Features Of Upper-level Air Temperature From Different Kinds Of Datasets On The Background Of Global Warming

As global surface warming has been an indisputable fact, research of the upper climate change gradually becomes one of the topic of atmospheric research. However, at present, the upper air research is facing uncertain results of long-term trends due to multi-source information, especially the reliability of the reanalysis is constantly being questioned. Therefore, three radiosonde datasets of RATPC, Angell, CMA-OBS (OBS), four reanalyses of the NCEP/NCAR (NCEP1), NCEP/DOE (NCEP2), ERA-40(ERA40), JRA-25(JRA) as well as Satellite-borne Microwave Sounding Unit (MSU) are compared on the characterization of long-term trend in different regions, different levels, different time scales.As far as typical stations all over the world, the correlation of the NCEP and radiosonde data is the best, JRA is the most different from the other datasets in the lower stratospher, and ERA40exists deviation in the upper troposphere at several stations, especially in low latitudes. As far as the the eastern part of China, the evaluation on the applicability of the four frequently used reanalysis datasets including NCEP/NCAR (NCEP1), NCEP/DOE (NCEP2), ERA-40(ERA40) and JRA-25(JRA) are also carried out. The main conclusions are as follows:(1) In different levels, except JRA, the temperature bias in each reanalysis relative to OBS is varying with altitude:in the stratosphere, as a boundary, below the tropopause the deviation is small while above large, and from500hPa to the ground it become continuous smaller and smaller;(2) Biases between most reanalysis datasets and OBS have significant interdecadal variability, for example, early in1990s the negative systematic bias of NCEP1and NCPE2is significant in the upper troposphere, while after2005the bias turns to be positive, but the bias between JRA and OBS is always negative in the stratosphere until1998and then change to be positive since early in the21st century, it is probably due to the ATOVS data replacing TOVS data during the process of assimilation in that time;(3) NCEP1is closer than others to OBS in climatology, NCEP2is strongest and JRA is weakest in the terms of the linear climate trends;(4) Contrary to the lower stratosphere and the upper troposphere, the lower troposphere is of a different changing trend. Moreover, where to shift from warming trend to cooling trend is different between summer and winter, the former is at700hPa and the latter is200hPa.In addition, the analysis of the relationship between the upper-air temperature and surface temperature over the eastern China are carried out in order to reveal the characteristics of response of the upper-air temperature to the global warming. On the basis of former studies, we combine160ground-based stations in the whole of China in the period of1960-2009to analyze the relationship between the upper-air temperature and surface temperature in order to improve the previous work. The results show that:(1) since1980the relationship between the surface air temperature and the annual average temperature of the each level over eastern China is more and more closer. The upper-air cooling is significantly related with the global warming.(2) Whether the warming rate at the mid-to-lower layer of troposphere or the cooling rate at upper layer of troposphere in the past30years, is much more intense than before1980s.(3) The amplitude of seasonal variation in northern China is larger than in southern China.(4) At the lower level (from surface layer to300hPa) of eastern China, the temperature has the different responses to the global warming at different regions and different seasons. For the regions, the boundary is35°N. There is no cooling to the northern of the boundary, vice visa. For the seasons, there is warmer in winter and cooler in summer.