Abstract: Atmospheric sounding plays a crucial role in climate change research, revealing how the three-dimensional structure of the atmosphere and atmospheric circulation respond to global warming. It has become one of the key cornerstones of global atmospheric reanalysis. Currently, the primary sounding methods include radiosonde, Aircraft Meteorological Data Relay (AMDAR), and satellite-based atmospheric detection using infrared, microwave, and radio occultation. Radiosonde measurements directly obtain atmospheric temperature, pressure, humidity, wind speed, and direction via instruments carried by balloons, providing continuous vertical profile data at approximately 5-10 meters per second. However, radiosonde observations are limited to twice daily, with sparse station density. Commercial aircraft equipped with specific sensors can provide temperature and wind speed data, while some are deployed with humidity sensors, addressing the temporal resolution limitations of traditional sounding and enabling coverage in remote areas such as oceans and deserts. Satellite infrared and microwave atmospheric detectors can invert the vertical structure of atmospheric temperature and humidity, providing global observations. However, the vertical resolution of the data is low, and there are significant errors in the lower atmosphere. Global navigation satellite system radio occultation measurements can provide observations of atmospheric temperature and humidity profiles in the troposphere, with high vertical resolution, self-calibration characteristics, and high long-term stability. Although there have been many successful efforts in eliminating non-climate signals (instrument changes, platform changes, algorithm updates) in observations, the long-term trend of changes given by different data still varies significantly. In the future, it is necessary to improve homogenization methods and using data assimilation methods to merge multi-source observation data and physical models to form optimal estimates.