Satellite-Retrieved Cloud Base Droplet Number Concentration Improved by In-Cloud Adiabatic Fraction

Abstract

Accurate retrieval of cloud droplet number concentration (N_d) is essential for understanding aerosol-cloud interactions (ACI) and their climatic impacts. Conventional satellite-based N_d retrieval methods typically assume adiabatic clouds (cloud adiabatic fraction (f_ad) = 1), leading to significant underestimations, particularly in environments characterized by low f_ad. Furthermore, conventional methods retrieve N_d near cloud tops, often much lower than near the cloud base. This study applies the recently developed f_ad retrievals to retrieve cloud base N_d accurately. The retrieval is based on the Visible Infrared Imaging Radiometer Suite (VIIRS) onboard polar orbiting satellites. The validation was done against ship-based measurements over the ocean at the Australian Great Barrier Reef. The results indicate that the traditional adiabatic assumption resulted in cloud base N_d underestimations by a factor of 2.73. To resolve this, we implemented two averaged-f_ad methods and one pixel-level f_ad correction. Averaged-f_ad-based corrections improve N_d accuracy but yield higher root mean square errors (RMSE) and lower correlation coefficients (R) due to mean − value reliance. Pixel-level f_ad-based correction minimizes bias by avoiding error amplification from f_ad spatial averaging. When N_d retrieved by pixels with cloud optical depth larger than the 50th value, the correction optimizes the following results: slope ∼1 (0.983 ± 0.088), lowest RMSE (30.365 ± 12.212 cm⁻³), highest R (0.728, P < 0.05), narrowest metrics confidence intervals, and ±30% validated error. This study underscores the importance of pixel-level f_ad correction in satellite-based N_d retrieval, offering improved accuracy for climate modeling and weather forecasting. Future work should expand validation to diverse regions and seasons to further assess the method's generalizability and limitations.