摘要
利用江西省吉安市地面气象观测资料,采用一元线性回归、均方差、M-K非参数检验法、EMD经验模态分解等分析方法,研究吉安市气温变化特征及其与气象因子的相关性。通过对年平均气温年际变化特征的分析,发现吉安市年平均气温呈增温趋势,每10a增温0.17℃;通过M-K检验,得出2001年为吉安市年平均气温突变年;通过年平均气温与其他气象因子相关性分析,得出气温与降水之间存在显著的负相关性关系。而后根据EMD分解第5个IMF分量得出,年平均气温在年际尺度上呈显著增温趋势,而年平均降水呈显著递减趋势。最后以年平均气压、年降水量、年平均相对湿度、年平均日照和年平均风速5个气象因子为自变量,以年平均气温为因变量,建立逐步回归方程,用于对年平均气温进行预报,模型预测效果较好。
Based on the surface meteorological observation data of Ji'an City, Jiangxi province, by means of linear re-gression, mean square deviation, M-K non parametric test method and EMD empirical mode decomposition method,the characteristics of air temperature change and its correlation with meteorological factors in Ji'an city were stud-ied. Through analyzing the characteristics of annual average temperature variation, it was found that the annual aver-age temperature in Ji'an city was increasing at a temperature of 0.17℃ every ten years; Through M-K test, it wasconcluded that 2001 was the year of the mean annual air temperature in Ji'an city; Through the correlation analysisbetween annual average temperature and other meteorological factors, it was concluded that there was a significantnegative correlation between air temperature and precipitation. Then according to fifth IMF components of the EMDdecomposition, the annual average temperature on the internal scales showed a significant warming trend, while theannual average precipitation showed a significant decreasing trend. Finally, with five meteorological factors of the an-nual average air pressure, annual precipitation, annual average relative humidity, average annual sunshine, and theannual average wind speed as the independent variables, the annual average temperature as the dependent variable,the stepwise regression equation was established to forecast the annual average air temperature, model prediction ef-fect is better.
出处
《河南科技》
2016年第21期157-160,共4页
Henan Science and Technology
关键词
气温
相关性分析
M-K检验
EMD经验模态分解
逐步回归
temperature
correlation analysis
MK test
EMD empirical mode decomposition
stepwise regression