We present a near-global observation-based explainable machine learning framework to quantify the response of cloud fraction (CLF) of marine low clouds to cloud droplet number concentration (N_d), accounting for the covariations with meteorological factors. This approach provides a novel data-driven method to analyse the CLF adjustment by assessing the CLF sensitivity to N_d and numerous meteorological factors as well as the dependence of the N_d–CLF sensitivity on the meteorological conditions.