The paper mainly investigated the temperature difference (ΔT_urban-rural) caused by urbanization in the central-eastern China(including:Beijing-Tianjin-Hebei region(JJJ), Huanghuai region(HH) and Yangtze River Delta(YRD)) from 1981 to 2014, by using the meteorological data of 424 meteorological stations and satellite reanalysis datasets. Results show that the ΔT_urban-rural was basically positive, and there was a significant day-and-night difference in ΔT between the northern areas (JJJ) and the southern areas (YRD):the ΔT_urban-rural of YRD was stronger than the ΔT_urban-rural of JJJ at daytime; on the contrary, the ΔT_urban-rural of JJJ was stronger than the ΔT_urban-rural of YRD at night; the characteristics of the ΔT_urban-rural of HH was between the two above regions. The results indicate that there was a good correlation between ΔT_urban-rural and ΔT_{d urban-rural} (the difference of the surface temperature between urban and rural areas) both at daytime and nighttime. In addition, the scale of the urban was not the cause of the day-and-night difference in ΔT_urban-rural between the North and South. Therefore, this article tried to explain its potential physical mechanism combining with the atmospheric heat sources (sinks) Q₁, water vapor sinks Q₂ and the other elements, and found that the Q₁ may be a better physical indicator to characterize the day-and-night difference in ΔT_urban-rural :(1)At daytime, the Q₁ in south was larger than in north both in rural areas and urban areas. The northern and southern urban underlying surface types were similar (Q₂ is basically the same), while the surface evaporation in the southern rural was intense, which makes the slowly warming up of the surface air temperature, making a larger ΔT_urban-rural in the southern areas than in the northern areas. (2) At nighttime, the absolute value of the negative value of Q₁ in northern rural was larger than that in the southern rural areas and in the northern urban areas. That is, the cooling effect of sensible heat in the northern rural was significantly stronger than that in the southern rural areas and stronger than that in the northern urban areas. Meanwhile, the Q₁ was basically the same both in the northern urban areas and the southern urban areas, making a larger ΔT_urban-rural in the northern areas than in the southern areas. The study showed that the ΔT_urban-rural at daytime was stronger in high temperature and humid areas, while the ΔT_urban-rural at night was stronger in low temperature and dry areas.