《Sounding Transformer Retrieval Method for Atmospheric Temperature and Humidity Profiles Based on Ground-Based Radiometer Simulation Data》Data

It contains 3713 valid paired samples and covers the period from January 1, 2020 to December 31, 2025. The observation station is Beijing Station (station number 54511, altitude about 32 meters).

The dataset mainly includes three types of data: (1) simulated microwave brightness temperature observed in the zenith direction of ground-based microwave radiometers, with a total of 46 channels, covering the K-band (21.46-31.54 GHz, 23 channels) and V-band (50.96-61.04 GHz, 23 channels); (2) True value of atmospheric temperature and humidity profile matched hourly with brightness temperature (from ground to approximately 10 km altitude, including temperature and absolute humidity stratification data); (3) Auxiliary feature data, including ground meteorological elements (2m temperature, relative humidity, weather phenomena, cloud cover, precipitation) as well as precipitation potential (PWV) and cloud liquid water content (CLW) inverted from ERA5 reanalysis data. This dataset is obtained based on the following method: taking the 12 hour sounding data from 54511 stations published by the University of Wyoming as atmospheric input, using the MonorTM v5.4 line by line radiative transfer model, and configuring zenith observation geometry according to the 46 microwave channels mentioned above for radiative transfer calculation, simulating the brightness temperature of ground-based microwave radiometers; At the same time, corresponding auxiliary features are obtained through ground meteorological observations and ERA5 reanalysis data.