Tomographic reconstruction algorithms for retrieving two-dimensional ice cloud microphysical parameters using along-track (sub)millimeter-wave radiometer observations

The submillimeter-wave radiometer operating in the along-track scanning mode continuously collects brightness temperature (TB) data over a two-dimensional (2D) cloud cross-section as the platform moves forward. TB observations from multiple positions and viewing angles show great promise in better constraining the 2D cloud microphysical properties compared to single-angle observations. In this study, we develop two types of tomographic reconstruction algorithms to retrieve 2D ice water content (IWC) profiles using multi-angle TB observations. The one-dimensional (1D) tomographic algorithm performs 1D retrievals beam by beam using each TB observation at a specific position and angle to derive cloud properties along the propagation path. It then integrates the 1D retrieval results to construct 2D cloud distributions. The 2D tomographic algorithm directly constrains the 2D cloud microphysical properties using multi-angle scanning TB observations. Starting with an initial assumption, the algorithm iteratively refines the 2D cloud microphysical quantities by minimizing discrepancies between TB simulations and observations under prior constraints. Both tomographic algorithms are developed based on a hybrid of Bayesian Monte Carlo Integration (MCI) and Optimal Estimation Method (OEM). A simulation experiment is conducted to evaluate the performance of two tomographic reconstruction algorithms. The experiment demonstrates stable convergence of both tomographic methods, with the 2D tomographic algorithm exhibiting superior performance. The experiment results highlight the significant advantage of using multi-angle observations to constrain 2D cloud structure. Compared to nadir-only retrievals, the tomographic technique provides a detailed reconstruction of ice clouds’ inner structure with high spatial resolution. Also, the technique significantly improves retrieval accuracy by correcting systematic biases and reducing the derivation of retrieval errors. Furthermore, the tomography technique effectively increases detection sensitivity for small ice cloud particles.