Advancements in ocean (and atmosphere) observations and modelling have been creating vast amounts of data to quantify the multi-scale variations of ocean dynamics (circulation, sea levels, hydrography, mixing, and sea-ice, etc.). The forcing mechanisms and predictability of these variations are being revealed through data analyses using classical statistical and dynamic analysis methods, and recently the advanced methods of Machine Learning. The analysis results are being applied to address issues in other disciplines of marine science. Further enhancement of these analyses can be “integrated” as key components of the “Digital Twin of the Ocean” (DTO). This talk will present examples of such analysis research that can potentially contribute to DTO for the protection of marine ecosystem and environment. These example are taken from the results of the speaker’s recent collaborative research, including 1) rapid drop of ocean temperature during cold-air outbreaks; 2) particle tracking modelling for assessing radiological risk of a nuclear power station; 3) research related to marine oil spill research and response; 4) marine heatwaves; and 5) ice phenology variations in ocean and lakes.