This paper examines the application of artificial neural networks to reconstructing vertical profiles of atmospheric temperature and relative humidity based on data from the microwave radiometers (MWTS and MWHS) installed on the FengYun-3 series satellites. The algorithm is based on a fully connected feedforward neural network. Brightness temperatures in the radiometer channels, simulated using the RTTOV fast radiation model, and the corresponding vertical temperature and relative humidity profiles from the ECMWF ERA5 reanalysis were used as training data. The resulting satellite-derived temperature and relative humidity profiles were compared with ground-based upper-air radiosonde data in Russia's Far East for the summer and winter of 2025. It was shown that the root-mean-square error in surface temperature retrieval does not exceed 3.5 K in summer and 5 K in winter, while in the troposphere and lower stratosphere it does not exceed 3 K. The error for relative humidity, regardless of season, does not exceed 20% in the 1000-100 hPa atmospheric column. An additional error analysis was conducted for coastal stations and mountainous regions, as well as for cases of temperature inversions in winter, which contribute most to the increase in retrieval errors.
artificial neural network method, vertical temperature profiles, microwave radiometers MWTS and MWHS
1. Zhao Y., Wu C., Wu P., Zhu K., Deng X. A microwave radiometer residual inversion neural network based on a deadband conditioning model // Journal of Marine Science and Engineering. 2023. V. 11(10):1887. DOI:https://doi.org/10.3390/jmse11101887
2. Saunders R., Hocking J., Turner E., Rayer P., Rundle D., Brunel P., Vidot J., Roquet P., Matricardi M., Geer A., Bormann N., Lupu C. An update on the RTTOV fast radiative transfer model (currently at version 12) // Geoscientific Model Development. 2018. V. 11. P. 2717–2737. DOI:https://doi.org/10.5194/gmd-11-2717-2018
3. Hersbach H. et al. The ERA5 Global Reanalysis // Quarterly Journal of the Royal Meteorological Society. 2020. V. 146. P. 1999–2049. DOIhttps://doi.org/10.1002/qj.3803
4. Durre I., Yin X., Vose R. S., Applequist S., Arnfield J. Enhancing the data coverage in the Integrated Global Radiosonde Archive // Journal of Atmospheric and Oceanic Technology. 2018, V. 35. No. 9. P. 1753–1770. DOI:https://doi.org/10.1175/JTECH-D-17-0223.1
