本文旨在利用因果分析算法PCMCI对成都市空气质量指数(AQI)进行预测。随着城市化的加速,空气污染问题日益严重,因此准确预测AQI对于改善城市空气质量具有重要意义。本文基于成都市AQI预测的背景与现状,介绍了PCMCI算法的基本原理及其在...本文旨在利用因果分析算法PCMCI对成都市空气质量指数(AQI)进行预测。随着城市化的加速,空气污染问题日益严重,因此准确预测AQI对于改善城市空气质量具有重要意义。本文基于成都市AQI预测的背景与现状,介绍了PCMCI算法的基本原理及其在因果分析中的应用。随后,对成都市2019至2023年的AQI数据进行因果分析,识别出与AQI相关的因果变量,并采用长短期记忆网络进行AQI预测。实验结果表明,本文所提方法的预测结果的均方根误差RMSE相较于传统的ARIMA预测模型和单变量LSTM模型预测精度分别提升了22.14%和9.58%,平均绝对百分比误差MAPE相较于传统的ARIMA预测模型和单变量LSTM模型预测精度分别提升了30.98%和11.59%。基于PCMCI提取的因果关系变量能显著提升AQI预测的准确性,能够为成都市的空气质量管理提供有效的决策支持。The aim of this paper is to forecast the air quality index (AQI) in Chengdu by using the causal analysis algorithm PCMCI. With the acceleration of urbanization, the problem of air pollution is becoming more and more serious, so it is of great significance to accurately predict AQI for improving urban air quality. Based on the background and current situation of AQI prediction in Chengdu, this paper introduces the basic principles of PCMCI algorithm and its application in causal analysis. Subsequently, the AQI data of Chengdu from 2019 to 2023 are causally analyzed to identify the causal variables related to AQI, and the long and short-term memory network is used for AQI prediction. The experimental results show that the root mean square error RMSE of the prediction results of the method proposed in this paper improves the prediction accuracy by 22.14% and 9.58% compared with the traditional ARIMA prediction model and univariate LSTM model, respectively, and the mean absolute percentage error MAPE improves the prediction accuracy by 30.98% compared with the traditional ARIMA prediction model and univariate LSTM model by 30.98% and 11.59%. The causal variables extracted based on PCMCI can significantly improve the accuracy of AQI prediction and provide effective decision support for air quality management in Chengdu.展开更多
目的为探究大气污染物与植被生长状况之间的相互影响,方法基于美国国家航空航天局(national aeronautics and space adminidtration,NASA)提供的归一化植被指数(normalized difference vegetation index,NDVI)与空气质量在线监测分析平...目的为探究大气污染物与植被生长状况之间的相互影响,方法基于美国国家航空航天局(national aeronautics and space adminidtration,NASA)提供的归一化植被指数(normalized difference vegetation index,NDVI)与空气质量在线监测分析平台提供的空气质量指数(air quality index,AQI),采用Kriging插值、一元线性回归和相关性分析等方法,对黄河中下游地区的河南省、山东省、山西省、陕西省,海河流域的河北省和重要城市(北京市和天津市)的AQI与NDVI时空分布特征进行解释,并分析其相关性。结果结果表明:(1)2014—2016年,AQI年均值呈显著下降趋势,每年AQI数值冬季最高,春秋相对较低,夏季最低,且呈现明显的区域差异性,出现中间高两侧低的“中心-外围”结构;2017年后,夏季AQI反而高于春秋季的。(2)2014—2020年,NDVI波动上升,上升斜率为0.0041/a。总体上,NDVI上升的面积占研究区总面积的93.76%,且极显著和显著上升趋势的面积分别占16.32%和12.09%。结论黄河中下游地区和海河流域AQI与NDVI时空格局及相关分析结果可以为气候变化对环境的影响研究提供理论基础。展开更多
文摘本文旨在利用因果分析算法PCMCI对成都市空气质量指数(AQI)进行预测。随着城市化的加速,空气污染问题日益严重,因此准确预测AQI对于改善城市空气质量具有重要意义。本文基于成都市AQI预测的背景与现状,介绍了PCMCI算法的基本原理及其在因果分析中的应用。随后,对成都市2019至2023年的AQI数据进行因果分析,识别出与AQI相关的因果变量,并采用长短期记忆网络进行AQI预测。实验结果表明,本文所提方法的预测结果的均方根误差RMSE相较于传统的ARIMA预测模型和单变量LSTM模型预测精度分别提升了22.14%和9.58%,平均绝对百分比误差MAPE相较于传统的ARIMA预测模型和单变量LSTM模型预测精度分别提升了30.98%和11.59%。基于PCMCI提取的因果关系变量能显著提升AQI预测的准确性,能够为成都市的空气质量管理提供有效的决策支持。The aim of this paper is to forecast the air quality index (AQI) in Chengdu by using the causal analysis algorithm PCMCI. With the acceleration of urbanization, the problem of air pollution is becoming more and more serious, so it is of great significance to accurately predict AQI for improving urban air quality. Based on the background and current situation of AQI prediction in Chengdu, this paper introduces the basic principles of PCMCI algorithm and its application in causal analysis. Subsequently, the AQI data of Chengdu from 2019 to 2023 are causally analyzed to identify the causal variables related to AQI, and the long and short-term memory network is used for AQI prediction. The experimental results show that the root mean square error RMSE of the prediction results of the method proposed in this paper improves the prediction accuracy by 22.14% and 9.58% compared with the traditional ARIMA prediction model and univariate LSTM model, respectively, and the mean absolute percentage error MAPE improves the prediction accuracy by 30.98% compared with the traditional ARIMA prediction model and univariate LSTM model by 30.98% and 11.59%. The causal variables extracted based on PCMCI can significantly improve the accuracy of AQI prediction and provide effective decision support for air quality management in Chengdu.
文摘目的为探究大气污染物与植被生长状况之间的相互影响,方法基于美国国家航空航天局(national aeronautics and space adminidtration,NASA)提供的归一化植被指数(normalized difference vegetation index,NDVI)与空气质量在线监测分析平台提供的空气质量指数(air quality index,AQI),采用Kriging插值、一元线性回归和相关性分析等方法,对黄河中下游地区的河南省、山东省、山西省、陕西省,海河流域的河北省和重要城市(北京市和天津市)的AQI与NDVI时空分布特征进行解释,并分析其相关性。结果结果表明:(1)2014—2016年,AQI年均值呈显著下降趋势,每年AQI数值冬季最高,春秋相对较低,夏季最低,且呈现明显的区域差异性,出现中间高两侧低的“中心-外围”结构;2017年后,夏季AQI反而高于春秋季的。(2)2014—2020年,NDVI波动上升,上升斜率为0.0041/a。总体上,NDVI上升的面积占研究区总面积的93.76%,且极显著和显著上升趋势的面积分别占16.32%和12.09%。结论黄河中下游地区和海河流域AQI与NDVI时空格局及相关分析结果可以为气候变化对环境的影响研究提供理论基础。
