湖南水稻干旱评估方法研究

A Study on Rice Drought Assessment in Hunan

为了探讨双季早稻、双季晚稻、一季稻干旱影响定量评估方法,统计各熟制水稻各生育期综合气象干旱指数、水分亏缺指数与水稻受旱率、成灾率、单位面积粮食减产量的相关,分析各熟制水稻干旱影响敏感期;基于相关系数,定量计算综合气象干旱指数、水分亏缺指数组合成综合评估指标所占的权重,组建各熟制水稻旱情综合评估指标;基于各熟制水稻旱情综合评估指标,结合抗旱能力,通过多元回归建立各熟制水稻受旱率、成灾率、单位粮食减产量评估模型。结果表明:(1)湖南季节性干旱特征明显,对于双季晚稻,各生育期干旱出现频率基本接近,干旱影响期为返青分蘖期和幼穗形成期;双季早稻、一季稻各生育期干旱出现频率差异较大,以水稻生长后期干旱出现频率高,而水稻干旱影响敏感期也在水稻生长后期。(2)双季早稻干旱灾害评估模型历史同级拟合率平均为77.6%,级差一级以下的拟合率为91.1%;双季晚稻同级拟合率为70.9%,级差一级以下的拟合率为83.0%;一季稻同级拟合率为61.8%,级差一级以下的拟合率为84.5%。基于降水、需水、可供水等因子建立水稻干旱灾害评估模型,物理意义清楚,准确率较高,可为水稻干旱影响定量评估提供技术支持。

In order to study the quantitative assessment of drought impacts on double cropping early rice, double cropping late rice and single cropping rice, the correlation analysis was applied between comprehensive meteorological drought index, water deficit index and drought covered ratio, drought inundated ratio, yield reduction per unit area in each growth stage of each cropping rice, then drought sensitive period of each cropping rice was determined. The weights of comprehensive meteorological drought index and water deficit index which were combined into a comprehensive evaluation index of each cropping rice were calculated, based on the correlation coefficient. Based on the comprehensive evaluation index and combined with the drought-resistant ability, the assessment models of each cropping rice about drought covered ratio, drought inundated ratio and yield reduction per unit area were built by multiple regression equation. The results showed that： （1） seasonal drought was a significant characteristic in Hunan Province. As for double cropping late rice, the frequency of drought had no significant difference among each growth stage; the drought sensitive period was the turning green to tillering stage and the spike differentiation stage. For double cropping early rice and single cropping rice, the frequency of drought had a significant difference between each growth stage, the high frequency appeared in the late stage of rice growth, and that was also the drought sensitive period. （2） The historical fitting rate of drought disaster assessment model for double cropping early rice was 77.6% in the same levels, it increased to 91.1% while the difference was under a rank. For double cropping late rice, the fitting rate was 70.9% in the same level and 83.0% while the difference was under a rank. For single cropping rice, the fitting rate became 61.8% and 84.5%. The rice drought disaster assessment model which was built based on the factors of precipitation, water demand and water supply had a clear physical meaning and a high accuracy. It could be used to provide technical support for quantitative assessment of drought impacts on rice.