the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Evaluation of high-resolution predictions of fine particulate matter and its composition in an urban area using PMCAMx-v2.0
- 1Department of Chemical Engineering, Carnegie Mellon University, Pittsburgh, PA, 15213
- 2Department of Physics, University of Patras, 26500, Patras, Greece
- 3Department of Civil and Environmental Engineering, Carnegie Mellon University, Pittsburgh, PA, 15213
- 4Department of Engineering and Public Policy, Carnegie Mellon University, Pittsburgh, PA, 15213
- 5Institute of Chemical Engineering Sciences (FORTH/ICE-HT), 26504, Patras, Greece
- 6Department of Chemical Engineering, University of Patras, 26500, Patras, Greece
- 1Department of Chemical Engineering, Carnegie Mellon University, Pittsburgh, PA, 15213
- 2Department of Physics, University of Patras, 26500, Patras, Greece
- 3Department of Civil and Environmental Engineering, Carnegie Mellon University, Pittsburgh, PA, 15213
- 4Department of Engineering and Public Policy, Carnegie Mellon University, Pittsburgh, PA, 15213
- 5Institute of Chemical Engineering Sciences (FORTH/ICE-HT), 26504, Patras, Greece
- 6Department of Chemical Engineering, University of Patras, 26500, Patras, Greece
Abstract. Accurately predicting urban PM2.5 concentrations and composition has proved challenging in the past, partially due to the resolution limitations of computationally intensive chemical transport models (CTMs). Increasing the resolution of PM2.5 predictions is highly desired to better inform air quality and emissions controls policies that protect public health, and also to address issues related to environmental justice. A nested grid approach using the CTM PMCAMx-v2.0 was used to predict PM2.5 at 36 × 36 km, 12 × 12 km, 4 × 4 km, and 1 × 1 km resolution for a domain largely consisting of Allegheny County and the city of Pittsburgh in southwestern Pennsylvania, US during February and July 2017. Performance of the model in reproducing PM2.5 concentrations and composition was evaluated at the finest scale using measurements from regulatory sites as well as a network of low-cost monitors. Total PM2.5 mass is reproduced well by the model during the winter period with low fractional error (0.3) and fractional bias (+0.05) when compared to regulatory measurements. Comparison with speciated measurements during this period identified small underpredictions of PM2.5 sulfate, elemental carbon (EC), and organic aerosol (OA) offset by a larger overprediction of PM2.5 nitrate (bias = +1.4 µg m-3, fractional bias = +0.81). In the summer period, total PM2.5 mass is underpredicted with fractional bias of -0.39. Here, PM2.5 nitrate is overpredicted again with a large fractional bias (+0.7) but significantly lower magnitude (+0.4 µg m-3). Underpredictions in PM2.5 sulfate and EC contribute to the negative prediction bias of total PM2.5 (-0.4 µg m-3 and -0.2 µg m-3, respectively), however the largest underprediction is seen for summer OA (bias = -1.9 µg m-3, fractional bias = -0.41). In the winter period, the model performs well reproducing the variability between urban measurements and rural measurements of local pollutants such as EC and OA. This effect is also captured well in the summer for EC, although the OA performance here is less consistent because much more of this OA is secondary and transported from outside of the inner modeling domain. Comparison with total PM2.5 concentration measurements from low-cost sensors yielded similar results with slightly higher overpredictions seen in the winter (fractional bias = +0.24) and lower underpredictions seen in the summer (fractional bias = -0.27). Inconsistencies in PM2.5 nitrate predictions in both periods are believed to be due to errors in partitioning between PM2.5 and PM10 modes and motivate improvements to the treatment of dust particles within the model. The underprediction of summer OA would likely be improved by updates to biogenic SOA chemistry within the model, which would result in an increase of long-range transport SOA seen in the inner modeling domain. These improvements are obvious topics for future work towards model improvement. Comparison with regulatory monitors showed that increasing resolution from 36 × 36 km to 1 × 1 km improved both fractional error and fractional bias by 0.04 in February 2017. In July 2017, fractional error decreased by 0.05 and fractional bias improved by 0.07 with increasing resolution. Improvements at all types of measurement locations indicated an improved ability of the model to reproduce urban-rural PM2.5 gradients at higher resolutions.
