Comment on essd-2021-299

Both raised the idea of adding more discussion on how these results will match policy discussions. In response we have added a full paragraph to the Discussion. At the same time ESSD is a venue for papers which describe a model and dataset. This paper is primarily a data descriptor paper, not a discussion paper. Referee 2 pointed out that the national inventory reports are themselves not perfectly certain. This is an interesting consideration. In the current version of OpenGHGMap we limit concern to uncertainty regarding the spatialization of national emissions, assuming that the reported national totals are correct. In future updates it would be interesting to include consideration of that fact that the national inventories are themselves uncertain. We have added a discussion about this to the Uncertainties section of the manuscript.

On the contrary, I am not fully convinced by the use of the results the authors envisage: "Our emissions inventory can support local authorities in their journeys towards climate neutrality in multiple manners. The inventory can help make local and regional sources of emissions more tangible for diverse politicians, city administrations and local communities and provides a good starting point, especially for communities that lack a detailed GHG emissions inventory (lines 698-700)." The experience from the GCoM and other local initiatives shows that local authorities need: 1) a precise estimation of the emissions they can influence by their policies and 2) tools helping them to evaluate the consequences of their actions. The dataset developed does not seem to offer this feature: for instance, attributing emissions equally to fuel stations does not "see" the effect of local traffic control measures. Even very important measures (such as e.g., closing the whole city to the internal combustion cars) would be extremely diluted across all fuel pumps of the nation, making difficult for the mayor to show benefits. Similarly for industries: the map shows several hotspots correspondence of the legal address of industries on which local authorities have presumably little influence.
In the results produced, both emissions manageable and unmanageable by local authorities are inextricably mixed and it is very difficult in some cases understanding how the information could support appropriate policies. In extreme cases, the approach taken could even push inappropriate measures: by absurd for instance, a mayor of a small town could be tempted of zeroing traffic related emissions in its jurisdiction simply closing a fuel station (or moving it to the neighbor town).
I would like to suggest the authors to reflect more on the possible uses of their results, and to present them in a more useful way. For instance, instead of showing results per macro-sectors, a finer or different subdivision could be more appropriate, or other solutions could be possible.
REPLY: These are good comments. We share the referee's vision of a large body of future research to make emissions reporting more granular and actionable. While we fully agree with the referee's comment, it is not clear what specific and feasible changes to make to the current manuscript based on this response. Obviously adding more spatial and temporal detail and linking to policy tools would be a good further step. But realizing that would require a major new modelling effort and paper. The model and paper already represent a substantial advance for the field. We hope to have further improved results to present in the future.
Regarding the request to "reflect on possible uses" -this is primarily a data and model description paper, not a discussion paper. That said, we appreciate the referees request. We have added a paragraph to the Discussion: "We foresee a number of use cases for the results presented here. For one, many local governments in Europe do not have an emissions inventory. The estimated inventory presented here presents a baseline initial estimate. This can be used to reveal which are the priority areas for reduction in each locale. For example, while vehicle electrification is highly promoted, it could be the case that for some regions emissions from light industry, buildings, or industrial sources are multipliers higher and thus represent more important reduction areas. The results presented here are not a full replacement for an inventory prepared using a tool like the GHG Protocol for Cities. A bespoke inventory will be more detailed and the process of preparing the inventory itself usually triggers discussions about solutions. As the body of solutions grows it is possible to imagine cities soon able to construct a Climate Action Plan based on a menu of options. An estimated inventory like the one presented here could be used to prioritize or filter a longer list of solutions into the shorter set most suitable for each city. Finally, the results presented here have some communication value. There is much discussion about decarbonization at the national and EU level, but many are curious about what this should look like at their town, building, or business level. The results presented here can help people translate macro-level concerns into a more tangible vision of what should change in their home town, and how they can participate in that transition.» On summary: the methodology chosen by authors to distribute national emissions down to the finer jurisdiction is surely one of the many possible ones and overall I think it is correct and transparently justified. But I am not sure it is the most appropriate for guiding local authorities in their decision on emissions control.

REPLY:
We agree that methodological innovation in this field is needed. We present just one possible approach to distribute national emissions to the jurisdiction.
As to which method is most appropriate for guiding local authorities in their decision on emissions control, this will be revealed as different local authorities make their choices about conducting an inventory, using this tool, or using other tools.
Specific comment: please consider changing the terminology used e.g., in the title: there are not 108,000 "cities" in Europe. See e.g https://ec.europa.eu/eurostat/web/nuts/tercetterritorial-typologies Here we present a new CO2 emissions inventory for all 116,572 municipal and local government units in Europe, containing 108,000 cities at the smallest scale used. The inventory spatially disaggregates the national reported emissions, using 9 spatialization methods to distribute the 167 line items detailed in the UN's Common Reporting Framework. The novel contribution of this model is that results are provided per administrative jurisdiction at multiple administrative levels, following the region boundaries defined OpenStreetMap, using a new spatialization approach.
Regarding our choice for the title, "cities" is catchier than "local administrative and jurisdictional divisions" and quickly conveys the sense of what we present in the results. In the abstract and Introduction we are quite clear from the outset about what administrative divisions are used so it is not the case that the paper is misleading.

