Iron (Fe) is one of most the important nutrients for phytoplankton growth in the ocean, making it a crucial element in the regulation of the ocean carbon balance and biogeochemical cycles. Atmospheric deposition of Fe to the ocean has been increased due to human activities, which can significantly alter the marine ecosystem. These necessitate a comprehensive understanding of how the ocean Fe cycling operates and how it will respond to human perturbations. In this work, we identify key mechanisms that control the ocean Fe cycle in various ocean basins and examine the responses of phytoplankton to an increasing Fe deposition using a global ocean biogeochemistry model, constrained by a new high-quality dataset of the ocean Fe distribution. In the first two parts of the work, we refine the Fe parameterization in an ocean biogeochemistry model and evaluate its ability in reproducing recent observations. We show that our new Fe scheme displays a remarkable improvement over the old scheme. Through a suite of model simulations, we reveal the crucial role of Fe release from particles and Fe retention by organic ligands in forming and maintaining the subsurface dissolved Fe (dFe) maxima observed in many ocean transects. The inclusion of spatially varying ligand classes with different binding strengths in the model is important to explain the observed dFe pattern. We also identify the relative roles of different external dFe sources in different ocean basins. While the atmospheric deposition is an important source of dFe in the Atlantic and Indian Oceans, sedimentary and hydrothermal dFe inputs are more important in the Pacific Ocean. In the third part of the work, we apply an unsupervised classification technique to analyze the dFe budget and the dFe distribution field simulated in different ocean Fe models. We suggest that the upper ocean dFe patterns are modulated by interior ocean processes and that without an appropriate representation of these processes, Fe models cannot reproduce observations, even with a correct magnitude of the external fluxes. Our analysis also emphasizes a much more complex picture of the ocean Fe cycling than that of other nutrients such as phosphorus (P) and nitrogen (N). In the last part, we incorporate our improved Fe scheme into an ocean ecosystem model to investigate the response of the Indian Ocean ecosystem to an increasing atmospheric deposition of Fe. We found that while the diatom growth and export carbon flux are enhanced in the south of 40ͦS, they decrease in some regions in the northern Indian Ocean, compensated by increases in the coccolithophores growth and carbonate carbon flux. These changes lead to a decrease in the carbon dioxide uptake over the Indian Ocean.