- Preprint
(1985 KB) -
Supplement
(1813 KB) - BibTeX
- EndNote
Brian T. Dinkelacker et al.
Status: closed
-
RC1: 'Comment on gmd-2022-145', Anonymous Referee #1, 29 Aug 2022
This manuscript presents an application of the CTM PMCAMx-v2.0 at increasing resolutions of 36 x 36, 12 x 12, 4 x 4, and 1 x 1 km for a domain covering Allegheny County and the city of Pittsburgh in southwestern Pennsylvania, USA to predict PM2.5 and its components. In general, the paper is very well structured and written, supported by relevant past works. The results are adequately presented, and extensively analysed covering the assessment of model performance for PM2.5 and its components, the influence of different horizontal resolutions and of the improvement of emissions allocation for cooking and traffic.
However, I have two main concerns regarding the presented work:
- The analysis presented and discussed is based on a modelling run for only two months – February and July – and the discussion and conclusions are drawn for Winter and Summer periods. An example is the sentence in line 327 “Total PM2.5 mass concentrations are underpredicted in the summer period”. It is clear that those months were chosen due to the availability of measurements, particularly PM2.5 compounds. Notwithstanding, the study would be much more robust and conclusion better supported if the modelling period was at least one year. Results could be assessed with the observations from fixed monitoring sites for the whole year, with a focus on those two months for which more data is available.
- The authors refer that “Garcia Rivera et al. (2022) investigated the effects of increasing grid resolution of model inputs and CTM output on source resolved predictions of PM2.5 concentration and population exposure at 36 km, 12 km, 4 km, and 1 km”. Is the current work based on the simulations performed by Garcia Rivera et al. (2022) or are there additional runs in the present work? Are the runs referred in lines 309 – 310 examples of additional simulations? What exactly is the novelty of this study compared to (Garcia Rivera et al.,2022)? This should be made clear in the manuscript and perhaps the “Model Application” section could be shortened since it is already explained in the published paper.
Specific comments
- The abstract is too long and has many details of the analysis of results that are unnecessary in an abstract. Conversely, the Conclusions lack a more comprehensive and focused text on the added value and limitations of this work and concrete ideas for future work.
- Section 3 has a subsection 3.1, that would only make sense if a 3.2 exists. Please restructure this section or remove the subsection title. One idea is to join sections 2 and 3 into one section.
- Figure 1 – Please include the metric scale, and draw the inner domain to allow to have an idea of where stations are located in relation to the modelling domains.
Technical corrections
- Line 323 - The reference (Kodros et al., 2021) is missing in the reference list.
- References Day et al., 2015, Skyllakou et al., 2021, included in the reference list are not cited in the text.
-
AC1: 'Response to the Comments of Referee 1', Spyros Pandis, 12 Oct 2022
Major comments
(1) The analysis presented and discussed is based on a modelling run for only two months – February and July – and the discussion and conclusions are drawn for Winter and Summer periods. An example is the sentence in line 327 “Total PM2.5 mass concentrations are underpredicted in the summer period”. It is clear that those months were chosen due to the availability of measurements, particularly PM2.5 compounds. Notwithstanding, the study would be much more robust and conclusion better supported if the modelling period was at least one year. Results could be assessed with the observations from fixed monitoring sites for the whole year, with a focus on those two months for which more data is available.
We do agree that longer evaluations are always better. However, significant resources are required to develop the emission inventory used for these PMCAMx simulations, so two months in two different seasons were chosen as a compromise. These simulations can provide sufficient information, so that the ability of the model to capture PM2.5 differences due to the seasonal variability in major emissions sources and meteorology can be adequately assessed. The primary goal of this study is to evaluate the predictions of PM2.5 concentrations at high spatial resolution (1 km x 1 km), produced using a corresponding high spatial resolution source-resolved emissions inventory. The high-resolution, source-resolved, and species resolved PM2.5 predictions evaluated in this study are novel and require significant evaluation against available measurements which can be sparse (especially for PM2.5 component analysis). Another factor in our decision has been the connection of the present work with the corresponding analysis and conclusions of Garcia Rivera et al. (2022) study who also studied the corresponding periods. A brief explanation and discussion of this point has been added in the revised paper.
