Climate Change 2001: Working Group I: The Scientific Basis

	Other reports in this collection

4.2.1.2 Nitrous oxide (N₂O)

The globally averaged surface abundance of N₂O was 314 ppb in 1998, corresponding to a global burden of 1510 TgN. N₂O abundances are about 0.8 ppb greater in the Northern Hemisphere than in the Southern Hemisphere, consistent with about 60% of emissions occurring in the Northern Hemisphere. Almost no vertical gradient is observed in the troposphere, but N₂O abundances decrease in the stratosphere, for example, falling to about 120 ppb by 30 km at mid-latitudes.

The known sources of N₂O are listed in Table 4.4 with estimates of their emission rates and ranges. As with methane, it remains difficult to assess global emission rates from individual sources that vary greatly over small spatial and temporal scales. Total N₂O emissions of 16.4 TgN/yr can be inferred from the N₂O global sink strength (burden/lifetime) plus the rate of increase in the burden. In the SAR the sum of N₂O emissions from specific sources was notably less than that inferred from the loss rate. The recent estimates of global N₂O emissions from Mosier et al. (1998b) and Kroeze et al. (1999) match the global loss rate and underline the progress that has been made on quantification of natural and agricultural sources. The former study calculated new values for N₂O agricultural emissions that include the full impact of agriculture on the global nitrogen cycle and show that N₂O emissions from soils are the largest term in the budget (Table 4.4). The latter study combined these with emissions from other anthropogenic and natural sources to calculate a total emission of 17.7 TgN/yr for 1994.

The enhanced N₂O emissions from agricultural and natural ecosystems are believed to be caused by increasing soil N availability driven by increased fertilizer use, agricultural nitrogen (N₂) fixation, and N deposition; and this model can explain the increase in atmospheric N₂O abundances over the last 150 years (Nevison and Holland, 1997). Recent discovery of a faster-than-linear feedback in the emission of N₂O and NO from soils in response to external N inputs is important, given the projected increases of N fertilisation and deposition increases in tropical countries (Matson et al., 1999). Tropical ecosystems, currently an important source of N₂O (and NO) are often phosphorus-limited rather than being N-limited like the Northern Hemispheric terrestrial ecosystems. Nitrogen fertiliser inputs into these phosphorus-limited ecosystems generate NO and N₂O fluxes that are 10 to 100 times greater than the same fertiliser addition to nearby N-limited ecosystems (Hall and Matson, 1999). In addition to N availability, soil N₂O emissions are regulated by temperature and soil moisture and so are likely to respond to climate changes (Frolking et al., 1998; Parton et al., 1998). The magnitude of this response will be affected by feedbacks operating through the biospheric carbon cycle (Li et al., 1992, 1996).

The industrial sources of N₂O include nylon production, nitric acid production, fossil fuel fired power plants, and vehicular emissions. It was once thought that emission from automobile catalytic converters were a potential source of N₂O, but extrapolating measurements of N₂O emissions from auto-mobiles in roadway tunnels in Stockholm and Hamburg during 1992 to the global fleet gives a source of only 0.24 ± 0.14 TgN/yr (Berges et al., 1993). More recent measurements suggest even smaller global emissions from automobiles, 0.11 ± 0.04 TgN/yr (Becker et al., 1999; Jiménez et al., 2000).

The identified sinks for N₂O are photodissociation (90%) and reaction with electronically excited oxygen atoms (O(¹D)); they occur in the stratosphere and lead to an atmospheric lifetime of 120 years (SAR; Volk et al., 1997; Prinn and Zander, 1999). The small uptake of N₂O by soils is not included in this lifetime, but is rather incorporated into the net emission of N₂O from soils because it is coupled to the overall N-partitioning.

Figure 4.2: Change in N2O abundance for the last 1,000 years as determined from ice cores, firn, and whole air samples. Data sets are from: Machida et al. (1995); Battle et al. (1996); Langenfelds et al. (1996); Steele et al. (1996); Flückiger et al. (1999). Radiative forcing, approximated by a linear scale, is plotted on the right axis. Deseasonalised global averages are plotted in the inset (Butler et al., 1998b).

Isotopic (¹⁵N and ¹⁸O) N₂O measurements are also used to constrain the N₂O budget. The isotopic composition of tropo-spheric N₂O derives from the flux-weighted isotopic composition of sources corrected for fractionation during destruction in the stratosphere. Typical observed values are ¹⁵N = 7 ‰ and ¹⁸O = 20.7 ‰ relative to atmospheric N₂ and oxygen (O₂) (Kim and Craig, 1990). Most surface sources are depleted in ¹⁵N and ¹⁸O relative to tropospheric N₂O (e.g., Kim and Craig, 1993), and so other processes (sources or sinks) must lead to isotopic enrichment. Rahn and Wahlen (1997) use stratospheric air samples to show that the tropospheric isotope signature of N₂O can be explained by a return flux of isotopically enriched N₂O from the stratosphere, and no exotic sources of N₂O are needed. Yung and Miller (1997) point out that large isotopic fractionation can occur in the stratosphere during photolysis due to small differences in the zero point energies of the different isotopic species, and Rahn et al. (1998) have verified this latter effect with laboratory measurements. Wingen and Finlayson-Pitts (1998) failed to find evidence that reaction of CO₃ with N₂ (McElroy and Jones, 1996) is an atmospheric source of N₂O. The use of isotopes has not yet conclusively identified new sources nor constrained the N₂O budget better than other approaches, but the emerging data set of isotopic measurements, including measurements of the intra-molecular position of ¹⁵N in N₂O isotopomers (Yoshida and Toyoda, 2000) will provide better constraints in the future.

Tropospheric N₂O abundances have increased from pre-industrial values of about 270 ppb (Machida et al., 1995; Battle et al., 1996; Flückiger et al., 1999) to a globally averaged value of 314 ppb in 1998 (Prinn et al., 1990, 1998; Elkins et al., 1998) as shown in Figure 4.2. The pre-industrial source is estimated to be 10.7 TgN/yr, which implies that current anthropogenic emissions are about 5.7 TgN/yr assuming no change in the natural emissions over this period. The average rate of increase during the period 1980 to 1998 determined from surface measurements was +0.8 ± 0.2 ppb/yr (+0.25 ± 0.05 %/yr) and is in reasonable agreement with measurements of the N₂O vertical column density above Jungfraujoch Station, +0.36 ± 0.06%/yr between 1984 and 1992 (Zander et al., 1994). Large interannual variations in this trend are also observed. Thompson et al. (1994) report that the N₂O growth rate decreased from 1 ppb/yr in 1991 to 0.5 ppb/yr in 1993 and suggest that decreased use of nitrogen-containing fertiliser and lower temperatures in the Northern Hemisphere may have been in part responsible for lower biogenic soil emissions. Schauffler and Daniel (1994) suggest that the N₂O trend was affected by stratospheric circulation changes induced by massive increase in stratospheric aerosols following the eruption of Mt. Pinatubo. Since 1993, the N₂O increase has returned to rates closer to those observed during the 1980s.

The feedback of N₂O on its own lifetime (Prather, 1998) has been examined for this assessment with additional studies from established 2-D stratospheric chemical models. All models give similar results, see Table 4.5. The global mean atmospheric lifetime of N₂O decreases about 0.5% for every 10% increase in N₂O (s = -0.05). This shift is small but systematic, and it is included in Table 4.1a as a shorter perturbation lifetime for N₂O, 114 years instead of 120 years. For N₂O (unlike for CH₄) the time to mix the gas into the middle stratosphere where it is destroyed, about 3 years, causes a separation between PT (about 114 years) and the e-fold of the long-lived mode (about 110 years).

Table 4.4: Estimates of the global nitrous oxide budget (in TgN/yr) from different sources compared with the values adopted for this report (TAR).
Reference:	Mosier et al. (1998b)		Olivier et al. (1998)		SAR	TAR
	Kroeze et al. (1999)
Base year:	1994	range	1990	range	1980s	1990s
Sources
Ocean	3.0	1-5	3.6	2.8-5.7	3
Atmosphere (NH3 oxidation)	0.6	0.3 -1.2	0.6	0.3 -1.2
Tropical soils
Wet forest	3.0	2.2 -3.7			3
Dry savannas	1.0	0.5 -2.0			1
Temperate soils
Forests	1.0	0.1 -2.0			1
Grasslands	1.0	0.5 -2.0			1
All soils			6.6	3.3 -9.9
Natural sub-total	9.6	4.6 -15.9	10.8	6.4 -16.8	9
Agricultural soils	4.2	0.6 -14.8	1.9	0.7 -4.3	3.5
Biomass burning	0.5	0.2 -1.0	0.5	0.2 -0.8	0.5
Industrial sources	1.3	0.7 -1.8	0.7	0.2 -1.1	1.3
Cattle and feedlots	2.1	0.6 -3.1	1.0	0.2 -2.0	0.4
Anthropogenic Sub-total	8.1	2.1 -20.7	4.1	1.3 -7.7	5.7	6.9a
Total sources	17.7	6.7 -36.6	14.9	7.7 -24.5	14.7b
Imbalance (trend)	3.9	3.1 -4.7			3.9	3.8
Total sinks (stratospheric)	12.3	9 -16			12.3	12.6
Implied total source	16.2				16.2	16.4
a SRES 2000 anthropogenic N2O emissions. b N.B. total sources do not equal sink + imbalance.

Table 4.5: Nitrous oxide lifetime feedback and residence time.
Models	Contributor	Lifetime	Sensitivity,	Decay Time
		LT (yr)	s=ln(LT)/ln(B)	of mode (yr)
AER 2D	Ko and Weisenstein	111	-0.062	102
GSFC 2D	Jackman	137	-0.052	127
UCI 1D	Prather	119	-0.046	110
Oslo 2D	Rognerud	97	-0.061
Lifetime (LTB) is calculated at steady-state for an N2O burden (B) corresponding to a tropospheric abundance of 330 ppb. The sensitivity coefficient (s) is calculated by increasing the N2O burden approximately 10% to B+B, calculating the new steady state atmospheric lifetime (LTB+B), and then using a finite difference approximation for s, ln(LTB+B/LTB) /ln(1+B/B). The perturbation lifetime (PT), i.e., the effective duration of an N 2 O addition, can be derived as PT = LT/(1.... s) or equivalently from the simple budget-balance equation: (B+B)/LTB+B = B/LTB + B/PT.