LUO Hui
,
LIU Jie
,
GONG Zaiwu
,
SHEN Jin
,
YANG Shengli
,
DU Yanxiu
,
XIAO Bo
,
DENG Xiaoli
,
MA Xiaomei
,
YAO Dongsheng
,
GAO Wuhu
Abstract:
An applied meteorological econometric model was built by using real time data of daily number of visitors 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 considering 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 special 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 visitors 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 Executive 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.