Retrieval of planetary boundary layer structure from a ground-based microwave radiometer based on the one dimensional variational method

Benefiting from their sensitivity to the lower atmosphere, ground-based microwave radiometers are well-established instruments for detecting atmospheric temperature and humidity profiles in the planetary boundary layer. The one-dimensional variational method utilizes the radiative transfer simulation and exhibits high retrieval reliability when applied to ground-based microwave radiometers. For solving the problems of low information content in ground-based microwave radiometers and the low convergence rate of one-dimensional variational method, surface meteorological observations and regularization parameters were introduced to improve the retrieval accuracy in the near-surface layer and the convergence rate. The performances of the methods mentioned above were evaluated using radiosonde observations. The results show that the root-mean-square errors of the temperature and humidity profiles at the surface are reduced by 78.2% and 55.5%, respectively, with the introduction of surface meteorological observations, and the upper atmosphere temperature and humidity profiles also improve slightly. The regularization parameter has little effect on the retrieval time and accuracy of the algorithm, but it can significantly reduce the number of non-converged profiles and improve the quality of the retrieved profiles.