Abstract:

 Based on 467 heatstroke cases, daily meteorological data and 1-4 days’ accumulated meteorological elements before heatstroke day from 2009 to 2012 in Beijing (sum 45 meteorological factors), relationships between heatstroke incidence and meteorological factors were analyzed by a correlation method. Improved forecast models of heatstroke were established by methods of a multiple linear regression and a nonlinear fitting, and the better model was selected to back substitution and verification for number of heatstroke patients. Also, this model was compared with the current used model. The results show?that temperature is a dominant factor causing heatstroke in summer in Beijing. Cumulative effects of temperature, vapor pressure, air pressure and precipitation in two days have a greater impact on heatstroke. In addition, new models have better fitting and forecasting effect for heatstroke, so is the grades of heatstroke. Thus, the models are of indicative to forecast heatstroke grades and prevent heatstroke.

Key words: Heatstroke, Meteorological factor, Correlation analysis, Forecast model