| Using the random weather generator (WeaGETS) to simulate the daily precipitation sequences in the past 200 years in Laibin, the simulated results compared with the measured data in last 50 years. Application of the simulated sequence analyzed the effect of precipitation to the sugarcane production in Laibin. The results showed that: The relative error of the simulated annual average precipitation was between 0.7% ~13.5%. The simulated monthly rainfall was obvious correlation with the actuality. The daily maximal precipitation and its change trend were trust. The difference between simulated each level rainy days with the real measured ones was 0.2～14.3 days. The simulated effect of daily precipitation probability for below moderate rain was better than the above heavy rain. In short, the simulated precipitation was close to the measured one. So it can be used in the climatic statistical analysis. Under the assurance ratio 80%, the rainfall in each stage of seeding, tillering, jointing and maturation of sugarcane was accounted 51.0%, 52.7%, 72.8% and 38.5% of the stage crop water requirements, respectively. In each growth stage, the no drought probability was 21.0%～40.0%, the light drought probability was 9.5%～29.5%, the moderate drought probability was 10.5%～26.5%, the severe drought probability was 7.5%～22.5% and the super drought probability was 6.0%～29.5%. So in Laibin the drought probability was comparatively large. The most important drought was in Spring and Autumn and major impacted the growth stages of seedling, tillering and maturity. |