Comparative evaluation of reanalysis-derived temperature profiles and low-level temperature inversion on the Tibetan Plateau

This research assesses four reanalysis datasets−ERA5, MERRA-2, JRA-55, and CRA40−utilizing radiosonde observations from seven stations collected between 2010 and 2021. The analysis differentiates between the mid-to-upper troposphere (above 500 hPa) and the lower troposphere, with a specific focus on near-surface temperature inversions. The results indicate a high degree of consistency in temperature measurements above 500 hPa, with biases decreasing with altitude; notably, CRA40 exhibits the smallest bias at 200 hPa. In contrast, significant discrepancies are observed below 500 hPa: MERRA-2, JRA-55, and ERA5 display pronounced cold biases, while CRA40 presents a warm bias, which can reach up to 5.80 K at 700 hPa. Furthermore, the study reveals a strong correlation between surface pressure and the distribution of temperature inversions, with ERA5 and MERRA-2 demonstrating the highest accuracy in inversion detection, despite seasonal and station-specific variations noted in CRA40 and JRA-55. These findings underscore the value of reanalysis datasets for meteorological monitoring and climate modeling in the Tibetan Plateau region.