(2) The authors refer that “Garcia Rivera et al. (2022) investigated the effects of increasing grid resolution of model inputs and CTM output on source resolved predictions of PM2.5 concentration and population exposure at 36 km, 12 km, 4 km, and 1 km”. Is the current work based on the simulations performed by Garcia Rivera et al. (2022) or are there additional runs in the present work? Are the runs referred in lines 309 – 310 examples of additional simulations? What exactly is the novelty of this study compared to (Garcia Rivera et al.,2022)? This should be made clear in the manuscript and perhaps the “Model Application” section could be shortened since it is already explained in the published paper.
This study uses the results of some of the simulations of Garcia Rivera et al. (2022) and adds several additional simulations. For example, additional simulations were performed to quantify the impact of the proposed emissions surrogates for traffic and cooking. The work of Garcia Rivera et al. (2022) is an analysis of the impact of increasing grid resolution on predicted human exposure to PM2.5 and the corresponding links with the PM2.5 sources. Garcia Rivera et al. (2022) did not address model performance and the corresponding challenges related to the different types of the available measurements. This evaluation is the primary purpose of the present study. The Model Application section now contains only the most essential information about the simulations so that the interested reader does not need to consult Garcia Rivera et al. (2022) to understand the major features of our simulations. These points have been clarified in the revised manuscript text.
Specific comments
(3) The abstract is too long and has many details of the analysis of results that are unnecessary in an abstract. Conversely, the Conclusions lack a more comprehensive and focused text on the added value and limitations of this work and concrete ideas for future work.
The content of the abstract and conclusion have been reassessed and changes have been made in the manuscript. Additional clarity on the scope of the work has also been highlighted in the revised manuscript.
(4) Section 3 has a subsection 3.1, that would only make sense if a 3.2 exists. Please restructure this section or remove the subsection title. One idea is to join sections 2 and 3 into one section.
This has been fixed in the manuscript.
(5) Figure 1 – Please include the metric scale and draw the inner domain to allow to have an idea of where stations are located in relation to the modelling domains.
Figure 1 (a) shows fully the entire inner domain. Figure 1 (b) shows the City of Pittsburgh limits, which are also shown in Figure 1 (a) to provide the reader with a scale reference. These notes have been included in the revised caption for clarity.
Technical Corrections
(6) Line 323 - The reference (Kodros et al., 2021) is missing in the reference list.
The missing reference has been added.
(7) References Day et al., 2015, Skyllakou et al., 2021, included in the reference list are not cited in the text.
These unused references have been removed.
-
RC2: 'Comment on gmd-2022-145', Anonymous Referee #2, 01 Sep 2022
Dinkelacker et al 2022 investigates the benefit of high-resolution modeling and emissions surrogates for the prediction of PM2.5 and its composition using PMCAMx. The authors present model results and measurement comparisons for model simulation at 36 km, 12 km, 4 km and 1 km simulations, to southwestern Pennsylvania while applying novel surrogates for on-road and commercial cooking emissions. Previously in the work of Rivera et al 2022 evaluated the effects of grid resolution for model predictions of particulate matter. This study enhances that work performing a model evaluation by comparing the modeling approach against EPA and low-cost sensor measurements. The study compares PM2.5 in Winter and Summer and for EPA monitor sites shows that fractional error and fractional bias improved by increasing resolution from 36 km to 1 km. In general, PM2.5 prediction performance improved in almost all cases. The underpredictions of PM2.5 at urban sites and overpredictions at rural sites were reduce as a result of increasing resolution.
Overall this paper is well written and logically organized. The subject matter is one that should appeal to atmospheric chemistry and air quality modelers. This study is valuable in showing the importance in developing improved emission inputs for models in particular for the spatial distribution to emissions from an inventory to a model grid. While a focus on improving model resolution to improve model results has its clear benefits, this study shows clearly that when we increase spatial resolution, it should be accompanied by improved and more accurate emission inputs.Comments:
- While it is clear that the different spatial distribution has an impact on the distribution of emissions for commercial cooking and on-road traffic, how are these two sectors typically distributed in the default surrogates? Please consider elaborating in Section 3 to provide some context to the reader.
