Reconstructing albedo from mean cloud properties

Abstract. Liquid marine clouds exert a substantial control on the Earth-atmosphere energy system through their large global coverage and high reflectivity of shortwave radiation, resulting in overall negative radiative impact. Previous studies showed that the two dominant factors determining their albedo are cloud fraction (CF) and liquid water path (LWP), but this relationship varies in regions of high aerosol loading. In this work, a simplified kernel was built to assess how well the top of atmosphere (TOA) all-sky albedo (α) can be estimated from the given properties of marine liquid clouds: CF, LWP and cloud droplet number concentration (N_d), and to what extent this approach applies globally. The study uses data retrieved from MODIS and CERES instruments for a near-global ocean domain (60º S–60º N) covering the period 2003–2021. The results showed that the albedo is only reconstructed to within 10 % in less than 40 % of cases. Several modifications of investigated method were tested for the improvement in albedo reconstructions. It was found that the number of biases decreases when the maximum solar zenith angle is considered, as well as if the CF–LWP–N_d–α kernel is calculated on a 1º latitude-longitude grid. The findings show that the relationship between the TOA albedo of a scene of clouds and the retrieved mean cloud properties is not universal and while accounting for regional variation is one way to address this, a better understanding of this effect is still needed to reduce uncertainty in aerosol-cloud interactions.