Ocean dissolved oxygen plays a significant role on biogeochemical cycle, and on ocean health evaluation as well. Also, the ocean deoxygenation is known to all, however, due to the lack of data, the status of global ocean dissolved oxygen is still unclear. World Ocean Atlas 2018 (WOA18), one known ocean dissolved oxygen of the public spatial datasets, which is based on several kinds of survey data, and has a climatology monthly scale. The climatology monthly scale limits to depict the dynamic characteristics of ocean dissolved oxygen, especially on the global change conditions. Array for Real-time Geostrophic Oceanography (Argo) is a major component of the Global Ocean Observing System, and the BGC-Argo has provided more than 25 thousands of oxygen profiles date to Sep. 2022, and will continuously obtain 2 thousands of profiles each year. Although, these oxygen profiles provide a promising source to analyze the dynamic characteristics of ocean dissolved oxygen at global scale, as driven by ocean currents, these profiles irregularly distribute in space, and the amount of data in time is still scarce. Thus, this paper develops a value-added spatiotemporal approach for exploring ocean dissolved oxygen from BGC-Argo profiles, named VAEDO. The VAEDO consists of two components. One component develops a dual-constraint interpolation method with the space-quality of oxygen profiles, which aims at dealing with the irregularly distribution and the data sparsity in space. The other component designs a based-random forest reconstruction model of oxygen profiles with Argo temperature and salinity profiles, which aims at resolving the data sparsity in time. Finally, using the VAEDO, this paper produces the climatology monthly and monthly global ocean dissolved oxygen spatial datasets during the period of Jan.2010 to Dec. 2021. The preliminary results show that at climatology monthly scale, the absolute error range covering -20.0~20.0umol/kg between the ocean dissolved oxygen datasets based on dual-constraint interpolation method and WOA18 dataset is more than 85.0%, and that at monthly scale, the absolute error range covering -10.0~10.0umol/kg between the reconstructed datasets based on random forest model and Argo oxygen profiles at 2022 reaches at 89.1%. The VAEDO and the developed ocean dissolved oxygen datasets may improve the application capabilities of Argo data, also will promote the realization of the ‘Global Ocean Oxygen Decade” proposed in the “The United Nations Decade of Ocean Science for Sustainable Development”.