Data from: Projecting nitrous oxide over the 21st century, uncertainty related to stratospheric loss
Data files
Jan 23, 2026 version files 910.06 KB
-
2025_N2O_PNAS-datasets.xlsx
86.89 KB
-
2025PNAS_N2O_Figures.pdf
813.47 KB
-
README.md
9.70 KB
Abstract
This data set includes the analysis and graphical data in the PNAS publication entitled "Projecting nitrous oxide over the 21st century, uncertainty related to stratospheric loss". The data is in the form of column vectors of various derived products involving N2O, NOy, O3, and QBO indices that are plotted in the paper. The abscissa can be year, month, or lag time (months). The data is provided in an Excel spreadsheet with tabs for each figure. The data is available for open use, and there are no ethical or legal considerations related to its use in subsequent research. The abstract and significance statement of the publication follow.
Abstract. Extending the N2O lifetime derived from Microwave Limb Sounder satellite observations, we find a mean value of 117 yr and a likely decrease of –1.4 ± 0.9 % per decade over the period 2004-2024. This trend is consistent with the previously published 2004-2021 value of –2.1 ± 1.2 % per decade. A more careful analysis of uncertainty now provides a more robust likely (one-sigma) range. From analyses of a range of factors controlling the N2O lifetime, we find that the decrease in lifetime can be explained by recent changes in stratospheric circulation and temperature. Projection of the lifetime change to 2100 shows that this effect is comparable to differences across the shared socioeconomic pathways used for climate projections and cannot be ignored. An updated evaluation of the N2O chemical feedbacks shows that this effect produces a relatively small shift in atmospheric abundance over the 21st century, but still an important shift, –11%, in the global warming potential of N2O.
Significance. Projecting atmospheric nitrous oxide (N2O) abundance is critical for climate and ozone assessments. Research has focused on projecting the changing emissions of N2O from direct anthropogenic sources, the dominant cause of the recent growth. Earth system models are now projecting natural sources perturbed by climate change. There has been little effort to understand how climate and compositional changes may change the stratospheric sink of N2O, which balances all these sources and also controls the atmospheric abundance. Here, we review recent observational and modeling evidence for an increase in the sink caused by decreasing N2O lifetime and show that it introduces uncertainties comparable to shifts across the different shared socioeconomic pathway (SSP) scenarios used in current assessments.
Dataset DOI: 10.5061/dryad.vmcvdnd6m
Description of the data and file structure
Description of the data and file structure
The data here are derived from public sources and are in the form of monthly means or 12-month running sums. The data from the analysis are provided in a spreadsheet (.xlsx format).
Files and variables
Files and variables
File: 2025PNAS_N2O_Figures.pdf
Description: Published figures along with captions. The captions give a more detailed description of the data being plotted. An alternative to Fig. 4, using the earlier N2O scenarios (RCPs), is provided.
File: 2025_N2O_PNAS-datasets.xlsx
Description: Monthly data used in this analysis and shown in the published figures are given in the spreadsheet. This analysis is accepted for publication in PNAS as # 2025-24123R on 26 Dec 2025.
Tab: monthly data
These monthly data are averages over the month and are plotted as the middle of the month. Please note that blank cells mean that there is NO DATA available for these times and the analysis did not use any assumed values for the variables.
A. month (#1 = Jan 2004)
B. year (and month as a fraction of the year)
C. N2O (ppb, mol/mol)
D. N2O Loss (kg-N/sec)
E. NOy Prod (kg-N/sec)
F. N2O Mass (Tg-N)
G. days / month
H. u wind @ 10 hPa from Singapore sonde (m/s)
Tab: annual mean data
These annual mean data are calculated as the day-weighted average of the monthly means. They are plotted at the first of each month, i.e., the 12-month running average of Jan through Dec of 2005 is plotted at 1 Jul 2005 (2005.5). All months are treated as being of equal length (1/12 of a year) for plotting purposes, but each monthly mean is weighted by the days in the month. February monthly data are always weighted by 28.25 days to give an accurate annual mean, assuming that the year is 365.25 days long (Julian calendar).
