| 02 Jul 2026
The JRC-FASST model: an updated version of the global atmospheric source-receptor model based on EMEP MSC-W chemical transport model
Abstract. Air pollution remains a major global challenge, affecting human health, agricultural productivity, and climate. Reduced-form models have been widely used to support rapid assessment of emission control strategies. This study presents an updated version of a global air quality screening tool, JRC-FASST, developed using the EMEP chemical transport model. JRC-FASST is a reduced-form Source-Receptor model, which translates changes in emissions of air pollutants from different source regions into changes in atmospheric concentrations. The new model version improves the representation of air pollution, by increasing spatial detail and refining the definition of emission source regions, leading to a more accurate estimation of pollutant concentrations and population and ecosystem exposure. A key enhancement of JRC-FASST is the expanded and refined set of source regions. JRC-FASST employs 76 source regions globally simulated at 0.5° × 0.5° resolution, including 41 individual countries in Europe simulated at 0.1° × 0.1° resolution. This allows a more detailed representation of national emissions and their transboundary impacts. It incorporates an expanded set of air quality indicators, including particulate matter, nitrogen dioxide, and multiple ozone metrics relevant for both health and crop impacts. JRC-FASST reproduces the results of the full atmospheric model with high accuracy, particularly for particulate matter and ozone, while maintaining computational efficiency suitable for policy screening. Overall, the updated model provides a more robust and flexible tool for exploring air quality policies at national, regional and global scales. Moreover, the current version establishes the methodological foundation for a continuously evolving platform.
This manuscript describes updates to the FASST source-receptor model, used for rapid screening of the impacts of potential emission controls, that include higher resolution, newer emissions, a greater number of source regions and a wider range of impacts metrics. These developments are useful, improving model skill and applicability. However, they are largely incremental, and while the study acknowledges that important methodological advances are needed to address nonlinear responses and changing background methane, these are left for future work. As a consequence the manuscript lacks the innovation and impact of the original TM5-FASST description paper. The manuscript merits publication as it documents useful developments of a model that will be of value to the broader community, but as a model description paper with no substantial new scientific results it would be better suited to GMD than to ACP.
I have two concerns about the paper in its current form that need to be addressed before publication. Firstly, there is no assessment of uncertainty or reliability in the results generated. The reduced form model is shown to match EMEP, but the study lacks further quantitative assessment of confidence in the results that are generated. Formal propagation of uncertainty is not needed here, but some measure of reliability or confidence is needed for anyone wishing to use the results. How dependent are they on the emissions assumptions, the meteorology, or the underlying model used?
Secondly, results sections 3.3-3.5 describing impacts are heavy on description but light on analysis, and their underlying purpose is not clear. The scenarios considered, 20% reductions in NOx and PM emissions, are effectively directly from the EMEP model simulations used to build JRC-FASST, and thus they illustrate EMEP results rather than JRC-FASST results. Focusing on a specific scenario that involves globally varying emissions of all precursors (e.g., CLE in 2050) would allow scientifically useful results to be generated, and provide a more critical test of the JRC-FASST tool by allowing validation against EMEP. This would give these sections a clear purpose and thus strengthen the paper.
General Comments
The weaknesses of the linearized approach are acknowledged in the results section, but little indication is given of the expected biases associated with them or the thresholds beyond which the results should not be believed. This undermines the value of the tool and needs to be addressed.
Section 3.2.3 loosely shows that O3 and PM are matched reasonably well in 2050 under CLE and MFR scenarios, but a much more critical and important test is to examine how well the change in these species is matched. Large errors in these changes may only show up as small errors in final concentrations, and thus the authors assertion in the conclusions (line 743) that the linear approach can "reproduce PM2.5 and O3 concentrations.... within... 5% uncertainty even for considerable reductions of key precursors emissions way above 80%" is vastly over-confident, as is clearly evident from the responses for ozone shown in Figure 7.
