 The study aims to develop a robust global model for estimating Sekai disk depth, SDD, using machine learning methods such as Extreme Gradient Boosting, XGBoost, and Random Forest, RF, which showed high precision with mean relative error of approximately 30% and good agreements with long-term in situ SDD in different waters. The results can support long-term global level water quality evaluation and informed development policy decisions. This article was authored by Abour Jong, Kuensha, Xiao Sun, and others.