Based on the long-term correlation of data, the relative change trend is used to construct the probability density function and the exceedance probability. The confidence limit of temperature relative change trend belonged to natural variability category is studied and calculated under a certain confidence level in China from 1951 to 2017. Determine whether the relative change trend is caused by unnatural factors (whether the temperature increase is significant), and explore the threshold value of temperature changes caused by unnatural factors in different regions, the corresponding transition time period and the evolution trend. The results show that: (1) 10% of the site temperature in Chinese 160 stations are overestimated when using traditional linear regression methods to calculate trends significance . These sites are mainly located in the northwest, southwest and east coastal areas of China. (2) From the perspective of the spatial distribution of temperature trends in the country, except for the cooling trend in the central and western regions of Xinjiang, the other regions are all warming trends. The relative temperature changes in Shanxi, Inner Mongolia, parts of Ningxia, southwestern Xinjiang, Yangtze River Delta and southwestern Yunnan are relatively large. The unnatural trends in Northeast China, Inner Mongolia and the northern Shanxi Province are large, and the temperature increase is significant. (3) From the spatial evolution of the significant inter-decadal warming areas, the North China and Northeast China regions took the lead in increasing temperature, and then gradually expanded to the south and west. During the period of 1966-2001, most of Chinese regions showed an increase in unnatural factors; for 1971-2006, the northeastern region and the northeastern Inner Mongolia region began to gradually decrease in temperature, while the significant warming area in southwestern China began to increase gradually; the number of significant warming sites was the largest during 1976-2011; from 1981 to 2016, the significant warming sites was mainly concentrated in the Yellow River, the Yangtze River Basin and the two major river basins and the southern part of China. In summary, there are prominent inter-decadal transitions in both time and space in Chinese significant warming areas caused by unnatural factors. This paper maybe provide new perspectives and new ways for the attribution and prediction research of temperature change in China , and strengthen the transformation and linkage of climate change research results to short-term climate prediction.