新疆喀什百日咳与气象因素的多元时间序列分析
|
摘要
目的研究气象因素对新疆喀什地区百日咳发病率的影响,建立多元时间序列(ARIMAX)模型对百日咳发病率进行短期预测,为百日咳的预防与控制提供决策依据。方法利用新疆喀什2011年1月—2016年6月的百日咳每月发病数和同期气象数据,绘制对应残差序列的CCF(互相关函数)图,来确定与百日咳发病率有关的气象因素,建立带有气象因素的ARIMAX模型并预测新疆喀什2016年6—12月的百日咳发病数。预测精度通过均方根误差(RMSE)来评估。结果滞后4阶的平均气压、滞后4阶的浮尘日数和滞后8阶的扬沙日数与百日咳病例数呈正相关,且纳入这3个气象因素的ARIMAX模型的AIC(最小信息量准则)值最小(AIC=394.72)。与一元时间序列ARIMA模型相比,拟合的RMSE降低了5.23%,预测的RMSE降低了4.33%,预测精度明显提高。结论本研究所建立的带有气象因素的ARIMAX(0,1,0)(0,1,1)12模型在短期内可以很精准的预测出新疆喀什地区百日咳的新发病数。其中平均气压、浮尘日数、扬沙日数可作为预测新疆喀什百日咳发病的指标,为相关政府部门提供可靠信息。
[Objective]To study the impact of meteorological factors on the incidence of pertussis in Kashi area of Xinjiang,establish a multivariate time series ARIMAX model to predict the incidence of pertussis in short term,and provide decisionmaking basis for the prevention and control of pertussis.[Methods]The monthly number of pertussis cases and the meteorological data of the same period from January 2011 to June 2016 in Kashi of Xinjiang were used to draw corresponding residual sequence of CCF,to determine the meteorological factors related to the incidence of pertussis,establish a ARIMAX model with meteorological factors and predict the incidence of pertussis from June to December in 2016 in Kashi of Xinjiang. The prediction accuracy was evaluated by root mean square error(RMSE).[Results]The average pressure of lag 4,the number of floating dust days of lag 4,and the number of blowing sand days of lag 8 are positively correlated with the number of pertussis cases,and the ARIMAX model incorporating these three meteorological factors has the smallest AIC value(AIC=394.72). Compared with the ARIMA model of one element time series,the RMSE of the fitting is reduced by 5.23%,the predicted RMSE is reduced by4.33%,and the prediction accuracy is obviously improved.[Conclusion]The ARIMAX(0,1,0)(0,1,1)12 with meteorological factors established in this study can accurately predict the number of newly pertussis cases in Kashi area of Xinjiang in the short term. The average air pressure,floating dust days and sand blowing days can be used as indicators to predict the incidence of pertussis,and provide reliable information for relevant government departments.
[Objective]To study the impact of meteorological factors on the incidence of pertussis in Kashi area of Xinjiang,establish a multivariate time series ARIMAX model to predict the incidence of pertussis in short term,and provide decisionmaking basis for the prevention and control of pertussis.[Methods]The monthly number of pertussis cases and the meteorological data of the same period from January 2011 to June 2016 in Kashi of Xinjiang were used to draw corresponding residual sequence of CCF,to determine the meteorological factors related to the incidence of pertussis,establish a ARIMAX model with meteorological factors and predict the incidence of pertussis from June to December in 2016 in Kashi of Xinjiang. The prediction accuracy was evaluated by root mean square error(RMSE).[Results]The average pressure of lag 4,the number of floating dust days of lag 4,and the number of blowing sand days of lag 8 are positively correlated with the number of pertussis cases,and the ARIMAX model incorporating these three meteorological factors has the smallest AIC value(AIC=394.72). Compared with the ARIMA model of one element time series,the RMSE of the fitting is reduced by 5.23%,the predicted RMSE is reduced by4.33%,and the prediction accuracy is obviously improved.[Conclusion]The ARIMAX(0,1,0)(0,1,1)12 with meteorological factors established in this study can accurately predict the number of newly pertussis cases in Kashi area of Xinjiang in the short term. The average air pressure,floating dust days and sand blowing days can be used as indicators to predict the incidence of pertussis,and provide reliable information for relevant government departments.
引文
[1]张凡,孙忠.传染病季节流行现象预测天津市百日咳疫情[J].中国预防医学杂志,2011,12(4):封三,封底.
[2]刘铁诚,宋凯军,张先慧,等.2005—2015年济南百日咳流行状况[J].中华预防医学杂志,2017,51(9):858-860.
[3]欧金玲,姚开虎,杨永弘.中国学者对百日咳的研究历史及现状[J].临床儿科杂志,2007,25(7):607-609.