A. year
B. N2O (ppb, mol/mol)
C. N2O Loss (Tg-N/yr)
D. NOy Prod (Tg-N/yr)
E. N2O Mass (Tg-N)
Tab: Fig. 1abcd
A. X-axis (mon)
B. Fig1a L-N2O (loss in Tg-N/yr)
C. Fig1b NOy yield (% N/N)
D. Fig1c Altitude (km) of L-N2O
E. Fig1c Altitude (km) of P-NOy
F. Fig1d Latitude (deg) of L-N2O
G. Fig1d Latitude (deg) of P-NOy
Caption: Fig. 1. (a) Monthly annual cycle of N2O loss rate (TgN/y) calculated from the MLS monthly data for N2O, O3, and T for Jan 2005 through Dec 2024. (b) Annual cycle of NOy yield (% per N2O lost as N per N). (c) Mean altitude (km, in pressure altitude) of the N2O loss ('o') and NOy production ('x'). (d) Mean latitude (degrees) of the N2O loss ('o') and NOy production ('x').
Tab: Fig. 2abcd
A. X-axis (yr)
B. Fig2a N2O mass (Tg-N)
C. Fig2a linear fit to N2O mass
D. Fig2b L-N2O (loss in Tg-N/yr)
E. Fig2b linear fit to L-N2O
F. Fig2c NOy yield (%N;N)
G. Fig2c linear fit to NOy yield
H. Fig2d N2O lifetime (yr)
I. Fig2d linear fit to N2O lifetime
J. Fig2d linear fit to N2O lifetime using only years 2004-2021
Caption: Fig. 2. Interannual variability (IAV) of key N2O quantities. All data here have the annual cycle removed (see Fig. 1) by summing or averaging over 1 year with daily weighting for each month from Aug 2004 through Dec 2024. Every Feb has 28.25 days. The first point covers Aug 2004 through Jul 2005 and is plotted as 1 Feb 2005. The last point is the sum over Jan 2024 through Dec 2024 and plotted as 1 Jul 2024. (a) N2O total atmospheric mass (TgN) derived from NOAA monthly global mean surface observations (ref) and a scaling factor of 4.79 TgN/ppb. The linear trend line fit (+3.0 %/dec) shows the positive residuals of the fit at both ends, indicating the parabolic pattern of the N2O growth rate, corresponding to a near uniform acceleration of 2.1 % yr2. (b) N2O loss rate (TgN/y) calculated from the MLS monthly data for N2O, O3, and T for Aug 2004 through Dec 2024. The trend line (+4.4 %/dec) shows that N2O loss is increasing faster than the burden. The IAV is much smaller than the annual cycle (Fig. 1a) and correlates with QBO winds (Fig. 3a). (c) Yield rate of NO (%) in terms of moles of N produced per moles of N lost of N2O. (d) Annual mean lifetime of N2O (y) derived from panels (a) and (b). The solid black trend line (–1.4 %/dec) for the extended period is compared with that from previous analysis using data through Dec 2021 (dotted line, –2.1 %/dec, Prather et al., 2023). The mean lifetime for either period is 117.3 y.
Tab: Fig. 3abc
A. X-axis (lag in mon)
B. Fig3a Xcorr (cross correlation) of u10 v. L-N2O
C. Fig3a Xcorr of u10 v. P-NOy
D. Fig3b Xcorr of P-NOy vs. L-N2O
E. Fig3c Xcorr of L-N2O vs. w @ 30hPa
F. Fig3c Xcorr of L-N2O vs. w @ 20 hPa
G. Fig3c Xcorr of L-N2O vs. w @ 10 hPa
H. Fig3c Xcorr of L-N2O vs. w @ 7 hPa
I. Fig3c Xcorr of L-N2O vs. w @ 5 hPa
Caption: Fig. 3. (a) Cross correlation of detrended N2O loss ('o') and NOy production ('x') against QBO metric (u10 = equatorial wind at 10 hPa), showing peak correlation with N2O-loss of –0.47 at –4 months (vertical dotted line, N2O lagging QBO). All data have been yearly averaged, month-by-month. (b) Cross correlation of NOy production against N2O loss, showing peak correlation of 0.98 and no significant lag. (c) Cross correlation of N2O loss against residual vertical velocity w*QBO at pressure levels: 30, 20, 10, 7, 5 hPa.
