中国沿海赤潮灾害时间序列特征的模拟与预测

Simulation and forecast of the red tide's time series characteristics in China seas

对赤潮灾害时间序列的特征分析是进行赤潮灾害防灾减灾工作的基础,对未来赤潮灾害的预测具有重要意义。通过观察发现中国沿海赤潮灾害年发生频次和分布面积的时间序列(2004—2019年)均呈现出趋势项和周期项并存的特征,因此利用灰色-周期外延组合预测模型对两个序列中的趋势项和周期项进行提取,模型的拟合度分别达到95.20%和95.24%,优于灰色模型、Fourier级数扩展模型和Holt-Winter指数平滑模型等。最后将该模型结果应用到对未来中国赤潮灾害的预测中,得到结论:2020和2021年赤潮灾害发生频次分别为39次和42次,与2019年相比略有上升,2020年赤潮灾害分布面积达到3168km^(2),与2019年相比增加了约59%,而2021年赤潮灾害分布面积出现回落,为1901km^(2)。

Analyzing time series characteristics of red tide is the basis of disaster prevention and mitigation,thus important for red tide predictions.There are trend components and periodic cycle components in annual time series of occurrence area and frequency of red tides during 2004-2019,so Gray-periodic extensional combinatorial model is used to extract these components.The fitting of occurrence frequency and area model can be up to 95.20%and 95.24%,respectively.These results are better than that of Gray model,Fourier series extension model,and Holt-winter exponential smoothing model.Consequently,it is used to forecast the occurrence frequency and area of red tide in 2020 and 2021,suggesting that the red tide events in 2020 and 2021 can be up to 39 and 41,respectively.The annual occurrence area of red tide in 2020 can be up to 3168 km^(2),which is about 59%more than 2019.For 2021,occurrence area is projected to fall to 1901 km^(2).