[4]陈佳,谢娜,吴秀峰,等.基于ARIMA乘积季节模型的新疆喀什百日咳流行趋势分析[J].新疆医科大学学报,2017,40(3):380-384.
[5]汪远征,徐雅静.多元平稳时间序列ARIMAX模型的应用[J].统计与决策,2007(18):132-135.
[6]陈聪聪.基于ARIMA模型和ARIMAX模型的山东省GDP的预测与分析[D].济南:山东大学,2016:1-49.
[7]YAN L,WANG H,ZHANG X,et al.Impact of meteorological factors on the incidence of bacillary dysentery in Beijing,China:A time series analysis(1970-2012)[J].Plos One,2017,12(8):1-13.
[8]骆鹏,张庶民.百日咳免疫控制研究进展[J].中国公共卫生,2010,26(5):651-653.
[9]黄海涛,高志刚.百日咳再现及其应对策略的研究进展[J].医学综述,2015,21(9):1630-1632.
[10]XION D,LONG Y,WANG S,et al.Spatiotemporal distribution of malaria and the association between its epidemic and climate factors in Hainan,China[J].Malaria J,2010,9(1):185-188.
[11]SHUIS Z,FANG H,JIAN JW,et al.Geographical,meteorological and vectorial factors related to malaria re-emergence in Huang-Huai River of central China[J].Malaria J,2010,9(1):1-9.
[12]李丽丽,刘起勇,林华亮,等.北京市房山区手足口病与气象因素的时间序列分析[J].疾病监测,2015,30(6):458-462.
[13]ZHANG WY,GUO WD,FANG LQ,et al.Climate Variability and Hemorrhagic Fever with Renal Syndrome Transmission in Northeastern China[J].Environ Health Persp,2010,118(7):915-920.
[14]张秉玲,王旭霞,张晓曙.甘肃省东南部地区气象因素与流行性乙型脑炎发病的时间序列分析[J].环境卫生学杂志,2014,4(5):434-437.
[15]李令军,王英,李金香,等.2000—2010北京大气重污染研究[J].中国环境科学,2012,32(1):23-30.
[16]杨柳,周国琪,邹纯朴.百日咳发病与气象因素的灰色关联分析[J].疾病监测与控制杂志,2008,2(10):583-584.
[2]刘铁诚,宋凯军,张先慧,等.2005—2015年济南百日咳流行状况[J].中华预防医学杂志,2017,51(9):858-860.
[3]欧金玲,姚开虎,杨永弘.中国学者对百日咳的研究历史及现状[J].临床儿科杂志,2007,25(7):607-609.
[4]陈佳,谢娜,吴秀峰,等.基于ARIMA乘积季节模型的新疆喀什百日咳流行趋势分析[J].新疆医科大学学报,2017,40(3):380-384.
[5]汪远征,徐雅静.多元平稳时间序列ARIMAX模型的应用[J].统计与决策,2007(18):132-135.
[6]陈聪聪.基于ARIMA模型和ARIMAX模型的山东省GDP的预测与分析[D].济南:山东大学,2016:1-49.
[7]YAN L,WANG H,ZHANG X,et al.Impact of meteorological factors on the incidence of bacillary dysentery in Beijing,China:A time series analysis(1970-2012)[J].Plos One,2017,12(8):1-13.
[8]骆鹏,张庶民.百日咳免疫控制研究进展[J].中国公共卫生,2010,26(5):651-653.
[9]黄海涛,高志刚.百日咳再现及其应对策略的研究进展[J].医学综述,2015,21(9):1630-1632.
[10]XION D,LONG Y,WANG S,et al.Spatiotemporal distribution of malaria and the association between its epidemic and climate factors in Hainan,China[J].Malaria J,2010,9(1):185-188.
[11]SHUIS Z,FANG H,JIAN JW,et al.Geographical,meteorological and vectorial factors related to malaria re-emergence in Huang-Huai River of central China[J].Malaria J,2010,9(1):1-9.
[12]李丽丽,刘起勇,林华亮,等.北京市房山区手足口病与气象因素的时间序列分析[J].疾病监测,2015,30(6):458-462.
[13]ZHANG WY,GUO WD,FANG LQ,et al.Climate Variability and Hemorrhagic Fever with Renal Syndrome Transmission in Northeastern China[J].Environ Health Persp,2010,118(7):915-920.
[14]张秉玲,王旭霞,张晓曙.甘肃省东南部地区气象因素与流行性乙型脑炎发病的时间序列分析[J].环境卫生学杂志,2014,4(5):434-437.
[15]李令军,王英,李金香,等.2000—2010北京大气重污染研究[J].中国环境科学,2012,32(1):23-30.
[16]杨柳,周国琪,邹纯朴.百日咳发病与气象因素的灰色关联分析[J].疾病监测与控制杂志,2008,2(10):583-584.