To understand and resolve uncertainties in weather and climate predictions, accurate observations must be obtained. Satellite remote sensing provides the best approach to achieve global observations of many atmospheric and surface variables, yet the assumptions required to retrieve these variables yield accuracies that are often outside of the necessary range. By utilizing all electromagnetic properties such as spatial, spectral, angular, and polarization, these accuracy levels may increase. The National Research Council recognizes this need and states that part of the future for remote sensing lie in developing retrievals derived from multi-angle, multi-wavelength, high resolution, and polarized observations. To meet this need the Jet Propulsion Laboratory in conjunction with the University of Arizona have developed the Multiangle SpectroPolarimetric Imager (MSPI).
In order to accurately interpret the data from MSPI, polarized radiative transfer models must be used. Currently only one-dimension version of models are available. When emphasizing the need for accurate observations, using bad assumptions, such as the plane parallel, that introduce large error to the retrievals need to be avoided. Therefore analysis of MSPI data must be performed by relying on three-dimensional polarized radiative transfer models – yet, no such model is accessible to the public.
This seminar introduces the development of the first publically available three-dimensional, polarized radiative transfer model (I3RC-POL). Benchmarking results have found that the I3RC-POL yields results within 1% error compared to unpolarized radiative transfer models. Polarized intensity benchmarks also return results within 3% error. Although initial benchmarking results are promising, some uncertainties must still be resolved. The hope is that once fully benchmarked, the open source framework of the I3RC-POL may provide a tool that will be shared and developed further by the remote sensing community.