NEIVAv1.0: Next-generation Emissions InVentory expansion of Akagi et al. version 1.0

This paper presents NEIVA, a new inventory of emission factors (EFs), emission ratios (ERs), and particle properties for biomass burning for 14 globally relevant fuel and fire types (e.g. temperate forest, boreal forest, cookstove, garbage burning). The inventory is an update and expansion of the widely used Akagi et al. (2011) emission factor inventory (hereafter Akagi) which was last updated in 2015. NEIVA provides significant updates to Akagi by incorporating new emissions data (2014 – 2023) available from recent advances in instrumentation and many recent laboratory and field studies. NEIVA expands upon Akagi in several significant ways – beginning with its structure: original, raw, intermediate, and final datasets are stored in connected tables. These tables include the original Akagi inventory, “raw” datasets from recent studies (2014-2023), recommended fuel and fire type average EFs as the intermediate dataset created along the way from integration and processing of Akagi and data from new studies. The data tables are accessible in a GitHub repository that includes tools to query and explore the data from the individual study level to recommended fuel and fire type averages derived from all of laboratory and field. The GitHub repository also includes the scripts used to generate the intermediate and recommended data tables as well as scripts for updating NEIVA with new data. NEIVA includes chemical and physical property data and model surrogate assignments for three widely used chemical mechanisms (SAPRC, MOZART-T1, and GEOS-Chem) for each of ~1000 non-methane organic compounds (NMOC). The paper includes a comparison of NEIVA recommended EFs with previous compilations and summarizes potential implications of NEIVA for atmospheric chemistry. The authors find that NEIVA recommended EFs result in emission mixture with greater fraction of intermediate volatility NMOG and when mapped to chemical mechanism show higher OH reactivity than previous EF compilations.

NIEVA will be a great benefit to researchers concerned with the effects of biomass burning on atmospheric composition, air quality, climate, and public health, especially those working at continental to global scales. The authors inclusion of laboratory studies is especially valuable, and I believe they have demonstrated inclusion of these data, following ER-based adjustments, is unlikely to introduce a systemic bias in the recommended EFs. The aggregation of global data into broad fuel and fire types and the approach chosen for averaging may make the recommended EFs less than ideal for regional (e.g. western U.S. and Canada) and finer scale applications. For example, temperate forest EF using equal weighting of studies based on many fires (e.g. Permar et al. and Gkzatelis et al.) with those based on only a few (e.g. Liu et al. 2017 and Travis et al. Blackwater fire). However, since NEIVA provides access to study level datasets, it also provides EF data needed for regionally focused applications.

I have only a few minor comments for the authors to address.

1. The GitHub repository does not include “.csv” tables as described in the manuscript; however, I was able to create and download using the provided tools (in Google Colab).
2. The inclusion of Travis et al. (2023) for prescribed burns of shrublands in the central U.S. with chaparral seems a bit odd. Did the authors consider introducing a separate shrubland category?
3. L346-348: “Rice straw EFs measured during a FIREX laboratory pile-burning simulation also were included (Koss et al.,2018; Selimovic et al., 2018; Gkatzelis et al., 2023; Travis et al., 2023).” Travis et al. and Gkatzelis et al. are both field studies and the latter does not report crop residue emissions. The sentence or references need updating.
4. Table S14. Travis et al. (2023) is a field study, the eastern portion of FIREX-AQ 2019. However, Table S14 lists Travis et al. (2023) as laboratory measurements and indicates the study’s EFs were adjusted in the creation of the NIEVA Processed EF dataset. The intermediate datasets (“.csv”) I extracted indicate the Travis et al. was correctly processed as field data suggesting the table is in error. However, I did not re-run the processing scripts to verify this is the case. This should be confirmed.
5. L507-508: “For boreal forest, the relatively high laboratory-based CO value is largely driven by EFs measured in boreal peat studies and reported by Yokelson et al. (1997).” Lab measurements of boreal peat should be included with peat. Is this inclusion in boreal forest fires a legacy of Akagi et al. (2011)? May have been better to extract subset of studies from Akagi et al (2011) instead of using full dataset if one lab burn has such a big impact.
6. A peat emissions field study worthy of consideration for future updates: Geron & Hays (2013) Air emissions from organic soil burning on the coastal plain of North Carolina, Atmospheric Environment, Volume 64, 2013, Pages 192-199, ISSN 1352-2310,https://doi.org/10.1016/j.atmosenv.2012.09.065.

General comments

This work compiles emission factors available in the literature, from laboratory and field experiments, to create an updated inventory of emissions from biomass burning for different fuel and fire types, with a focus on gaseous species. The data provided in this database is of great importance for the atmospheric chemistry modeling community. The proposed mapping onto commonly used chemical mechanisms (referred to as “surrogate” in the manuscript) is a great value-added that makes the database practical to use. The methodology is well documented and the differences with other inventories are thoroughly discussed. The discussion on applications and impacts in modeling is a good addition. The manuscript is very well written and easy to follow.

Minor comments

The PM category “PM2.5* = PM1-PM5” is not very clearly defined. Is it PM between 1μm and 5μm? Why report this particular category, which is not very common? I think some clarification is needed here.

Section 5: the manuscript is generally quite long and I do not think Section 5 helps the manuscript, especially since the Github is well commented and documented. I would take out Section 5 or move it to supplementary material.

Supplementary material: some references need fixing (“Error! Reference source not found.” several times)