Abstract: Clouds have long represented a significant source of uncertainty in the estimation of the Earth Radiation Budget (ERB) from space- or airborne imagery observations. In passive cloud remote sensing, three-dimensional (3D) cloud structures observed in nature are often captured as homogeneous (1D) representations. On the one hand, this limits our ability to access process-level understanding of interactions playing out in an inhomogeneous atmosphere. On the other, this simplification leads to systematic biases during: (a) the retrieval of cloud parameters such as optical thickness and droplet size from imagery, referred to as “3D-bias-RS,” and (b) the subsequent estimation of radiative fluxes from those retrievals, referred to as “3D-bias-RE.” These 3D biases are scale-dependent and anticipated to become increasingly significant in the future, as the radiation science teams supporting the new missions, such as NASA’s AOS and ESA’s EarthCARE, aim to deliver radiation products at process scale (~1 km), in contrast to the traditional aggregation scale (~20 km). In this seminar, we will present approaches developed within our group to address these 3D biases. We introduce a novel radiance-based radiative closure method, utilizing a 3D radiative transfer (RT) model, which allows for the direct quantification of 3D-bias-RS (and other biases in satellite retrievals) by leveraging satellite radiance observations as “ground truth.” Building upon this closure approach, we have also developed a physically based correction method capable of effectively mitigating the bias to the first order. Furthermore, we demonstrate how direct radiation measurements from our legacy airborne instruments can be employed to quantify 3D-bias-RE. Recent advancements in the Libera wide-field-of-view camera provide new remote sensing capabilities, enabling multi-angle observations of clouds and improving the 3D characterization of cloud properties. By integrating innovations in satellite instrumentation, remote sensing techniques, and RT modeling, we propose an operationally feasible solution that addresses 3D biases, thereby contributing to a more accurate cloud representation in ERB estimates from space.
a joint institute of 
and 
and