基于小波与ARIMA的黄骅市降水和气温的分析研究

Research of Precipitation and Temperature in Huanghua Based on the Wavelet and ARIMA Analysis

为探究黄骅市降水和气温的变化特征及演化趋势,基于1956—2020年黄骅市降水和气温资料,运用Mann-Kendall检验、累积距平、小波分析的方法,研究黄骅市降水和气温的演变特征与周期性规律,并基于1956—2015年的数据建立ARIMA模型,对该市2016—2020年的降水量、温度进行了拟合与对比分析。结果表明:1956—2020年,黄骅市多年平均降水量588.6 mm,多年平均温度12.7℃。1956—2020年黄骅市气温显著上升,降水量呈下降趋势。在55年的时间尺度上,气温和降水的小波方差均存在一个明显的峰值,且二者在该时间尺度上对应的周期均为37年。年均温度与月降水量的最优模型分别为ARIMA(0,1,2)、ARIMA(1,0,0),实测与模拟值相对误差不超过22%。

To investigate the characteristics and evolutionary trends of temperature and precipitation in Huanghua,based on the local temperature and precipitation data from 1956 to 2020,methods including Mann-Kendall test,accumulated anomaly and wavelet analysis were utilized to study the evolutionary characteristics and periodic laws of precipitation and temperature in Huanghua.Based on the meteorological data from 1956 to 2015,the ARIMA model was established to fit and compare the precipitation and temperature during 2016—2020 in the city.The results show that from 1956 to 2020,the multi-year averaged precipitation is 588.6 mm and the multi-year averaged temperature is 12.7℃in Huanghua.The temperature of Huanghua has presented a significantly increasing trend and the precipitation has performed a descending trend from 1956 to 2020.There is a significant peak in the wavelet variance for both temperature and precipitation on a 55-year time scale,which corresponds to a period of 37 years.The optimal models for annual average temperature and monthly precipitation are ARIMA(0,1,2)and ARIMA(1,0,0),respectively,with the relative errors smaller of 22%between the observed and simulated values.