基于改进自回归理论的非平稳气温变化预测模型研究

Research on Non-Stationary Temperature Change Prediction Model Based on Improved Autoregressive Theory

气温变化在时间尺度上通常表现为非平稳性,这给气温预测模型的研究带来了挑战。为了能有效地对未来的非平稳气温进行预测,基于自回归模型理论提出一种能处理非平稳气温变化的预测模型,选择1951—2020年常德市57662号站点中2月、5月、8月、11月的平均气温数据对该模型的精度进行了检验,预测了该站点未来15年中2月、5月、8月、11月平均气温的变化情况。研究结果表明:改进的自回归模型在拟合精度上优于传统的自回归模型,其拟合系数最高可达到0.96,证明了该模型在处理非平稳气温变化预测问题上的可行性。

Temperature changes typically exhibit non stationarity on a time scale,which poses challenges to the research of temperature prediction models.In order to effectively predict future non-stationary temperatures,a prediction model that can handle non-stationary temperature changes is proposed based on the theory of autoregressive models.The accuracy of the model is tested by selecting the average temperature data of February,May,August,and November from Station 57662 in Changde City from 1951 to 2020,and predicting the changes in average temperature of February,May,August,and November at the station over the next 15 years.The research results show that the improved autoregressive model has better fitting accuracy than traditional autoregressive models,with a maximum fitting coefficient of 0.96,proving the feasibility of the model in dealing with non-stationary temperature change prediction problems.