The construction of a yield loss evaluation index for the cold vortex type light-temperature-water composite adversity during rice flowering period in Northeast China is important for elucidating the impacts of cold vortex type composite disasters on rice yield loss in middle and high latitude areas. Moreover, it can provide meteorological support to ensure safe production of high-quality japonica rice in China and contribute to regional disaster reduction and efficiency improvement. By combining growth period data, meteorological data, and yield data, we delineated and constructed the composite stress occurrence index of cold vortex type light-temperature-water at the flowering stage of japonica. We analyzed the relationship between factors causing disasters and yield structure, as well as the relationship between different yield structures and yield by employing BP neural network method. We further dissected the processes involved in the causation of combined disasters. Based on the K-means clustering method and historical typical disaster years, we quantified the critical thresholds and disaster grades, and established an evaluation index and model for assessing yield loss caused by combined stress from cold vortex type light-temperature-water. Finally, we examined the spatial and temporal variations of low temperature, abundant rainfall, and reduced sunlight during the flowering period in the three provinces of Northeast China. Results showed that the critical thresholds for light, temperature, and water stress index during the flowering stage of mild, moderate, and severe cold vortex types were [0, 0.21), [0.21, 0.32), and [0.32, 0.64], respectively. The rates of yield loss were [0, 0.03), [0.03, 0.08), and [0.08, 0.096], respectively. Based on the verification results of a total of 751 samples in 11 random years from 1961 to 2020, the percentage of stations for which the production reduction grade, as calculated by the composite index developed in this study, aligning with the actual production reduction grade was 63.7%, consistently exceeding 58.0% annually. Moreover, the proportion of sites with a similarity or difference level of 1 stood at 88.3%, surpassing 85.0% in each year. The index could effectively assess the extent of rice yield loss caused by cold vortex disasters in Northeast China.
构建东北地区水稻开花期冷涡型光-温-水复合逆境产量损失评估等级指标,对阐明中高纬度地区冷涡型复合灾害叠加效应对水稻产量损失的影响机制具有参考意义,可为保障中国优质粳稻安全生产、区域减灾增效提供气象支撑。本研究利用生育期资料、气象资料、产量资料,界定并构建寒稻开花期冷涡型光-温-水复合逆境发生指数,采用BP神经网络法分层次分析致灾因子与产量结构的关系度、不同产量结构与产量的关系度,解析复合致灾过程,基于K-均值聚类方法及历史典型灾害年确定灾害临界值和等级,建立冷涡型光-温-水复合逆境产量损失评估指标及评估模型,并分析东北三省水稻开花期低温多雨寡照的时空分布特征。结果表明: 研究区水稻开花期轻度、中度、重度冷涡型光-温-水复合逆境产量损失评估指标临界阈值分别为[0,0.21)、[0.21,0.32)、[0.32,0.64],产量损失率分别为[0, 0.03)、[0.03, 0.08)、[0.08, 0.096]。基于1961—2020年随机11年总计751条样本的验证结果显示,利用本研究构建复合指数计算的减产等级与实际减产等级一致的站点比例为63.7%,各年均超过58.0%;一致或相差1级的站点比例为88.3%,各年均超过85.0%。该指标能够很好地评估冷涡型复合灾害造成的东北地区水稻产量损失率。.
Keywords: cold vortex composite disaster; composite index; flowering period; rice; yield loss.