The paper is very comprehensive and provides the most complete and accurate N2O budget published to date. Estimates for almost all known sources/sinks are included and disaggregated spatially and temporally. One of the main strengths is various bottom up and different top down inversion methods are used and it is encouraging that the estimates are mostly consistent. The information presented is very useful and anticipate that this work will be well cited. Some relatively minor points of clarification/suggestions for improvement:

Lines 131-132 are #s based on BU or TD?

Line 143 and elsewhere manure should be mature

Figure 3, why no anthropogenic source for coastal?

Lines 447 and 1170 were FAOSTAT emission factors for N additions based on the 2006 guidelines or the 2019 refinement?

Line 481 does this mean that 56% of N inputs were assumed to be anthropogenic and consequently 56% of total N2O from this source is anthropogenic?

Line 489 define book keeping approach; is it the same as mass balance?

Line 503 does the 5 Tg N refer to NOx?

Figure 13 state that blue is BU and red TD

Line 795 replace The sections followed with The following sections

Figure 13 KAJ seems low. Perhaps this is related to Figure 15, the green bar for TD shows a net sink; this does not seem correct, please double check.

Figure 14 does ensemble include BU and TD?

Figure 15 I think (blue) should be (red) and (yellow) should be (green)

Figure 16b why is non-ag error bar so large? 16d what are A B C D E and why such large bars for A?

Line 1177 mentions higher tier estimates. In this context, mention that the USA uses a Tier 3 approach for most agricultural soils.

Supplement line 15 says 6 models accounted for manure N but line 69 says 5 models

Line 87 equation is missing

Line 126 use more descriptive text than intense. Perhaps state if microbes or plant roots have 1^st shot at available N

Lines 294 and 303 - 2006 guidelines or 2019 refinement?

Good effort but lacks key uncertainty info. Details in attachment.

This paper is an ambitious and detailed compilation of information across many data sources and different modeling approaches.  As such, it will provide a useful reference for the scientific community, including policy makers.  I have listed my most major comments below.

• The paper does not always clearly distinguish between results that are highly speculative and purely model based (e.g., responses to CO2 and climate change) and those that have a more solid grounding in data (e.g., responses to fertilizer and manure inputs). For example, lines 123-131 are stated as though they are facts.  They should be qualified with, “according to BU estimates,” as is appropriately done for the presentation of model results in the next paragraph starting on line 132. 

It also would be useful to include a graphical depiction of the relative uncertainty of different budget terms in Figure 1, e.g., perhaps with solid arrows for more robustly known fluxes and more faded arrows for speculative fluxes.  (The current arrows have more faded colors that transition to more solid colors but it is not clear what this transition represents.) 

• Line 217 “Reducing N2O emissions is a required net-zero greenhouse gas (GHG) emission and the recovery of stratospheric ozone.” First, this sentence does not make sense grammatically. Second and more importantly, the concept of “net zero” is not appropriate for N2O.  Earth has always been, and always will be, a net natural source of N2O, which is then destroyed photochemically in the stratosphere. “Net zero” is mentioned again on line 298 so line 217 does not seem to be just a typo.  I bring this up because I have seen essays arguing for “N2O neutrality” as a feasible and desirable policy goal, to the point that spraying chemicals on natural landscapes is considered as a way to stop nitrification.  The idea that humans should be actively trying to stop natural N2O emissions seems likely to have potentially bad unintended consequences.  The concept that “net zero” is not logical for N2O (which is not analogous to CO2) should be clearly communicated to policy makers, who may not be familiar with Earth’s natural biogeochemical cycles. 
• There is a confusing switching back and forth between 3 alternative time frames: 1997-2020, 1980-2020 and 2010-2019. Please state clearly somewhere early in the methods why these 3 time frames were chosen and why each is significant.
• The frequent reporting of rates of increase in TgN/yr-2 units is not intuitively meaningful and arguably not mathematically correct. Strictly speaking, it assumes that the source can be fit with a parabolic (t^2) parabolic dependence on time, which is not obviously the case for many regions, based on figure 14.  At minimum please explain how these rates of rates of increase were calculated and why there is so much emphasis on reporting them.

Below is a list of more detailed minor comments.

Line 84. It would be useful to provide an estimate of N2O's relative contribution to enhanced GH forcing (e.g., 6% or whatever the latest value is).

Line 109. It's odd to mention 10% here when the abstract cited "nearly 25%" from the preindustrial, indicating that 15% of the rise was prior to 1980.  This isn't wrong necessarily but it sends a confusing message.

