| A prominent issue in climate research is the trend of the upper air temperature. Since the first satellite TIROS-N placed in orbit in 1978, there have been over 30 years-long observations for the upper air, which is a principal supplement to sounding data in climate research area. As for the scarcity work using MSU/AMSU (STAR, UAH, RSS) long time series in the upper air of China, it is essential for analyzing and comparing the trend characteristics in different time and space scale of three different MSU data, which is the prerequisite for the MSU/AMSU long time series application. Therefore, on the basis of the former work, this paper analyzes the trends characteristics in the middle troposphere (850hPa-300hPa) and lower stratosphere (100hPa-50hPa) of China under the background of global warming. By comparing and anglicizing the reasons for different trend pattern, this paper explores the homogeneity of MSU/AMSU time series and the impact for climate using different correction methods, which lay the foundation for self-building long time series of satellite data. In addition, the quality of the traditional radiosonde long time series in China is evaluated by MSU/AMSU. For these purposes, this paper draws conclusions in the following:1. In the lower stratosphere, the air temperature is cooling by -0.572℃/10a,-0.401℃/10a,-0.417℃/10a, respectively for UAH,RSS and STAR, while in the mid-troposphere, the temperature is warming by 0.115℃/10a,0.160℃/10a,0.150℃/10a. In the four climate sub-regions of mid-troposphere, MSU/AMSU series show that the warming speed of northern China is faster than the south, and the temperature changes in the north dominate the whole China, which consists with the results published already, while in the lower stratosphere, the trend of spatial variation is small. Among the three different MSU/AMSU series, RSS and STAR show more consistent in the trend and annual variations, especially in the lower stratosphere.2. By comparing the trend discrepancy of UAH,RSS and STAR, it concluded that different correction methods for homogeneity applied to MSU/AMSU datasets get different results. The most important and hardest correct is the LECT drift and hot target temperature change. In the lower stratosphere, hot target temperature change dominates the inhomogeneity, while in the mid-troposphere, the impact of LECT drift contributes more than other factors, especially in the Qinghai-Tibetan Plateau area.3. MSU/AMSU time series exhibit warmer trend in troposphere after eliminating the stratosphere contribution using measured brightness temperatures of MSU channels 2 and 4, developed by Q. Fu. This method requires high precision for the input radiosonde data, while the inhomogeneity radiosonde of China is a big obstacle to obtain reliable correction coefficients. For application purpose, the method can effectively rectify the stratosphere cooling signal in the raw MSU2 if homogeneity and stable sounding comes up.4. In the mid-troposphere, Radiosonde shows the largest warming speed, which makes a difference to MSU/AMSU. The differences between Radiosonde and MSU/AMSU attribute to the stronger cooling after volcanic eruptions of St. Helens and El Chicho′n in 1980 and 1982, to the stronger warming after ENSO event in 1997/1998, and to the cooling by the change of instruments and correction methods after 2000. However, Radiosonde exhibits the largest cooling trend in the stratosphere, which also dues to the stronger cooling by the change of instruments and correction methods after 2000.
|
PDF Full Text Request