Answers are grounded in this paper and cite their sources.
The accurate and efficient retrieval of boundary layer temperature and humidity profiles is crucial for the study of weather forecasting and vertical transport of the pollutant and energy. Aiming at this problem, this paper proposes an integrated inversion algorithm to obtain the boundary layer temperature and humidity profiles from ground-based infrared hyperspectral radiation by using Radial Basis Function Neural Networks (RBFNN). By using this method, the nonlinear atmospheric radiative transfer processes can be perceived from observational data directly, and the temperature and humidity profiles can be simultaneously inversed without iterative solutions. To investigate the performance of RBFNN, a systematic comparison with the eigenvector method and physical retrieval method is conducted, and the performance of joint and non-joint retrieval strategies for temperature and humidity are also analyzed. The results indicate that: (1) RBFNN outperforms traditional methods (e.g., eigenvector method, physical retrieval method) on the whole, with RMSE of 0.8110 K and 1.0070 g/kg for temperature and humidity profiles respectively; (2) RBFNN demonstrates robust performance in the vertical direction, with near-surface temperature deviations controlled within ±1 K and humidity deviations maintained below ± 1 g/kg throughout the entire layer. (3) The combination of temperature and humidity channel in RBFNN enhances retrieval accuracy, with the correlation coefficients (R2) for temperature and humidity inversions reaching 0.9913 and 0.9461, respectively, outperforming the non-joint inversions. It is confirmed that RBFNN can effectively improve the retrieval accuracy and stability of the temperature and humidity profiles, providing an efficient and reliable approach for ground-based remote sensing of the boundary layer.
No takes yet. Share an insight, caveat, or question.
Xiao et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: