HERMES_&Delta; v1.0.1: an open-source emission processor for translating official air pollutant and greenhouse gas inventories into model-ready emissions

Guevara, Marc; Tena, Carles; Camps, Paula; Oliveira, Kevin; Gehlen, Johanna; Albarracin, Angie; Castesana, Paula; Collado, Oscar; Herrero, Laura; Legarreta, Oliver; Lombardich, Ivan; Piñero-Megías, Carmen; Rizza, Luca; Slater, Jessica; Viñas, Artur; Macchia, Francesca; Montane, Gilbert; Jorba, Oriol; Pérez García-Pando, Carlos

doi:10.5194/egusphere-2026-4146

Preprints

https://doi.org/10.5194/egusphere-2026-4146

Preprints

23 Jul 2026

| 23 Jul 2026

Status: this preprint is open for discussion and under review for Geoscientific Model Development (GMD).

HERMES_Δ v1.0.1: an open-source emission processor for translating official air pollutant and greenhouse gas inventories into model-ready emissions

Marc Guevara, Carles Tena, Paula Camps, Kevin Oliveira, Johanna Gehlen, Angie Albarracin, Paula Castesana, Oscar Collado, Laura Herrero, Oliver Legarreta, Ivan Lombardich, Carmen Piñero-Megías, Luca Rizza, Jessica Slater, Artur Viñas, Francesca Macchia, Gilbert Montane, Oriol Jorba, and Carlos Pérez García-Pando

Abstract. This paper presents the High-Elective Resolution Modelling Emission System delta version (HERMES_Δ), an open-source, parallel and flexible emission processing tool to bridge the gap between official inventories and model-ready emissions for air quality and greenhouse gas (GHG) modelling applications. HERMES_Δ processes and combines the inventories reported to the Convention on Long-Range Transboundary Air Pollution, the EU National Emission Ceilings Directive and the UN Framework Convention on Climate Change with a large collection of detailed spatio-temporal proxies and speciation profiles to create gridded, hourly and speciated emissions for input into a variety of mesoscale modelling systems, including CHIMERE, CMAQ, MOCAGE, MONARCH and WRF-CHEM. Official emissions at sectoral and industrial facility-level of NO_x, SO_x, CO, NH₃, non-methane volatile organic compounds (NMVOC), PM₁₀, PM_2.5, black carbon (BC), CO₂ and CH₄ are combined, whereafter a detailed horizontal, vertical, temporal and speciation treatment is applied to produce ready-to-use modelling emissions compatible with multiple user-defined grids and chemical mechanisms. As a results, HERMES_Δ provides self-consistent air pollutants and GHG high resolution emissions, thus enhancing the integrated analysis of emissions affecting both air pollution and climate change. The paper exemplifies the use of HERMES_Δ by processing the Spanish official inventories and describing the methods and sources of information considered to construct the required input information, which can be used as a guidance for potential applications in other countries. Illustrations of HERMES_Δ emission outputs and of its computational performance are provided to show its capabilities. The model is designed following a flexible and transparent framework, so that it can be applicable to any European country where the required input data are available. The HERMES_Δ code is distributed through a Zenodo repository (https://doi.org/10.5281/zenodo.21393948, last access: 10/07/2026).

How to cite. Guevara, M., Tena, C., Camps, P., Oliveira, K., Gehlen, J., Albarracin, A., Castesana, P., Collado, O., Herrero, L., Legarreta, O., Lombardich, I., Piñero-Megías, C., Rizza, L., Slater, J., Viñas, A., Macchia, F., Montane, G., Jorba, O., and Pérez García-Pando, C.: HERMES_Δ v1.0.1: an open-source emission processor for translating official air pollutant and greenhouse gas inventories into model-ready emissions, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2026-4146, 2026.

