25 pqlm* STP
Vertical profile of NH, (solid lines) and S02 (dashed lines) concentrations according to Georgii and
Miiller (1974). (By courtesy of Tellus)
According to the observations of Georgii and Millier (1974) the following conclusions can be drawn concerning the vertical profile of NH3:
(a) The profiles suggest that NH3 sources are at the surface.
(b) The concentration decreases with increasing height in the lower two kilometers of the troposphere, then it becomes nearly constant.
(c) When the temperature is higher, concentrations are also greater, especially near the surface.
(d) The annual variation can be detected even at an altitude of 3000 m, where the S02 level is already independent of seasons. The vertical transport of NH3 is influenced by thermal structure and convection.
(e) In the lowest 2 kilometers the SOz concentration is greater in the winter, while the NH3 level has a maximum during summertime. Since a large part of the sulfur dioxide over continents is due to combustion processes (see later) this inverse relationship means that NH3 is of biological origin.
On the basis of data obtained in the atmosphere Soderlund and Svensson (1976) propose that the atmospheric ammonia burden is 0.9 x 1061 expressed in nitrogen. For the calculation of this figure the authors mentioned assumed characteristic mixing heights for different climatic regions. This height varied between 1 and 2 km.
A significant fraction of NH3 is converted into ammonium containing aerosol particles in the atmosphere. These particles are generally composed of ammonium sulfate, the formation of which will be discussed later (Subsection 3.S.3). We only note here that the concentration of NH4 in the lower troposphere is comparable to that of NH3 gas. Even, in the upper troposphere the particulate concentration may be greater than the level of gaseous NH3. For this reason Soderlund and Svensson (1976) speculate that the atmospheric NH4 burden is twice the global mass of NH3 (both expressed as nitrogen).
There is a considerable body of evidences suggesting that the major part of the NH3 is of biospheric origin. This biological source is provided by the decomposition of nitrogenous organic matters in aquatic and terrestrial ecosystems. Animal urea is also believed to be an important source of atmospheric NH3 as proposed by Healy et al. (1970). The global strength of ammonia sources was estimated originally by Robinson and Robbins (1970). They gave, however, a very high figure (1160 x 106 t yr"1) as compared to more recent and more reliable estimates18 (Galbally, 1975; Soderlund and Swensson, 1976; Dawson, 1977; Georgii and Lenhard, 1978). Thus Soderlund and Swensson's speculations led to a global emission of (113-244) x 1061 yr"Other workers give values of less than 100 x 1061 yr"1 ;e.g. Dawson calculates a source intensity of 47 x 106 t yr"
Combustion of coal was also suggested as a source of NH3-N. In the study of Soderlund and Svensson (1976) a figure of (4-12) x 106 t yr "1 is proposed for the global strength of this anthropogenic source, which is in acceptable agreement with earlier estimate of 3.5 x 106 t yr"1 given by Robinson and Robbins (1970).
A large part of the NH3, and of particulate ammonium formed from the gaseous phase, is washed out of the air by cloud and percipitation elements. The size of this sink term can be estimated on the basis of the results of chemical analyses of atmospheric precipitation (e.g. Eriksson, 1952). Soderlund and Svensson (1976) propose global values between 38 x 106 and 85 x 1061 yr"1 for wet deposition. The
18 This is due to the fact that Robinson and Robbins (1970) wanted to balance a very high unrealistic dry deposition rate(~900x 10" t yr-1).
greater amount of this quantity is deposited on land: (30-60) x 1061 yr ~1 (expressed as nitrogen). Another fraction of NH3 and ammonium particles is removed from the air by dry deposition. Many authors assume in calculating this sink strength that the dry deposition velocity19 is equal to the observed value for sulfur dioxide (~ 1 cm s"1) in the atmosphere. It is also argued (see Subsection 5.2.1) that the corresponding value for ammonium particles is much smaller. Thus Soderlund and Svensson (1976) computed by this questionable procedure dry deposition values between 19 x 106 and 53 x 106 t yr-1 for ammonia-nitrogen over the terrestrial system. For oceanic areas the corresponding figure is (6-17) x 1061 yr ""1. In contrast to this estimate Dawson (1977) speculates that the dry deposition of ammonia can be neglected. The author of this book believes that lower limit given by Soderlund and Svensson is more realistic that the upper one.
