People

PACE has implemented two different cloud remote sensing techniques – for OCI observations we are obtaining bispectral cloud retrievals; and for HARP-2 observations we are obtaining polarimetric cloud retrievals. The implemented Bispectral retrieval algorithm has a very extensive heritage from radiometric imagers such as MODIS/VIIRS, whereas the polarimetric retrieval algorithm itself will be the first spaceborne implementation of this approach from spatially collocated hyperangular observations. Currently, the cloud algorithms implemented for PACE operate entirely independently from one another, with the bispectral algorithm for OCI being largely a re-implementation of the heritage approach (though lacking thermal bands used for cloud top height estimation and thermodynamic phase classification). This project aims to develop synergistic refinements to both the HARP-2 polarimetric (developed by this proposal team under prior PACE SAT funding) and OCI bispectral cloud retrievals that will take advantage of the unique information offered by each approach. With the development of cross-product synergy retrievals, we hope to improve the PACE cloud products and subsequently the enabled science. For example, estimates of the cloud droplet number concentration (CDNC), which are based on the retrieved cloud optical thickness ( and effective radius), can be improved via synergistic constraints on further enhancing studies of aerosol-cloud interactions using PACE observations.

We also aim to implement new techniques that take advantage of the HARP-2 hyperangular reflectances to mitigate sources of uncertainty in various retrieval products. For example, knowledge of the three-dimensional cloud shape can be obtained from hyperangular data and used to retrieve cloud base and top height among other features. Additionally, our group has shown that cloud shape can potentially be used to quantify the impact of 3D radiative effects within a cloud-object by constraining radiative fluxes across the surface boundary of the cloud (section 3.II.c). Currently 3D radiative effects are generally unquantifiable at the pixel level, but a geometric cloud-object approach may offer a path toward addressing this persistent issue.