Received: 10 Jul 2026 – Discussion started: 23 Jul 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Preprint (PDF, 2538 KB)

Supplement (1480 KB)

Download & links

Status: open (until 18 Oct 2026)

Post a comment Subscribe to comment alert

RC1:
'Comment on egusphere-2026-4146', Anonymous Referee #1, 30 Sep 2026 reply
General comments
This manuscript describes HERMES_Δ, an open-source, MPI-parallel Python emission processor that converts officially reported inventories (CLRTAP, NECD, UNFCCC, combined with facility-level LPS and IEPR/PRTR data) into gridded, hourly and speciated emissions for several 3D chemistry-transport models. The Spanish application illustrates the workflow, the input data required and the type of outputs obtained.
Starting from sectoral, facility-level official inventories rather than from pre-gridded products (CAMS-REG, CAMS-GLOB-ANT) is a methodological advance over most comparable tools, and the joint, self-consistent treatment of air pollutants and GHG is timely and relevant for the new Ambient Air Quality Directive and for GHG monitoring initiatives. The processing chain itself is valid and clearly described. However, the manuscript currently provides no quantitative evaluation of the resulting emissions against any independent dataset, and no assessment of the uncertainty associated with several consequential methodological choices (plume rise, NEI/LPS/IEPR reconciliation, PRTR gap-filling). The code (Zenodo, GPLv3), a benchmark case and automated tests are provided. The main limitation is that part of the Spanish pre-processing relies on data obtained "upon reasonable request" or by personal communication.
The presentation is very clear with a number of broken cross-references, minor English lapses and notation/reference inconsistencies to fix (see technical corrections).
In the paper there is no evaluation of the emission outputs against independent data; the absence of any uncertainty or sensitivity assessment for methodological choices the authors themselves flag as alternative options; and an undisclosed but potentially significant gap in source coverage (maritime navigation and cruise traffic).
To improve the paper, I suggest considering the following comments:
The spatial correlation with fossil-fuel CO₂ could partly reflect proxies shared across sectors and species. It could be useful to add: a table showing that annual totals per pollutant (and, where relevant, per SNAP sector) are conserved between the input inventory and the gridded/speciated output; and a comparison with the official 0.1°×0.1° EMEP gridded emissions reported by Spain and/or with CAMS-REG (aggregated to a common grid), so to demonstrate the added value of this approach.

It could be useful to add a sensitivity analysis, or a quantified discussion, of the effect of several assumptions, as for example, the fixed plume-rise factor of 2, the fixed NO/NO₂ split of 90/10 outside road transport, the CO₂-ratio gap-filling of PRTR emissions.

The supplement shows that national and international sea traffic (including international bunkers) and domestic/international cruise traffic are "not included in HERMES_Δ". This is nowhere stated in the main text, and shipping emissions can be a major contributor to NOx and SOx along the Spanish coast and in port cities. The lack of these emissions could also alter the results presented in section 3.2 and the correlation among pollutants, and they represent an important input in CTMs. Please state this limitation explicitly in the main text, quantify the share of national NFR totals it represents for the relevant pollutants, and clarify whether this is a permanent design choice or something it is planned to add in future developments.

The manuscript credits HERMESv3 as a direct predecessor, but does not explicitly state what HERMES_Δ adds relative to it. Given that HERMESv3 is a tool still actively used, a paper introducing a new version should clarify, in a short paragraph or table: which specific functionalities are new in Δ; and what practical reasons would lead an existing HERMESv3 user to migrate to HERMES_Δ.

Please, quantify how often each of the three scenarios occurs in the Spanish case (share of source category–region–pollutant combinations and share of emissions), and clarify: the definition of the correction factor, since the notation is inconsistent between Eq. (2) and the text (subscripts a vs. r); how the case NEI < LPS alone is handled (a negative correction factor), and the case IEPR = 0; how LPS and PRTR facilities are matched to avoid double counting.

A plume-rise factor of 2 is applied to all facilities, while a more detailed treatment based on plant-level vertical profiles has been developed by some of the same authors for power plants. Please clarify why the latter is not used at least for SNAP01, or use it as a benchmark. In Sect. 3.1.1, "the altitude of the stack is obtained using Google Earth elevation": please clarify whether this is terrain elevation, and align the wording with Table 2's "Stack_height" description.

Clarify whether OC or organic matter is emitted, and which OM/OC ratio, if any, is used ("expressed in full molecular mass" is unclear). A brief comparison of the NMVOC composition in Fig. 7 with the speciation used in CAMS-REG would help contextualise the results.

The fourth-level SEI split relies on data obtained by personal communication, and the Spanish input database is available only upon request. Please specify the inventory submission year and LPS/PRTR versions used, and release, as far as licences allow, the open parts of the input database (proxies, profiles) alongside the code.

The claim of applicability to "any European country" is plausible given the European coverage of several proxies, but it is not demonstrated beyond the Spanish case. Please, add a short description of the minimum input data needed and of fallback options when only national NFR totals and European-wide datasets are available.

Technical corrections
Main text:
Abstract, L. 20: "As a results" → "As a result".

L. 49: missing space in "national total,sectoral".

L. 98, 112, 115, 302: "Sect. Error! Reference source not found." must be replaced with the correct section number.

L. 102–103: "Sect. 3.3.2 presents the main conclusions" → Sect. 4.

L. 241: "NEDC" → "NECD".

Eq. (1) and surrounding text (L. 125–134): notation inconsistency Eout(x,v,h,e) vs. Eout(x,v,h)e.

Eq. (2) and L. 274: CFp,s,r vs. "administrative division a" in the text and in the equation.

Table 1: descriptions of Source_category_description and Source_category_code appear swapped; "srt" → "str" in the caption; descriptions of P_hour and P_day refer to "weekly" profiles (also Table 2).

L. 560 ("CLC" 2020) vs. Table 3 ("CORINE land cover 2018"): please harmonise the dataset year cited (the supplement consistently uses "CLMS (2020)").

L. 569: "natural have distribution network" → "natural gas distribution network".

L. 763: "which are both reflect" → "which are both reflected".

L. 774: "nigh-time" → "night-time".

L. 822: "raging" → "ranging".

L. 857: "analyses analyzes", there is a repetition;

L. 1005–1006: "propose" → "proposes"; "a fix plume rise factors" → "fixed plume rise factors".

"Marenostrum 5" (L. 852) vs. "MareNostrum5" elsewhere: please harmonise.

Citations vs. reference list: "Seeman et al., 2024" (L. 986) vs. "Seemann" in the reference list; "Yarwood, 2005" (L. 819) vs. "Yarwood et al., 2005"; "Binkowsk" (L. 437) → "Binkowski"; Benavides et al. (2019) and Copernicus Land Monitoring Service (2020) appear to be cited but missing from the reference list.

Supplement:
Table S3 (SNAP07 section, road transport): "See Table S 4 Error! Reference source not found. for details" — same unresolved Word cross-reference issue as in the main text; please fix throughout the supplement as well.

Reply
Citation: https://doi.org/10.5194/egusphere-2026-4146-RC1

Supplement

https://doi.org/10.5194/egusphere-2026-4146-supplement

Data sets

HERMES_Delta Benchmark Dataset Carles Tena, Paula Camps, and Marc Guevara https://doi.org/10.5281/zenodo.21471814

Model code and software

HERMES_Delta v1.0.1 Carles Tena, Johanna Gehlen, Luca Rizza, and Marc Guevara https://doi.org/10.82201/5DVMYR

Viewed

Total article views: 377 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
209	133	35	377	82	43	32

HTML: 209
PDF: 133
XML: 35
Total: 377
Supplement: 82
BibTeX: 43
EndNote: 32

Views and downloads (calculated since 23 Jul 2026)

Month	HTML	PDF	XML	Total
Jul 2026	66	22	12	100
Aug 2026	52	32	3	87
Sep 2026	62	55	14	131
Oct 2026	29	24	6	59

Cumulative views and downloads (calculated since 23 Jul 2026)

Month	HTML	PDF	XML	Total
Jul 2026	66	22	12	100
Aug 2026	52	32	3	87
Sep 2026	62	55	14	131
Oct 2026	29	24	6	59

Viewed (geographical distribution)

Total article views: 338 (including HTML, PDF, and XML) Thereof 338 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 12 Oct 2026

Short summary

We present the High-Elective Resolution Modelling Emission System delta version (HERMES_Δ), an open-source and flexible emission processing tool that allows making official national inventories fit for air quality and greenhouse gas modelling needs. HERMES_Δ converts raw inventories into gridded, vertically discretised, hourly and speciated emission numerical datasets to make them usable for model simulations. The use of HERMES_Δ is exemplified by processing the Spanish official inventories.


Total:	0
HTML:	0
PDF:	0
XML:	0