It was also proposed in the literature (e.g. Crutzen, 1974) that a non-negligible part of NH3 is converted in the troposphere into nitrogen oxides. The first step of this conversion is the reaction between ammonia and OH free radicals:
—-Biogenic sources and sinks 1 exDressed
--»Anthropogenic sources and sinks > m10°
Chemical transformation J in yr I I Atmospheric reservoirs in 10® t
—-Biogenic sources and sinks 1 exDressed
--»Anthropogenic sources and sinks > m10°
Chemical transformation J in yr I I Atmospheric reservoirs in 10® t
Atmospheric pathways of nitrogen compounds. Note: NH, released from domestic animal urea is considered biogenic and not pollution
19 Deposition velocity is defined by the ratio of the deposition (quantity deposited per unit time on a unit surface) to the concentration (see Subsection 5.2.1).
which is followed by the NH2-»NOx transformation. Crutzen argues that this process is a very important source of nitrogen oxides, while Soderlund and Svensson's work indicates values between 3 x 106 and 8 x 106 t yr~
On the right-hand side of Fig. 15 the atmospheric pathways of NH3 are plotted. Values of different terms are based on the references mentioned. In the figure, NH3 and NH4 burdens are also given. All numbers are expressed in nitrogen equivalents. In the total deposition value, a dry deposition of 10 x 10b t yr_1 is postulated. It should be noted that the accuracy of the values is not better than a factor of 2 or 3. It follows from these data that the residence time of NH3 in the atmosphere is around 5 days.
During denitrification processes in acidic soil NO is produced by chemical destruction of nitrite. Under aerobic conditions or in the air, NO is oxidized to form N0220:
If the 02 partial pressure is large, the equilibrium NO concentration is small. Thus, if the NO 2 concentration in the air is 10 fig m "3 the NO level is theoretically equal to 10"5 /jg m~3 (Junge, 1963). For this reason the majority of earlier nitrogen oxide analyses carried out in clean tropospheric air aimed to determine the N02 concentration. However, further studies, done by oxidizing NO to N02 before sampling, showed that in background air the concentration of NO is comparable to the NOz level (Lodge and Pate, 1966; Lodge et a/., 1974). This means that NO gas is released from the soil and that equilibrium is not attained in the time available.
On the basis of the results of different authors Robinson and Robbins (1970) estimated concentrations of 2 ppb (NO) and 4 ppb (N02) for continental areas between 65° N and 65° S, while for other territories they proposed 0.2 ppb and 0.5 ppb, respectively. On the other hand, Soderlund and Svensson (1976) speculate that the sum of NO and N02 concentrations in clean tropospheric air, except temperate regions, is less than 1 ppb. Over temperate regions the NOx level is about 4 ppb. The above NOx levels for remote areas were confirmed recently by Cox (1977), who measured mean concentration of 0.12 ppb NO and 0.34 ppb N02 mean concentrations on the Irish West Coast.
Soderlund and Swensson (1976) calculate for the global NOx—N burden values between 1 x 106 and 4 x 106 t by using the above tropospheric concentrations.
As mentioned earlier, a large portion of NOx is of biological origin. The global strength of this source was calculated to be (21-89) x 1061 yr"1 by balancing source and sink terms (Soderlund and Svensson, 1976). Another NOx source at the Earth's surface is provided by man's activities. This source mainly arises from coal, gasoline
20 Other possible routes of oxidation are reactions [3.21] and [3.33], and oil combustion.21 Robinson and Robbins (1970) estimate that the anthropogenic NOv—N production is 16 x 106 t yr "This figure is essentially confirmed by the more recent speculation of Soderlund and Svensson (1976) who give for this source strength a value of 19 x 106 t yr"'.
There are different opinions in the literature concerning the role of lightning in NOx production. According to earlier estimates (Junge, 1963; Georgii, 1963) the effect of this process can be neglected. However, some more recent papers (e.g. Noxon, 1976; Chameides et al„ 1977) indicate that this source is at least comparable to the NOv source intensities discussed above. It is obvious from this discrepancy that much more research is needed in the future to establish this production rate in a reliable way.
It is well documented that N02 can be transformed in the air to particulate nitrate. For this reason the nitrate content of aerosol particles in clear air is around 0.1 fig m "3 (Soderlund and Svensson, 1976). This gas-to-particle conversion process is initiated either by reaction [3.24] or by the gas phase hydrolysis of nitrogen dioxide:
The nitric acid vapour produced in this way condenses together with H20 molecules to form acid solution droplets which can be neutralized by some cations. Theory shows (Kiang ei al. 1973) that the homogeneous condensation of HNO, vapour is rather improbable under normal atmospheric conditions. This means that the condensation takes place mainly on existing aerosol particles which serve as nuclei for the phase transition.
An interesting version of this process is the interaction of nitric acid vapour and sea salt particles (Cadle, 1973). These particles are composed mainly of sodium chloride, so that sodium nitrate is formed as a result with the liberation of gaseous hydrochloric acid. It is very probable that the high nitrate content of seaside atmospheres may be explained by this interaction (Junge, 1963).