- In the PMCAMX model simulations, are there removal processes such as dry deposition and wet scavenging? If so, how were these handled between the different grid resolutions? Are these something that must be interpolated from WRF or is it something that PMCAMx explicitly captures?
- The WRF simulations were conducted at 12 km resolution with the information interpolated to higher resolutions of 1 km and 4 km. First, what was the procedure to create 36 km inputs for PMCAMx? Second, why was this design choice made rather than simulating WRF at 36 km and interpolating for all higher resolution grids? Or perhaps simulate meteorology at 4 km with interpolating to 1 km grid and extrapolating to 12 and 36 km grids?
- For those unfamiliar with emissions processing and emission surrogates, it may prove useful to further emphasis how using different surrogates merely change the spatial distribution of the total amount. The authors perhaps made this clear in Section 6 and the Supplementary material but the reader could benefit from this stated more clearly and earlier.
- Related to the previous point, at times in the paper it reads like the novel surrogates for commercial cooking and on-road emissions were applied to all grid resolution configurations and other times it reads like they were only applied to 1 and 4 km simulations (see Line 183-184).
- There is no mention of the novel emission surrogates for commercial cooking and on-road emissions in the abstract. I feel like this point could use some emphasis in terms of mentioning that these were used and developed and briefly mention the evaluation of these novel surrogates.
- The paper could in general benefit from making it more clear how this study is both related and also different from Rivera et al 2022.
Minor comments:
- Table 4 could benefit from explicitly mentioning in the caption that it is for the $1 \times 1$ km resolution.
- In supplementary material, the captions for Figure S7 and S8 should be checked and corrected to avoid any confusion. Figures S7 and S8 captions incorrectly start off with mentioned commercial cooking and also both for February 2017. Presumably, S7 and S8 are for on-road traffic for February and July respectively as they mention using simulated traffic approach.
-
AC2: 'Response to the Comments of Referee 2', Spyros Pandis, 12 Oct 2022
Comments
(1) While it is clear that the different spatial distribution has an impact on the distribution of emissions for commercial cooking and on-road traffic, how are these two sectors typically distributed in the default surrogates? Please consider elaborating in Section 3 to provide some context to the reader.
The spatial distributions of cooking and traffic emissions using both the new and old surrogates are illustrated in the Figures S5, S6, S7 and S8 located within the supplementary material. This was not made clear in the main text and some text has been added to both explain the approach used and to also point the reader to these additional figures.
(2) In the PMCAMx model simulations, are there removal processes such as dry deposition and wet scavenging? If so, how were these handled between the different grid resolutions? Are these something that must be interpolated from WRF or is it something that PMCAMx explicitly captures?
These processes are included in PMCAMx, and this has been clarified in the included model description. Wet deposition is dependent on precipitation, which indeed relies upon the meteorological input data. The need to interpolate meteorology is a limitation of the study, however the evaluation process described in the supplementary material (Figures S1 and S2 with corresponding discussion) provides confidence in this specific application.
(3) The WRF simulations were conducted at 12 km resolution with the information interpolated to higher resolutions of 1 km and 4 km. First, what was the procedure to create 36 km inputs for PMCAMx? Second, why was this design choice made rather than simulating WRF at 36 km and interpolating for all higher resolution grids? Or perhaps simulate meteorology at 4 km with interpolating to 1 km grid and extrapolating to 12 and 36 km grids?
The WRF simulations and PMCAMx simulations use the same domain, meaning we can directly aggregate the nine 12 km x 12 km cells that make up a single 36 km x 36 km cell to create input data for the coarse grid. For the higher resolution grids, we use the interpolation process used in previous work (Fountoukis et al., 2013) that has been evaluated extensively for use in higher resolution air quality studies. An evaluation for the specific meteorological data used in this study is provided for the reader in the supplementary material (Figures S1 and S2 with corresponding discussion).
