作　　者： (李琳琳）; (王婷）; (李雨鸿）; (刘青）; (宋晓巍）; (赵振宇）;

机构地区： 辽宁省气象科学研究所,辽宁沈阳110161

出　　处： 《大麦与谷类科学》 2017年第4期50-54,共5页

摘　　要： 利用辽宁省33个气象站点1993—2012年水稻产量、生育期内的旬平均气温、旬降水量及旬日照时数等资料,应用统计分析方法建立水稻的产量动态预报模型。使用5年滑动平均法分离水稻趋势产量,分析气象产量与水稻生育期内逐旬气象要素的相关性,确定5月上旬平均气温、5月下旬平均气温、6月下旬降水量、7月下旬降水量、8月上旬日照时数、9月中旬平均气温和9月下旬日照时数为关键气象因子,建立水稻产量动态预报模型,并对预报结果进行验证。结果表明:对1993—2012年进行模拟预报及回代检验,平均准确率在93%以上;对2013年的产量进行预报,准确率为93.97%~99.67%,预报准确率较高。预测结果基本可以反映水稻产量的变化情况,能够满足业务服务的需要。 In the current research, a dynamic prediction model was established for predicting single-season rice yield by using statistic methods to analyze the data of single-season rice yields, 10-day average air temperature, 10-day precipitation, and 10-day average sunshine hours in the rice growth periods during 1993—2012 in Liaoning Province. By employing the 5-year moving average method,the single-season rice yield trend was isolated. The correlation between meteorological yield and the 10-day meteorological elements in rice growth period was determined, which led to identification of seven key meteorological factors as follows： early-May mean air temperature, late-May mean air temperature, late-June precipitation, late-July precipitation, early-August sunshine hours,mid-September mean air temperature, and late-September sunshine hours. Based on these key meteorological factors, a dynamic prediction model was established for single-season rice yield. Simulation with the model from 1993—2012 showed that average accuracy for yield prediction was more than 93%; the yield prediction with the model for 2013 showed that the accuracy was 93.97%~99.67%. This indicates that the dynamic prediction model can predict single-season rice yield and basically meets the demand for business services.

关 键 词： 水稻 关键气象因子 气象产量 动态预报