲁任翔，王翔，马会勤. 无人机高光谱遥感监测葡萄长势与缺株定位. 园艺学报. 2021(08): 1626-1634 . 百度学术
6. 刘雅婷，龚龑，段博，方圣辉，彭漪. 多时相NDVI与丰度综合分析的油菜无人机遥感长势监测. 武汉大学学报(信息科学版). 2020(02): 265-272 . 百度学术
7. 陈晓凯，李粉玲，王玉娜，史博太，侯玉昊，常庆瑞. 无人机高光谱遥感估算冬小麦叶面积指数. 农业工程学报. 2020(22): 40-49 . 百度学术
8. 落莉莉，常庆瑞，武旭梅，杨景，李粉玲，王琦. 夏玉米叶片光合色素含量高光谱估算. 干旱地区农业研究. 2019(04): 178-183 . 百度学术
9. 张良培，刘蓉，杜博. 使用量子优化算法进行高光谱遥感影像处理综述. 武汉大学学报(信息科学版). 2018(12): 1811-1818 . 百度学术
10. 李亚妮，鲁蕾，刘勇. 基于PROSAIL模型的水稻田缨帽三角-叶面积指数模型及其应用. 应用生态学报. 2017(12): 3976-3984 . 百度学术
其他类型引用(17)
计量
- 文章访问数: 1052
- HTML全文浏览量: 294
- PDF下载量: 91
- 被引次数: 27