It is reasonable to expect that nitric acid vapour should react with atmospheric NH3 to form an aerosol of ammonium nitrate, NH4N03. However. Stelson et a*. (1979) calculate from the available thermodynamic data that solid NH4N03 should be in equilibrium with NH3 and HN03 such that, at 20 C, 11 fig m~3 of an equimolar mixture of the two gases would remain in the atmosphere. At 30 C the figure is 38 fig m~3. Since these figures are of the order of magnitude of concentrations in only slightly polluted air, NH4N03 can form during cold nights and decompose during the day, and it can be collected on filters and lost when the filter samples are brought into a warm laboratory. This finding raises questions about the adequacy of our understanding of atmospheric particulate nitrate.
21 It should be mentioned here that, due to this pollution source, the N02 level in polluted cities can reach 100 ppb and under these conditions nitrogen dioxide can stimulate a chain reaction by photolysis (see [3.32]) leading to smog formation (e.g. Cadle, 1966; Haagen-Smit and Wayne 1976).
It should be mentioned that N02 can take part in many other chemical conversions leading to the formation of particulate matter. However, these reactions were studied mainly in polluted air (e.g. Cadle, 1966 and 1973). Thus, application of the results obtained to clean air would be questionable. Nevertheless, it is possible thai under background conditions terpenoid hydrocarbons (see Subsection 3.3.3.) play an important part in these transformations. Furthermore, as we shall see in the next section, the presence of NOz can also be important in the oxidation of sulfur dioxide.
On the basis of the results of nitrate measurements in the atmosphere Robinson and Robbins( 1970) calculated that the total atmospheric nitrate mass is 0.9 x 106 t, which is comparable to the nitrate-niirogen burden (C.5 x 1061) given by Sóderlund and Svensson (1976).
It is believed that a large proportion of both NOv and nitrate-containing particles is removed from the troposphere by wet deposition. Sóderlund and Svensson (1976) propose a global wet deposition between 18xl06 t yr-1 and 46xl06 t yr"1 (expressed as nitrogen) on the basis of nitrate concentration analyses in precipitation. The corresponding value used by Robinson and Robbins (1970) is 70 x 10" t which arose from a wet deposition calculation of Eriksson (1952).
Another possible sink mechanism is the dry deposition. Robinson and Robbins (1970 assumed in their study a value of 1 cm s" 1 for N02 dry deposition velocity. This value was based on experimental data obtained over alfalfa and oats by Tingey (see Robinson and Robbins (1970)). Using this figure they calculated a very high N02 dry deposition. In contrast, Sóderlund and Svensson (1976) applied a dry deposition velocity of 0.3-0.8 cms"1 for NOA. Taking into account their model concentrations mentioned above, their calculation resulted in a dry deposition term of (25 70) x 106 t yr"'. Although the above deposition velocity range is in good agreement with the result of laboratory studies made by Judeikis and Wren (1978) with selected soil and cement surface, the laboratory model experiments of Bótger (1978) indicate an order of magnitude smaller N02 deposition velocities than the above values22 (see Chapter 5). Thus, the deposition values given by Sóderlund and Svensson would have to be decreased by a factor of ten. Furthermore, the dry deposition of nitrate particles was found to be insignificant due to the small nitrate concentration in the air and to the small deposition velocity of aerosol particles. The Swedish workers mentioned give a nitrate-nitrogen dry deposition range of 0.3 x 106-3 x 10" t yr"1.
On the basis of the foregoing discussion a schematic atmospheric NOA—N and N03—N cycle has been constructed as shown in Fig. 15. It has been postulated that the sum of dry deposition of gases (4 x 106 t yr"1) and particles (2 x 106 t yr"1) is only 6 x 1061 yr "For the wet deposition a global value of 34 x 1061 yr-1 has been accepted, the midpoint of the range proposed by Sóderlund and Svensson (1976). The strength of biological sources has been calculated by balancing the source and
22 According to Botger's experiments the dry deposition velocity of NO is even smaller.
sink terms23. One can see that biological source intensity received is equal to the NO, production rate by pollution sources. It should be mentioned that our figure for global natural source intensity essentially agrees with the emission rate calculated by Galbally (1975) by using the 50 % concentration difference between altitudes of 1 m and 2 km. In Fig. 15 no value for the NO, formation by lightning is given due to the uncertainties mentioned above. This does not exclude the possibility, however, that this term might be signilicant in the atmospheric pathways of nitrogen oxides. It goes without saying that the numbers represented have a considerable uncertainty which is not less than a factor of 2-3. It follows from the data given that the residence time of nitrogen oxide molecules in the atmosphere is about 9 days.
Finally, for the sake of completeness, the cycle of N20 is also plotted in Fig. 15. The values of the budget terms in this case are also very uncertain. However, the global N20 burden is rather well established by atmospheric measurements (see Subsection 3.5.2). We have to emphasize here that the quantity of NO, molecules of surface origin reaching the stratosphere is relatively small as compared to the mass of nitrous oxide, owing to the relative atmospheric concentrations of these species.
Was this article helpful?