Investigating Aerosol-Cloud-Interactions Radiative Impacts combining a New Global Satellite Joint-Dataset and Radiative Transfer Model.

Aerosol-Cloud Interactions (ACIs) remain one of the largest sources of uncertainty in climate projections. Satellite observations provide essential constraints to estimate ACI-induced radiative forcing (e.g., Twomey, 1974; Albrecht, 1989), yet large discrepancies among studies persist due to measurement limitations. Passive sensors such as the Moderate Resolution Imaging Spectroradiometer (MODIS) cannot simultaneously retrieve aerosol and cloud properties, leading to biases when absorbing aerosols above clouds (AACs) alter cloud optical retrievals. These biases become particularly pronounced during extreme aerosol events - such as wildfires, dust outbreaks, or volcanic eruptions - when AACs distort satellite-derived cloud effective radius (CER) and cloud optical thickness (COT). Previous studies over the Southeast Atlantic and Saharan regions have shown that AACs can lead to underestimated COT and either over- or underestimated CER (Haywood et al., 2004; Alfaros and Contreras, 2013; Costantino and Bréon, 2010, 2013).

   To address these issues, we develop a new methodology combining data from MODIS (and VIIRS) with TROPOMI to construct a high-resolution aerosol–cloud joint dataset. This synergy enables separation of distinct aerosol–cloud configurations - (i) aerosol below cloud top (BCT), (ii) aerosol above cloud and attached (ACTa), and (iii) aerosol above cloud top and separated (ACTs) - facilitating a clearer quantification of their respective influences on cloud properties hence, radiative forcing. The dataset provides global coverage from 2019 to the present, and is applied here to three case studies: the 2019/2020 Australian fires, the 2020 California fires, and the recurrent Namibia/Angola fire season (July-October).

    Our results highlight that accounting for aerosol-cloud vertical configuration substantially improves the quantitative evaluation of ACIs, with cloud droplet number concentration (Nd) exhibiting distinct responses across scenarios. Additionally, we use the Successive Order of Scattering (SOS) radiative transfer model (Lenoble et al., 2007) to simulate aerosol-cloud radiative effects, generate lookup tables (LUTs) to correct cloud retrieval biases in MODIS and other passive sensors and generate aerosol index to better understand its dependency on aerosol layer height and cloud cover.