(4) For those unfamiliar with emissions processing and emission surrogates, it may prove useful to further emphasis how using different surrogates merely change the spatial distribution of the total amount. The authors perhaps made this clear in Section 6 and the Supplementary material, but the reader could benefit from this stated more clearly and earlier.
A short description of this is now provided in Section 3 (Model Application). A clarifying discussion about how these surrogates might change the emissions provided to the model has been added.
(5) Related to the previous point, at times in the paper it reads like the novel surrogates for commercial cooking and on-road emissions were applied to all grid resolution configurations and other times it reads like they were only applied to 1 and 4 km simulations (see Line 183-184).
They were used for the 1 km simulations. Conflicting statements have been removed and a clarifying note has been added to the caption of Table 4 highlighting that these results are for the high-resolution simulations only.
(6) There is no mention of the novel emission surrogates for commercial cooking and on-road emissions in the abstract. I feel like this point could use some emphasis in terms of mentioning that these were used and developed and briefly mention the evaluation of these novel surrogates.
A sentence has been added to the abstract, noting the use and evaluation of new surrogates for these sources at high spatial resolution.
(7) The paper could in general benefit from making it more clear how this study is both related and also different from Rivera et al 2022.
These studies are of course related. This work is meant to be a detailed quantitative evaluation of the simulation results used to investigate the changes in predicted concentration and exposure with increasing resolution in Garcia Rivera et al. (2022). None of this evaluation is included in the previous work. A clarifying statement has also been added, noting the additional simulations performed to quantify any differences in the use of the novel spatial surrogates for commercial cooking and on-road traffic.
Minor comments
(8) Table 4 could benefit from explicitly mentioning in the caption that it is for the 1 x1 km resolution.
The recommended change has been made in Table 4.
(9) In supplementary material, the captions for Figure S7 and S8 should be checked and corrected to avoid any confusion. Figures S7 and S8 captions incorrectly start off with mentioned commercial cooking and also both for February 2017. Presumably, S7 and S8 are for on-road traffic for February and July respectively as they mention using simulated traffic approach.
These are correctly identified typos. They have been corrected in the manuscript.
Status: closed
-
RC1: 'Comment on gmd-2022-145', Anonymous Referee #1, 29 Aug 2022
This manuscript presents an application of the CTM PMCAMx-v2.0 at increasing resolutions of 36 x 36, 12 x 12, 4 x 4, and 1 x 1 km for a domain covering Allegheny County and the city of Pittsburgh in southwestern Pennsylvania, USA to predict PM2.5 and its components. In general, the paper is very well structured and written, supported by relevant past works. The results are adequately presented, and extensively analysed covering the assessment of model performance for PM2.5 and its components, the influence of different horizontal resolutions and of the improvement of emissions allocation for cooking and traffic.
However, I have two main concerns regarding the presented work:
- The analysis presented and discussed is based on a modelling run for only two months – February and July – and the discussion and conclusions are drawn for Winter and Summer periods. An example is the sentence in line 327 “Total PM2.5 mass concentrations are underpredicted in the summer period”. It is clear that those months were chosen due to the availability of measurements, particularly PM2.5 compounds. Notwithstanding, the study would be much more robust and conclusion better supported if the modelling period was at least one year. Results could be assessed with the observations from fixed monitoring sites for the whole year, with a focus on those two months for which more data is available.
- The authors refer that “Garcia Rivera et al. (2022) investigated the effects of increasing grid resolution of model inputs and CTM output on source resolved predictions of PM2.5 concentration and population exposure at 36 km, 12 km, 4 km, and 1 km”. Is the current work based on the simulations performed by Garcia Rivera et al. (2022) or are there additional runs in the present work? Are the runs referred in lines 309 – 310 examples of additional simulations? What exactly is the novelty of this study compared to (Garcia Rivera et al.,2022)? This should be made clear in the manuscript and perhaps the “Model Application” section could be shortened since it is already explained in the published paper.
Specific comments
- The abstract is too long and has many details of the analysis of results that are unnecessary in an abstract. Conversely, the Conclusions lack a more comprehensive and focused text on the added value and limitations of this work and concrete ideas for future work.
- Section 3 has a subsection 3.1, that would only make sense if a 3.2 exists. Please restructure this section or remove the subsection title. One idea is to join sections 2 and 3 into one section.
- Figure 1 – Please include the metric scale, and draw the inner domain to allow to have an idea of where stations are located in relation to the modelling domains.
Technical corrections
- Line 323 - The reference (Kodros et al., 2021) is missing in the reference list.
- References Day et al., 2015, Skyllakou et al., 2021, included in the reference list are not cited in the text.
-
AC1: 'Response to the Comments of Referee 1', Spyros Pandis, 12 Oct 2022
Major comments
(1) The analysis presented and discussed is based on a modelling run for only two months – February and July – and the discussion and conclusions are drawn for Winter and Summer periods. An example is the sentence in line 327 “Total PM2.5 mass concentrations are underpredicted in the summer period”. It is clear that those months were chosen due to the availability of measurements, particularly PM2.5 compounds. Notwithstanding, the study would be much more robust and conclusion better supported if the modelling period was at least one year. Results could be assessed with the observations from fixed monitoring sites for the whole year, with a focus on those two months for which more data is available.
We do agree that longer evaluations are always better. However, significant resources are required to develop the emission inventory used for these PMCAMx simulations, so two months in two different seasons were chosen as a compromise. These simulations can provide sufficient information, so that the ability of the model to capture PM2.5 differences due to the seasonal variability in major emissions sources and meteorology can be adequately assessed. The primary goal of this study is to evaluate the predictions of PM2.5 concentrations at high spatial resolution (1 km x 1 km), produced using a corresponding high spatial resolution source-resolved emissions inventory. The high-resolution, source-resolved, and species resolved PM2.5 predictions evaluated in this study are novel and require significant evaluation against available measurements which can be sparse (especially for PM2.5 component analysis). Another factor in our decision has been the connection of the present work with the corresponding analysis and conclusions of Garcia Rivera et al. (2022) study who also studied the corresponding periods. A brief explanation and discussion of this point has been added in the revised paper.
(2) The authors refer that “Garcia Rivera et al. (2022) investigated the effects of increasing grid resolution of model inputs and CTM output on source resolved predictions of PM2.5 concentration and population exposure at 36 km, 12 km, 4 km, and 1 km”. Is the current work based on the simulations performed by Garcia Rivera et al. (2022) or are there additional runs in the present work? Are the runs referred in lines 309 – 310 examples of additional simulations? What exactly is the novelty of this study compared to (Garcia Rivera et al.,2022)? This should be made clear in the manuscript and perhaps the “Model Application” section could be shortened since it is already explained in the published paper.
This study uses the results of some of the simulations of Garcia Rivera et al. (2022) and adds several additional simulations. For example, additional simulations were performed to quantify the impact of the proposed emissions surrogates for traffic and cooking. The work of Garcia Rivera et al. (2022) is an analysis of the impact of increasing grid resolution on predicted human exposure to PM2.5 and the corresponding links with the PM2.5 sources. Garcia Rivera et al. (2022) did not address model performance and the corresponding challenges related to the different types of the available measurements. This evaluation is the primary purpose of the present study. The Model Application section now contains only the most essential information about the simulations so that the interested reader does not need to consult Garcia Rivera et al. (2022) to understand the major features of our simulations. These points have been clarified in the revised manuscript text.
Specific comments
(3) The abstract is too long and has many details of the analysis of results that are unnecessary in an abstract. Conversely, the Conclusions lack a more comprehensive and focused text on the added value and limitations of this work and concrete ideas for future work.
The content of the abstract and conclusion have been reassessed and changes have been made in the manuscript. Additional clarity on the scope of the work has also been highlighted in the revised manuscript.
(4) Section 3 has a subsection 3.1, that would only make sense if a 3.2 exists. Please restructure this section or remove the subsection title. One idea is to join sections 2 and 3 into one section.
This has been fixed in the manuscript.
(5) Figure 1 – Please include the metric scale and draw the inner domain to allow to have an idea of where stations are located in relation to the modelling domains.
