Abstract: Based on the information of intensity of wheat powdery mildew and climate in Jiaozuo during 1980 and 2005,the quantitative relationship between wheat powdery mildew levels and the local climate indicators was identified by means of order statistical and Bayes criterion,and a prediction model was established.Four factors for prediction model were:(1)the average temperature of October last year(T10)≥16.5℃,(2)the temperature of March this year(T3) ≤9.5℃,(3)the sum of rainfall in August,September and October last year(R8+R9+R10)≤180mm,(4)the rainfall of March this year(R3)≥25mm.A severe wheat powdery mildew disaster would take place if 3 or 4 factors were met at the same time,and the disaster would not happen if non or only 1 factor was met.The historical matching rate for the forecast results was over 88% during the 1980-2005.The forecasted results were in line with the factual records of 2006 and 2007.

Key words: Wheat powdery mildew, Wheat powdery mildew, Prediction, Climate index