Line 110.  What does “it” refer to?

Line 113.  I suggest to delete "with a substantially lower resolution" because it is confusing.

Lines 123-131.  This paragraph should state explicitly that these results (i.e., trends in different sources) are based on BU approaches, since TD approaches cannot generally distinguish individual sources.  As such, the BU results are in large part based on speculative model-based estimates.  The paragraph states them as though they are known facts.

Lines 132-144. This paragraph seems to overlap considerably with the previous paragraph (lines 117-131).  Is the previous paragraph covering the period 1997-2020 (or 1980-2020?), while the current paragraph covers 2010-2019?   Please distinguish the 2 paragraphs more clearly and consider deleting one of them unless there is a compelling reason to distinguish the last 10 years from the last 23-40 years.

Line 149. "since" is confusing.  Do you mean "by"?

Line 151. Is this a second derivative (Tg N/yr-2)?  I think this will be confusing to many readers and not intuitively meaningful.  I suggest to present as the rate of growth in TgN/yr in the 1980s contrasted with the higher rate of growth in the most recent decade. 

Line 169. What is "manure forest conversion"?

Line 198. 66/270 = 24.4%, which is not “more than 25%”.

Line 203. Please update the NOAA reference.

Lan, X., E.J. Dlugokencky, J.W. Mund, A.M. Crotwell, M.J. Crotwell, E. Moglia, M. Madronich, D. Neff and K.W. Thoning (2022). Atmospheric Nitrous Oxide Dry Air Mole Fractions from the NOAA GML Carbon Cycle Cooperative Global Air Sampling Network, 1997-2021, Version: 2022-11-21, https://doi.org/10.15138/53g1-x417.

Line 207. Please clarify that this means since the period of observations began.  As written, it seems to imply that the growth rate was higher prior to 1980.

Line 217. What does "is a required net-zero GHG emission" mean? 

Line 223. This suggests that up to 44% of N2O is produced abiotically.  This doesn't seem right.  Furthermore the cited 56-70% seems at odds with Figure 1, in which only 2.3/18.1 TgN (12.7%) is from a non-microbial source.

Line 248. Grammar note: missing "a" before small. 

Substance note: the effect on the N2O source from OMZs in the ocean will not necessarily be small, but perhaps this sentence refers to the fact that N2O contributes a relatively small part of GHG warming (even if the expanding OMZ effect on N2O emissions is large).

Figure 3.  Please delete page number 275.

Line 298. Again, it is not possible or even desirable to achieve net zero emissions of N2O.   To do so would involve severe disruptions to Earth's natural nitrogen cycle. This needs to be made clear to policy makers who are not familiar with  biogeochemistry.

Line 355. It is debatable whether models are accurately capturing nitrification, denitrification, and other key processes (e.g., Nevison, C., Goodale, C., P. Hess, W.R. Wieder, J. Vira and P.M. Groffman (2022). Nitrification, and denitrification in the Community Land Model compared to observations at Hubbard Brook Forest. Ecological Applications.  https://doi.org/10.1002/eap.2530.)

1. 15, Table 1. Please specify whether the shelf products represent natural or anthropogenic emissions or both.

p.16, Table 1. Please indicate what type of emissions are modeled with these "other" approaches.

Line 419. That this part of the natural flux was ASSUMED to be constant should be acknowledged in the abstract around line 130, which states that natural sources were relatively constant, as though this were a result.

Line 423.  44% of what?

Line 492. Please comment here on whether any data exist to evaluate these model predictions.  This seems like a highly speculative flux to include in the budget, with a less solid grounding in data and observations than, e.g., direct agricultural emissions from fertilizer or manure.

Line 505. Is this in the stratosphere or the troposphere?

Line 587. Why such a small increase in manure management when the other 3 sources increased by 50% or more during the same period?

Line 626. What is EDGAR/NMIP2?  I do not see it described above in the methods.

Line 700. Perhaps comment on whether the trend in the posterior fluxes differed substantially from the assumed trend in the prior emissions.

Figure 13. Please label TD and BU in the figure legend and/or describe in the caption.

Line 798. Similar to my comment in the executive summary, why are 1997-2020 and 1980-2020 used as alternative historical periods?

Line 898. Was this decrease due mainly to a reduction in fertilizer use from 1980 to 2020?  If so, readers might be interested to know if the reduction was this achieved due to deliberate mitigation strategies or rather to the collapse of the Soviet Union?

Line 998 and elsewhere.  Again, the use of the second derivative is confusing.

Line 1112.  Again, what is manure forest conversion?