学术报告通知

报告题目：Forecasting Ozone Levels: Life before CMAQ: The World of the Perfect Prog and MOS (Model Output Statistics)

报告人： Jerry Davis

报告人职务： Atmospheric Modeling Division, U.S. EPA

报告时间：11月6日（周二）下午14:30

报告地点：中意环境节能楼205会议室

主办研究所：环境系大气污染与控制研究所

内容简介：

The purpose of this seminar is to demonstrate that ozone lends itself to prediction using only meteorological covariates. This is in sharp contrast to the last seminar where the most important covariates in the prediction of total nitrate were chemical in nature. In addition the results from this purely statistical model are comparable to those obtained from CMAQ. Techniques for obtaining forecasts for the covariates will also be discussed. 

The objective of the original research project was to develop a statistical model to predict one day in advance both the maximum hourly and 8 h (10am to 5pm) average ozone for Houston (TX).  A smoothing function approach was taken to model development. Ozone data (1983 to 1991) from ten stations in the immediate Houston area were used in the study. The meteorological data came from the Houston International Airport. The models were developed using data from April through October for 1983 to 1987 and 1989 to 1990. Forecasts were developed for 1988 and 1991. The final model contained the following covariates: the east/west and north/south wind components, opaque cloud cover, yesterday’s maximum ozone, today’s maximum temperature and the morning mixing depth.

Individual forecasts were generated for all ten stations in the Houston area using observed meteorology. The root-mean-square prediction error for the 8 h average forecasts ranged from 13.2 to 16.3 ppb (with R2 ranging from 0.66 to 0.73) for the individual stations and from 18.5 to 22.0 ppb (with R2 ranging from 0.61 to 0.68) for the maximum hourly ozone.