Abstract:

To reveal the variation characteristics in temperature over the Zhongyue mountain region in the troposphere near the bottom of the free atmosphere in recent 60 years, based on homogenizing the monthly mean air temperature data at Dengfeng meteorological station, the simulated models of the monthly mean air temperature at Songshan meteorological station were constructed using the monthly mean air temperature homogeneous data of Dengfeng meteorological station.The missing data from 1990 to 2002 at Songshan meteorological station were interpolated using the monthly models to get the monthly mean air temperature data from 1956 to 2017.The trend of temperature change was analyzed using a linear regression method.The results show that an obvious correction effect is obtained using the homogeneous treatment to deal with the inhomogeneous effect of the monthly mean air temperature caused by the unnatural factors of station migration.After homogenization, the significant rising rate of the annual average air temperature at Dengfeng station from 1969 to 2017 increases from 0.218℃ per decade to 0.310℃ per decade.The increasing rate is 42%.The verification of the model with independent data shows that, on the whole, the linear correlation coefficient and slope between calculated model values and measured values are 0.999 and 0.989, respectively (n=204, p < 0.01).From January to December, the mean correlation coefficient of model verification and test parameters is 0.958, root mean square error is 11.7%, the average absolute deviation is 0.3℃, the average deviation is 0.1℃, the fitting index is 0.973, and simulation efficiency is 0.900.The model has a good simulation effect.From 1956 to 2017, the annual mean air temperature of Songshan station increases significantly, with a rate of 0.223℃ per decade.Among the four seasons, the temperature increasing rate in spring is the highest, with 0.350℃ per decade, followed by winter and autumn.The mean air temperature in summer does not increase obviously.The warming rate in February is the highest, at 0.445℃ per decade.

Key words: Mountain observatory, Homogeneity, Interpolation