Tab: Fig. 4
A. X-axis (years)
B. projected N2O abundance (ppb) assuming constant growth of +3 %/dec, and fixed lifetime
C. projected N2O abundance assuming lifetime decrease of -2.1 %/dec
D. projected N2O abundance assuming lifetime decrease of -1.4 %/dec
E. projected N2O abundance assuming lifetime decrease of -1.4 %/dec and including chemical feedback
F. (intentionally left blank)
G. second X-axis (decadal years)
H. SSP1-2.6 decadal projections for N2O (ppb)
I. SSP2-4.5 ibid
J. SSP3-7.0 ibid
K. SSP4-6.0 ibid
L. SSP5-8.5 ibid
Caption: Figure 4. Observed (2000-2024) and projected (2025-2100) N2O annual mean tropospheric abundance (gray line) assuming the observed 2005-2024 +3.0 %/decade growth rate continues to 2100. N2O emissions (2025-2100) are calculated from this projection assuming a fixed lifetime of 120 yr. Using these emissions, the N2O abundance is projected (black lines) using three different lifetime assumptions: (solid line) lifetime decreases at –1.4 % per decade; (thin dotted line) same lifetime trend plus a chemical feedback factor that reduces the lifetime with increasing burden (i.e., dln(N2O lifetime) / dln(N2O burden) = –0.065; Prather and Hsu, 2010); and (dashed line) the lifetime decreases at –2.1 % per decade as found from the earlier period 2004-2021 (Prather et al., 2023). The reduction in 2100 N2O abundance for these three cases are –4.3, –4.5, and –6.5 %, respectively. A range of SSP emissions scenario projections (Shared Socio-economic Pathways: SSP1-2.6, SSP2-4.5, SSP5-8.5, SSP4-6.0, SSP3-7.0, in ascending order) calculated by Meinshausen et al. (2020) for the IPCC assessments are shown for perspective.
Code/software
Code/software
The Figure/ data sets are given in a .xlsx file with separate page tabs for each Figure.
Access information
Access information
Data used in this analysis was collected from the following sources:
N2O surface abundance from NOAA: G.S. Dutton et al., Combined Atmospheric Nitrous Oxide Dry Air Mole Fractions from the NOAA GML Halocarbons Sampling Network, 1977-2024, Version: 2024-02-21, doi: 10.15138/GMZ7-2Q16, creation date: 14 Mar 2025 (Xin Lan), https://gml.noaa.gov/webdata/ccgg/trends/N2O/N2O_mm_gl.txt, (accessed 30 March 2025)
MLS O3profiles from JPL: M. Schwartz, L. Froidevaux, N. Livesey, W. Read, R. Fuller, MLS/Aura Level 3 Monthly Binned Ozone (O3) Mixing Ratio on Assorted Grids V005, Greenbelt, MD, USA, Goddard Earth Sciences Data and Information Services Center (GES DISC), doi: 10.5067/Aura/MLS/DATA/3546 (accessed 30 March 2025).
MLS temperature profiles from JPL: M. Schwartz, N. Livesey, W. Read, R. Fuller, MLS/Aura Level 3 Monthly Binned Temperature on Assorted Grids V005, Greenbelt, MD, USA, Goddard Earth Sciences Data and Information Services Center (GES DISC), doi: 10.5067/Aura/MLS/DATA/3550 (accessed 30 March 2025).
MLS N2O profiles from JPL: A. Lambert, N. Livesey, W. Read, R. Fuller, MLS/Aura Level 3 Monthly Binned Nitrous Oxide (N2O) Mixing Ratio on Assorted Grids V005, Greenbelt, MD, USA, Goddard Earth Sciences Data and Information Services Center (GES DISC), doi: 10.5067/Aura/MLS/DATA/3545 (accessed 30 March 2025).
Singapore u winds at 10 hPa: T. Kerzenmacher, P. Braesicke, QBO: monthly zonal stratospheric winds from tropical radiosonde data (mainly Singapore), https://zenodo.org/records/14037052, (accessed 20 Jum 2025)
MERRA-2 w profiles from NASA:* P.A. Newman, L. Coy, S. Pawson, The Quasi-biennial Oscillation (QBO), Vertical residual velocity (mm/s) monthly means from MERRA-2 (5S-5N), 1980/01-2025/05, https://acd-ext.gsfc.nasa.gov/Data_services/met/qbo/qbo.html, and https://acd-ext.gsfc.nasa.gov/Data_services/met/qbo/QBO_MERRA2-Wvals_05S-05N_GSFC.txt (accessed 23 Jun 2025)