Figure 1 (a) shows fully the entire inner domain. Figure 1 (b) shows the City of Pittsburgh limits, which are also shown in Figure 1 (a) to provide the reader with a scale reference. These notes have been included in the revised caption for clarity.
Technical Corrections
(6) Line 323 - The reference (Kodros et al., 2021) is missing in the reference list.
The missing reference has been added.
(7) References Day et al., 2015, Skyllakou et al., 2021, included in the reference list are not cited in the text.
These unused references have been removed.
-
RC2: 'Comment on gmd-2022-145', Anonymous Referee #2, 01 Sep 2022
Dinkelacker et al 2022 investigates the benefit of high-resolution modeling and emissions surrogates for the prediction of PM2.5 and its composition using PMCAMx. The authors present model results and measurement comparisons for model simulation at 36 km, 12 km, 4 km and 1 km simulations, to southwestern Pennsylvania while applying novel surrogates for on-road and commercial cooking emissions. Previously in the work of Rivera et al 2022 evaluated the effects of grid resolution for model predictions of particulate matter. This study enhances that work performing a model evaluation by comparing the modeling approach against EPA and low-cost sensor measurements. The study compares PM2.5 in Winter and Summer and for EPA monitor sites shows that fractional error and fractional bias improved by increasing resolution from 36 km to 1 km. In general, PM2.5 prediction performance improved in almost all cases. The underpredictions of PM2.5 at urban sites and overpredictions at rural sites were reduce as a result of increasing resolution.
Overall this paper is well written and logically organized. The subject matter is one that should appeal to atmospheric chemistry and air quality modelers. This study is valuable in showing the importance in developing improved emission inputs for models in particular for the spatial distribution to emissions from an inventory to a model grid. While a focus on improving model resolution to improve model results has its clear benefits, this study shows clearly that when we increase spatial resolution, it should be accompanied by improved and more accurate emission inputs.Comments:
- While it is clear that the different spatial distribution has an impact on the distribution of emissions for commercial cooking and on-road traffic, how are these two sectors typically distributed in the default surrogates? Please consider elaborating in Section 3 to provide some context to the reader.
- In the PMCAMX model simulations, are there removal processes such as dry deposition and wet scavenging? If so, how were these handled between the different grid resolutions? Are these something that must be interpolated from WRF or is it something that PMCAMx explicitly captures?
- The WRF simulations were conducted at 12 km resolution with the information interpolated to higher resolutions of 1 km and 4 km. First, what was the procedure to create 36 km inputs for PMCAMx? Second, why was this design choice made rather than simulating WRF at 36 km and interpolating for all higher resolution grids? Or perhaps simulate meteorology at 4 km with interpolating to 1 km grid and extrapolating to 12 and 36 km grids?
- For those unfamiliar with emissions processing and emission surrogates, it may prove useful to further emphasis how using different surrogates merely change the spatial distribution of the total amount. The authors perhaps made this clear in Section 6 and the Supplementary material but the reader could benefit from this stated more clearly and earlier.
- Related to the previous point, at times in the paper it reads like the novel surrogates for commercial cooking and on-road emissions were applied to all grid resolution configurations and other times it reads like they were only applied to 1 and 4 km simulations (see Line 183-184).
- There is no mention of the novel emission surrogates for commercial cooking and on-road emissions in the abstract. I feel like this point could use some emphasis in terms of mentioning that these were used and developed and briefly mention the evaluation of these novel surrogates.
- The paper could in general benefit from making it more clear how this study is both related and also different from Rivera et al 2022.
Minor comments:
- Table 4 could benefit from explicitly mentioning in the caption that it is for the $1 \times 1$ km resolution.
- In supplementary material, the captions for Figure S7 and S8 should be checked and corrected to avoid any confusion. Figures S7 and S8 captions incorrectly start off with mentioned commercial cooking and also both for February 2017. Presumably, S7 and S8 are for on-road traffic for February and July respectively as they mention using simulated traffic approach.
-
AC2: 'Response to the Comments of Referee 2', Spyros Pandis, 12 Oct 2022
Comments
(1) While it is clear that the different spatial distribution has an impact on the distribution of emissions for commercial cooking and on-road traffic, how are these two sectors typically distributed in the default surrogates? Please consider elaborating in Section 3 to provide some context to the reader.
