江西省早稻高温热害灾损评估模型

Loss estimation model of high temperature damage to early rice in Jiangxi Province

关键词:  早稻  高温热害  灾损评估模型  主成分回归  江西省

Key words:  early rice  high temperature damage  estimation model of yield loss  principal component regression method  Jiangxi Province

摘要:

针对江西省早稻高温热害灾损缺乏定量化评估技术的问题，采用主成分回归方法，对江西省早稻高温热害灾损评估模型进行研究和验证。结果表明:江西省早稻高温热害过程可分为相对高温胁迫和混合高温胁迫2类；在高温热害的影响因子中，热害积温、最高气温、热害日数对水稻热害具有同等重要的作用，单个或多个因子明显偏高亦可加重热害。本研究建立的早稻高温热害灾损评估模型验证结果表明:验证样本实际产量与模拟产量的平均相对误差为2.4%；分期播种试验期间，遭遇高温热害影响3 个播期实际减产率与模型评估灾损率分别相差2.4%、0.7%、3.6%。研究表明模型模拟的结果比较客观，可利用该模型对高温热害年的早稻灾损率进行模拟和预测。

Abstract:

Due to the lack of quantitative assessment techniques for the yield loss of early rice caused by high temperature damage in Jiangxi Province, a loss estimation model of high temperature damage to early rice was developed and verified by using principal component regression method. The results showed that: The process of high temperature damage to early rice in Jiangxi Province was divided into two types, namely, relative high temperature stress and mixed high temperature stress; Among the influence factors of high temperature damage to early rice, the accumulated hot damage temperature, maximum temperature and hot damage days were equally important. When single or multiple factors were obviously high, the hot damage wouldbe aggravated. The verification results of the model established showed the average relative error between actual and simulated yield was 2. 4%. During the stage sowing test, the differences between actual reduction rate and estimation yield loss of three sowing periods under high temperature stress were 2. 4%, 0. 7% and 3. 6%, respectively. In conclusion, the simulation results by the new model constructed in this study were objective, and the model could be used to simulate and predict the yield loss of early rice in the year of high temperature damage.