The conclusions are greatly overstated: "the new JRC-FASST represents a significant methodological and technical advancement" (line 727). In practice it is a valuable but incremental improvement with no substantial change in methodology, only in underlying emissions and resolution. This is acknowledged later: "This work represents the first step towards the full development of the JRC-FASST model" (line 775) which is a much fairer assessment.
Specific Comments
Line 40: the effect of increased temperature on ozone formation is relatively small; bigger impacts are through its influence on precursor emissions and meteorology.
Line 42: reduced form models haven't yet been introduced in the paper (only in the abstract)
Line 46: Please spell out key acronyms on first use: CTM, DDM, RSM, etc.
Lines 76-79: The objective of the work is stated here, but a clearer purpose for the improvements needs stating. Is this development incremental (just because it is now possible), or is there a specific new application that will only be possible with the improvements?
Line 114: "The model is set up using WRF Single-moment 5-class". This statement is unclear to a general reader and requires explanation and context.
Lines 114-122: this section would benefit from some rewriting, as the technical detail is not all needed for the paper. A "recommendation" (line 121) is not suitable for a description of what has been done.
Line 124: "has... been evaluated well". This is ambiguous and should be rephrased. Do you mean well evaluated (lots of evaluation done, results so-so) or evaluated and shown to perform well (some evaluation done, results are good)?
Line 127: "combined with". Are the simulations independent or nested/coupled? If independent then this phrase should be replaced with "and". The simulations were not combined, the outputs were combined.
Table 1: If PM2.5 is the sum of POM, BC and RemPM, then either switch the PM2.5 and PMcoarse columns so that the corresponding columns are adjacent, or drop the PM2.5 column and label the remaining columns PM2.5(POM), PM2.5(BC), PM2.5(Other).
Line 153: the statement here implies that all reduced-form models are linear functions, but this is not the case. Please rephrase to explain that FASST is based on linear functions. Note that the function does not approximate the processes (line 154), it approximates the effect of the processes on concentrations (this is not the same thing) so please rephrase.
Line 157, Equation 1: if using subscript notation to show dependency, then to ensure consistency the SRC term must have subscripts of R, m, S and p.
Line 165: It would be helpful to identify some of the key metrics here in the text (PODy, SOMO35, AOT40, etc.) so that the definitions do not have to be squeezed into the caption for Table 2.
Line 193: the list of EU countries is not needed here, as the information is already provided in Table A3.
Line 240 and Equation 2: simplify notation by using subscript G rather than GLB. Consider using subscript 0 to replace base. Subscript x is not explained and can probably be removed (otherwise it would be needed on the C terms too).
Equation 2 assumes that contributions from each source region are independent and do not interact with each other. This is an important assumption (which will be violated for anything other than small perturbations) so the assumption should be stated clearly in the text.
Line 272: Threshold-based ozone metrics do not respond linearly to emission perturbations, so the linear summation in Equations 2 and 3 is not fully appropriate. Are these dealt with differently, or are the uncertainties ignored?
Line 288: Health Impacts: how does this approach differ from that used in the original FASST tool? It is important to highlight which aspects are new here.
Line 320: These metrics have already been defined earlier.
Lines 353-364: The first section of the results is poorly written and needs rewording. Fig 4 shows only the EMEP results, so there is no evidence of any "significant improvement" (leave this discussion for later in the section). The statement that biases have a "minor impact" when considering relative changes is a much-repeated fallacy (it assumes biases are systematic and proportional, and they usually are not), and the authors indeed acknowledge this in the next sentence (line 361). The following sentence notes that EMEP is "extensively and regularly evaluated", but it is the results of the evaluation that matter, not the frequency of evaluation. Please rewrite this section.
Line 378: If a quantitative comparison between the baselines is meaningless, why is it included in this form? It would make more sense to use the TM5 version to generate 2015 conditions so that the comparison can be like-for-like. Note here that better resolution of spatial structure at higher resolution is no surprise, but does not demonstrate that the model actually represents observations better. This is the only section of the paper comparing EMEP-FASST and JRC-FASST, and it is thus important to present a fair comparison to demonstrate the improvements made.
Figure 6: please use the same axis ranges for a given species so that the panels can be compared side by side. TM5 panels are labeled Obs 2015 but the text indicates that the observations are for 2001.
Line 427: The tests do not "validate" the approach, they invalidate it, so it would be better to describe this as characterizing the uncertainties associated with the approach.
Line 431: Two different concepts are confused here: linearity in response associated with the size of the perturbation, which reflects chemical saturation and ultimately titration, and additivity, which reflects interactions between the changes in different precursors or regions. Please clarify the discussion here.
Line 434: "reverse response" would be clearer as "decrease". 20% emission changes change the prevailing O3 regimes, but there are subtle balances between reduced titration and reduced production that partly compensate when averaged in time and space, and JRC-FASST does not resolve these.
Figure 7: the source regions are identified, but what receptor regions are shown? (the same issue arises for Figs 8-10).
Line 511: Please emphasize that year 2018 is different from the year used to create the FASST tool (2015), and thus that this is an effective test of FASST.
Figs 11-13: the country labels in the panels are too small to read and distract the reader from the quality of the agreement. Please remove them. Please also consider combining the three figures into a single 3x3 panel figure so that the reader can compare species and scenarios in a single place.
Line 540: Why does JRC-FASST underestimate NO2 on CLE but overestimate it on MFR, given that both scenarios involve strong reductions in NOx emissions? This is worthy of comment.
Figure 14: Why are 2015-2040 changes included in this figure? Year 2040 is not considered anywhere in the manuscript.
Line 568: Please add a statement to explain the purpose of this section. Are these examples? Why were these specific emission perturbations considered?
Figure 15: Please use consistent x-axis labels on the two panels so that they can be compared (ideally 0.2 ppb intervals). The labels need to be larger and clearer. Consider using the same order of regions on the y-axis so that the difference in chemical regimes is clearer in the differing responses to NOx and VOC.
Figures 16-18: As for Fig 15, please use consistent x-axis labels and larger labeling.
Line 780: inclusion of nitrogen and sulfur deposition would be valuable, but this does not impact the representation of ozone as suggested here.
Appendix Figs A1-A3: the distinction of the different panels is unclear and should be introduced in the caption(s), following the text at line 387. The large difference in the x and y-axes ranges distorts these figures, please use more similar ranges so that the 1:1 line lies closer to the diagonal.
Technical Corrections
Line 72: largely improved -> greatly improved
Line 96 "to study to perform" - please clarify
Line 98: "with the first level about 45 m". Add "deep " or "in depth" at the end of this sentence, otherwise this sounds like the middle of the level.
Lines 99, 104: The format of the citation of the EMEP status reports requires adjustment so that the reader is clear that these are citations. Add date in parentheses?
Line 108: consist in -> consist of
Line 112: there is no need to explain the units, otherwise they should be written out in full e.g., 10, 5 or 2 arc minutes, etc.
Line 137: "the" -> "a"
Line 139: remove "the"
Line 151: set -> set of. Note that CTM has already been defined in the introduction.
Line 182: of them -> of which
Line 192: the resolution -> a resolution
Line 203: replace "perform for all the" with "consider all"
Line 205: remove "were included"
Line 223: SRs -> SR relationships?
Line 224: Sentence starting "In Sect 3.2..." is poorly phrased, please rewrite.
Line 228: but did not vary -> and were not altered
Figs 1 and 2: axes and/or frame needed for figures
Table 3: The model resolution entries for Europe in the last three rows should be 0.1 x 0.1 deg? Please consider simplifying the simulation names: "G0" and "E0" would be simpler and more intuitive than "P0_Global" and "P0_Europe".
Line 336: corps -> crops
Line 350: mode -> model
Line 387: ad -> and
Line 496: scatter plots -> responses (these are line graphs not scatter plots)
Line 539: remove "the" before unity.
Line 551: training -> train, test -> testing
Line 700: below -> greater than (and remove minus sign); also add "occur" before only.
Line 714: above -> greater than (and remove minus sign)
Line 916: Reference incomplete - doi/isbn needed.