Reply to Referee 2
Interactive comment on "Estimating CO2 Emissions for 108,000 European Cities" by Moran et al.
This study represents valuable modelling work as a first attempt on estimating the CO 2 emissions from cities across Europe. The authors do a great job in identifying and mapping the emission (point) sources and their CO 2 emissions, and the methodology is clear and well explained, a very difficult task given the current state of the proxies' availability and the incomparability between countries administrative units. As an exercise, the authors compare their estimates with data from two global inventories (ODIAC and EDGAR v6.0) and include the Norwegian case-study.
In the current context of the Paris Agreement and reduction pledges, countries should not base their reductions only on national level inventories but take into account cities' individual contributions. In this context, this study is a valuable piece of information to be shared with local authorities. However, the authors should be careful on the sort of concrete messages they want to pass on to the cities. Would be of interest to add a section on how authors see their findings as contributing to local reductions (something like result-mitigation format) and which would, in their vision, be the most critical activities to challenge, in terms of CO2 reduction.

REPLY: We agree this is an interesting topic for exploration, but in this ESSD publication we prefer to keep a tight focus on a clear description of the data and underlying model. The discussion proposed merits a fuller treatment, including probably a closer comparison between the city-level inventory tools like GHG Protocol for Cities, which is very action-Oriented, and top-down models like this an EDGAR.
To further respond to this comment we have added a new paragraph to the Discussion: "We foresee a number of use cases for the results presented here. For one, many local governments in Europe do not have an emissions inventory. The estimated inventory presented here presents a baseline initial estimate. This can be used to reveal which are the priority areas for reduction in each locale. For example, while vehicle electrification is highly promoted, it could be the case that for some regions emissions from residential or commercial buildings, or industrial sources are multiple times higher than from private cars and thus represent more important reduction opportunities. The results presented here are not a full replacement for an inventory prepared using a tool like the GHG Protocol for Cities. A bespoke inventory will be more detailed but the approach presented here can act as a starting point, help with classifying emissions and provide a benchmark against which estimates can be compared or even calibrated. The process of preparing the inventory itself usually triggers discussions about solutions. As the body of solutions grows it is possible to imagine cities soon able to construct a Climate Action Plan based on a menu of options. An estimated inventory like the one presented here could be used to prioritize or filter a longer list of solutions into the shorter set most suitable for each city. Finally, the results presented here have some communication value. There is much discussion about decarbonization at the national and EU level, but many are curious about what this should look like at their town, building, or business level. The results presented here can help people translate macro-level concerns into a more tangible vision of what should change in their home town, and how they can participate in that transition.» As mentioned below in the specific comments, I miss a paragraph on uncertainty calculation. The UNFCCC NGHGIs include in their NIRs (Annexes) uncertainties reported for each sector, sub-sector and activities. Solazzo et al., 2021 estimated as well uncertainties for EDGAR v5.0, year 2015. Even if uncertainty calculation is still not implemented in ESCI, I would strongly advise the authors to include some reference values for uncertainties from the above mentioned sources, as an idea of magnitude.
REPLY: Good point. We have renamed section 7 to "Limitation, Uncertainties, and Future Work" and added the following paragraph: OpenGHGmap treats the CRF National Inventory Reports (NIRs) as authoritative. However, these inventories are contain uncertainties. The NIR reports contain annexes which discuss uncertainties at the sector, sub-sector, and activity levels. The current version of the OpenGHGMap model does not exploit this uncertainty information, but future versions may. Related to this it is noteworthy to mentioned related work on intercomparison of national emissions totals (Elguindi et al., 2020) and an assessment of uncertainty in the bottom-up EDGAR v6.0 model (Solazzo et al., 2021). Since OpenGHGMap treats national inventories as a fixed constraint with no uncertainty, the sources of uncertainty in the model are purely related to the spatialization of emissions. These uncertainties, and modeling choices, are discussed in the relevant section of Methods above.
Overall the paper is well written and has a clear structure. The authors highlighted as well the limitations and future work, in a transparent and honest way, and to which I also contribute with some ideas (see specific comments). I recommend it for publication, subject to addressing the changes and suggestions, as highlighted above and below in the specific comments.

Specific comments:
Title: A bit confusing (108,000 cities) given the number in the abstract.

REPLY:
We've clarified this so the abstract now reads: "Here we present a new CO2 emissions inventory for all 116,572 municipal and local government units in Europe, containing 108,000 cities at the smallest scale used. The inventory spatially disaggregates the national reported emissions, using 9 spatialization methods to distribute the 167 line items detailed in the UN's Common Reporting Framework. The novel contribution of this model is that results are provided per administrative jurisdiction at multiple administrative levels, following the region boundaries defined OpenStreetMap, using a new spatialization approach." Line 23: the authors refer to ESCI as "inventory", I would introduce here the ESCI model name and rename it as "modelled estimates" or "modelled city inventory". REPLY: Good suggestion. We've changed that to read: «Here we provide a new pan-European model estimating emissions at the municipality level»