The spatial distributions of cooking and traffic emissions using both the new and old surrogates are illustrated in the Figures S5, S6, S7 and S8 located within the supplementary material. This was not made clear in the main text and some text has been added to both explain the approach used and to also point the reader to these additional figures.
(2) In the PMCAMx model simulations, are there removal processes such as dry deposition and wet scavenging? If so, how were these handled between the different grid resolutions? Are these something that must be interpolated from WRF or is it something that PMCAMx explicitly captures?
These processes are included in PMCAMx, and this has been clarified in the included model description. Wet deposition is dependent on precipitation, which indeed relies upon the meteorological input data. The need to interpolate meteorology is a limitation of the study, however the evaluation process described in the supplementary material (Figures S1 and S2 with corresponding discussion) provides confidence in this specific application.
(3) The WRF simulations were conducted at 12 km resolution with the information interpolated to higher resolutions of 1 km and 4 km. First, what was the procedure to create 36 km inputs for PMCAMx? Second, why was this design choice made rather than simulating WRF at 36 km and interpolating for all higher resolution grids? Or perhaps simulate meteorology at 4 km with interpolating to 1 km grid and extrapolating to 12 and 36 km grids?
The WRF simulations and PMCAMx simulations use the same domain, meaning we can directly aggregate the nine 12 km x 12 km cells that make up a single 36 km x 36 km cell to create input data for the coarse grid. For the higher resolution grids, we use the interpolation process used in previous work (Fountoukis et al., 2013) that has been evaluated extensively for use in higher resolution air quality studies. An evaluation for the specific meteorological data used in this study is provided for the reader in the supplementary material (Figures S1 and S2 with corresponding discussion).
(4) For those unfamiliar with emissions processing and emission surrogates, it may prove useful to further emphasis how using different surrogates merely change the spatial distribution of the total amount. The authors perhaps made this clear in Section 6 and the Supplementary material, but the reader could benefit from this stated more clearly and earlier.
A short description of this is now provided in Section 3 (Model Application). A clarifying discussion about how these surrogates might change the emissions provided to the model has been added.
(5) Related to the previous point, at times in the paper it reads like the novel surrogates for commercial cooking and on-road emissions were applied to all grid resolution configurations and other times it reads like they were only applied to 1 and 4 km simulations (see Line 183-184).
They were used for the 1 km simulations. Conflicting statements have been removed and a clarifying note has been added to the caption of Table 4 highlighting that these results are for the high-resolution simulations only.
(6) There is no mention of the novel emission surrogates for commercial cooking and on-road emissions in the abstract. I feel like this point could use some emphasis in terms of mentioning that these were used and developed and briefly mention the evaluation of these novel surrogates.
A sentence has been added to the abstract, noting the use and evaluation of new surrogates for these sources at high spatial resolution.
(7) The paper could in general benefit from making it more clear how this study is both related and also different from Rivera et al 2022.
These studies are of course related. This work is meant to be a detailed quantitative evaluation of the simulation results used to investigate the changes in predicted concentration and exposure with increasing resolution in Garcia Rivera et al. (2022). None of this evaluation is included in the previous work. A clarifying statement has also been added, noting the additional simulations performed to quantify any differences in the use of the novel spatial surrogates for commercial cooking and on-road traffic.
Minor comments
(8) Table 4 could benefit from explicitly mentioning in the caption that it is for the 1 x1 km resolution.
The recommended change has been made in Table 4.
(9) In supplementary material, the captions for Figure S7 and S8 should be checked and corrected to avoid any confusion. Figures S7 and S8 captions incorrectly start off with mentioned commercial cooking and also both for February 2017. Presumably, S7 and S8 are for on-road traffic for February and July respectively as they mention using simulated traffic approach.
These are correctly identified typos. They have been corrected in the manuscript.
Brian T. Dinkelacker et al.
Brian T. Dinkelacker et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 291 | 88 | 13 | 392 | 27 | 3 | 4 |
- HTML: 291
- PDF: 88
- XML: 13
- Total: 392
- Supplement: 27
- BibTeX: 3
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1