Retrieval of Cloud Optical Properties From Multiple Infrared Hyperspectral Measurements: A Methodology Based on a Line-by-Line Multiple-Scattering Code

A methodology to retrieve cloud optical properties using high spectral resolution (HSR) infrared (IR) measurements is presented. This new method has the ability to easily adapt to multiple instruments and viewing angles. The retrieval uses a line-by-line multiple-scattering simulation and HSR IR measurements to retrieve spectrally resolved cloud optical depths (ODs). The spectral ODs are compared to a precomputed OD database generated from an ensemble of cloud particle-size distributions and precomputed single-scattering and single-particle optical properties for a variety of ice-crystal habits. Cloud microphysics are retrieved by finding the closest fit to the database. Results are independent of first-guess optical property assumptions on size and habit. The retrieval method has been applied to aircraft, satellite, and uplooking HSR measurements with results evaluated against coincident HSR lidar and radar measurements. Analysis of retrieval errors produced by assumptions and uncertainties in the atmospheric state demonstrates different sensitivities to atmospheric parameters when uplooking or downlooking data are analyzed. For both viewing geometries, the retrieval is most sensitive to the uncertainties in the assumed cloud boundaries. It is also found that nonuniform vertical distribution of cloud OD can result in significant biases in the IR retrieved cloud ODs.