西安世园会客流影响及预测的气象计量经济分析  

Econometric Analysis on the Effects of Meteorological Factors on Daily Visitors to Expo 2011 Xi'an and Its Application for Predicting the Number of Visitors

作　　者：罗慧 刘杰[2,3] 巩在武[4] 沈瑾 杨胜利 杜岩岫 肖波 邓小丽 马小梅[6] 姚东升[7] 高武虎 

机构地区：[1]陕西省西安市气象局,西安710016 [2]中国社会科学院研究生院,北京100005 [3]国家气象中心,北京100081 [4]南京信息工程大学经济管理学院,南京210044 [5]西安世园执委会世园投资集团投资策划处,西安710024 [6]中国气象局气象干部培训学院,北京100081 [7]陕西省防雷中心,西安710015

出　　处：《气象》2012年第11期1408-1416,共9页Meteorological Monthly

基　　金：国家自然科学基金(70901043和71171115);教育部人文社科基金(09YJC630130);公益性行业(气象)科研专项(GYHY201106037);陕西创新基金(2011M-62)共同资助

摘　　要：以2011年西安世界园艺博览会(简称世园会)的入园客流数量和逐日气象资料为基础,以最高温度、平均风速和平均相对湿度为定量解释变量,以降水、节假日、每一天和月份为虚拟解释变量,引入随机误差项的AR(2)结构,通过逐步优化的建模思想,构建了2011年西安世园会入园客流的气象计量经济模型。以该模型为核心,以网络技术和数据库技术为基础,构建世园会入园客流数量预测的气象服务业务化系统。2011年8月底至10月22日,气象部门和世园会管理层同时开展了基于网络的业务化应用,运用逐日温度和降水等天气预报信息,结合入园票务政策变化、优惠措施出台、指定日特别活动等信息,联合开展对逐日、未来一段时间和整个运营期的可能入园客流数量的定量预测,为世园执委会部署和控制客流数量、确保世园会安全运营提供重要参考依据。结论表明,气象计量经济学模型能很好地解释气象因素、节假日、每一天以及不同月份变化对入园客流的影响,准确模拟入园客流数量的逐日与逐月变化特征,能较好地预测未来两周的入园客流人数。在8月提前两个月准确预测出西安世园会可能入园客流总人数逾1600万人,与实际人数约1573万很接近。An applied meteorological econometric model was built by using real-time data of daily number of visi- tors and meteorological factors from April 28 to August 15, 2011. The factors in the model included the number of daily visitors, the daily weather data during this period, and the quantitative explanatory variables of the daily high temperature, mean wind speed and mean relative humidity, meanwhile, the virtual explanatory variables were con- sidering the precipitation, holidays, daily variables, and monthly variables. Besides, a random error factor of AR （2） was introduced. With the above, the meteorological econometric model was created. Furthermore, the spe- cial meteorological service system for predicting the number of visitors of the Expo was developed and the operational tests as well as application were carried out. The system for predicting the number of daily vis- itors has been built and applied in predicting the number of visitors with quantitative analysis on different weather conditions, and this helped the decision-makers of the Executive Committee of Expo 2011 Xi＇an control the number of individual visitors in daily and safely operating. The system has been proved to be successful in the short-range and medium-range predictions which simulated the characteristics of the daily and monthly changes of the numbers, and predicted the number of visitors in 15 days accurately, while the 1--7 d prediction was even better than that of 8--15 d. The result offered continuous support to the Exec- utive Committee for decision-making. According to the prediction of October 2011, 10 million visitors were expected till 22 of October, and the total （including the securities, staff） would exceed 16 million, which was very close to the actual total number, roughly 15.73 million.

关 键 词：西安世界园艺博览会 气象服务 气象计量经济模型 气象服务系统 客流预测 

分 类 号：P414[天文地球—大气科学及气象学]