I really enjoy reading this well-written and highly interesting manuscript presented by Yi et al. The authors use a Shipping Emission Inventory Model to carefully analyze and generate a powerful set of global emission inventory. The analytical process was thoroughly stepped through in the manuscript, and the data was well-presented. These figures and data would be useful to policy-makers and technology-developers, in addition to scientists in atmospheric and ocean sciences. I’d recommend publication with minor revision.

One small suggestion I have for the authors is that, the definitions of some terms could be better clarified so that the general audience who are not so familiar with the field can understand more easily. For example, I am not 100% sure what AIS signal means – does this mean how many ships there are in this dataset? If so, the authors can simply say this in the captions of Table 1 and Figure 2, and also define this term in the text. Also, please define HC, BC, and other abbreviations that appear in the manuscript.

In addition, figures could have higher resolutions. Right now, it is hard to read the legends. Figure 3i, ‘N2O’ does not have the subscript.

Sijia Dong

This manuscript is a study that established the global ship emissions inventories from 2013, 2016 to 2021 using the SEIMv2.2 model, providing a comprehensive analysis of the patterns of spatiotemporal variations in ship emissions throughout the years. The authors conducted a thorough cleaning of the global AIS data, correcting for spatial drift, excessively long-time intervals, and data misalignment, and employed compression techniques to make the data computable. This extraordinary work is quite challenging, but it ultimately ensures the quality of the data, allowing for a broad analysis of ship emissions from multiple angles, including the composition of ship types, age distribution, temporal changes, spatial variations, and analyses across different intersecting dimensions. The article's visualization of the data results is also very clear and intuitive. The dataset it provides could be useful for future scholars in the fields of atmosphere, ocean, and environment. Overall, this is a good paper that deserves to be published in ESSD. Nevertheless, some minor issues must be clarified.

First, in the analysis of temporal changes, the authors could include some discussion on international policies, particularly how recent fuel-switching regulations impact changes in international shipping emissions, especially within Emission Control Areas (ECA).

Second, could the authors add a paragraph discussing on the uncertainties and limitations of the model in the conclusion section? This could include a discussion on the accuracy of AIS data, the uncertainties in emission factors, and potential future work in these areas.

Finally, in the conclusion section, it would be beneficial if the authors could emphasize the global impacts revealed by the spatial heterogeneity of emissions structure and intensity shown in the high-resolution ship emission inventories. For example, how might this spatial variation affect environmental impacts, or what insights could it provide for emission reduction strategies?