Exploring the capabilities of synergistic passive and active remote sensing with a new aerosol retrieval testbed
Abstract
The present generation of space-based instrumentation generally permits measurements spanning the majority of the shortwave spectrum, or observations at multiple viewing angles, but rarely both. Moreover, very few of these instruments possess sensitivity to polarization, and those that do lack the accuracy and resolution required to fully utilize this quantity. This limited information content generally confines current operational aerosol retrievals to only a few parameters, like aerosol optical depth and fine-mode fraction. In the next decade, a variety of satellites are expected to launch with sensors that will exceed these current observational limitations, including platforms that will make coincident measurements of polarimetric radiances with atmospheric profiles of backscatter, extinction, and depolarization from LIDAR. Increased information content in these new datasets is expected to drive significant improvements in aerosol remote sensing capabilities but, if this additional information is to be fully utilized, novel retrieval approaches will have to be developed and evaluated. In this work, observationally constrained surface parameterizations and aerosol models are used to simulate realistic top-of-atmosphere polarimetric radiances as well as elastic and inelastic backscatter profiles. These synthetic observations are fed into the Generalized Retrieval of Aerosol and Surface Properties (GRASP) algorithm and the resulting inversion products are compared with the aerosol state parameters used to produce the simulated observations (i.e. the modeled "truth") in order to obtain estimates of retrieval uncertainty for different instrument and retrieval configurations. This new testbed is used to assess inversion capabilities under various observation scenarios, with an emphasis on the retrieval improvements afforded by polarization and the inclusion of LIDAR profiles. Special attention is paid to cases where the aerosol assumptions modeled by the retrieval diverge from the particle properties used in the forward modeling of the simulated scenes. In the context of polarimetric measurements, it is found that errors in the retrieval forward model can significantly impact retrieval performance, especially incorrect assumptions regarding coarse and fine mode particle shape.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2020
- Bibcode:
- 2020AGUFMA208...04E
- Keywords:
-
- 3311 Clouds and aerosols;
- ATMOSPHERIC PROCESSES;
- 3359 Radiative processes;
- ATMOSPHERIC PROCESSES;
- 3360 Remote sensing;
- ATMOSPHERIC PROCESSES