Abstract: Daily precipitation serial data of 200 years is created by weather generator and is used to compared with history data of 40 years from 1961 to 2000.The result shows that the mean precipitation and probability distribution of simulation value is close to actual value,so the simulated precipitation serial data can be used to analysis the risk of rainfall to crop production.The relationship between rainfall and cotton growth shows that the precipitation is usually sufficient to cotton growth during growing season,but it presents usually drought in June,September and October.The precipitation is usually sufficient to cotton growth in July and August,but few years it also presents drought or waterlogging.The raifall of North China does not fit to wheat growth.The precipitation is not usually sufficient to wheat growth during growing season except October and November,and the precipitation is only supply one third of requiring water during growing season and 20% to 30% in key period April and May.The water is usually shortage during growth season in North China.According to areain 2003,the water of wheat field is shortage of more than 30 billion m~3 during growth season in North China.

Key words: Weather generator, Precipitation, Crop production risk, Cotton, Wheat

CLC Number: 

• S16