Isotopic Composition of the Martian Atmosphere Author(S): Alfred O

Home , Hargraves (crater)

Isotopic Composition of the Martian Atmosphere Author(s): Alfred O. Nier, Michael B. McElroy and Yuk Ling Yung Reviewed work(s): Source: Science, New Series, Vol. 194, No. 4260 (Oct. 1, 1976), pp. 68-70 Published by: American Association for the Advancement of Science Stable URL: http://www.jstor.org/stable/1742562 . Accessed: 18/01/2013 12:54

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

American Association for the Advancement of Science is collaborating with JSTOR to digitize, preserve and extend access to Science.

http://www.jstor.org

This content downloaded on Fri, 18 Jan 2013 12:54:12 PM All use subject to JSTOR Terms and Conditions in B2, being smaller than those in land at any other site could result in a prelanding observations. These results B3, were better covered by the bigger landing delay and significant additional will be reported on when completed. Ten- dunes apparent in the B2 pictures, and operational complexity. Acceptance of tatively, Viking 2 rests in a deflation that the smaller dunes in B3 might not this additional complexity was not justi- hollow. cover the ejecta from the larger craters fied, based on the B3 assessment cited H. MASURSKY there as well. Site B3 was favored by above. U.S. Geological Survey, still others independent of the dune argu- Subsequently, the B3 a and / ellipses Flagstaff, Arizona 86001 ment since it appeared to them that B3 (Fig. 8) were coalesced into one with pre- N. L. CRABILL was smoothed by uniform mantling. liminary coordinates of 48?N, 226?W. Fi- NASA Langley Research Center, In the B sites, where dunes and aeo- nal coordinates (47.89?N, 225.86?W) were Hampton, Virginia 23665 lian mantle were observed, an attempt selected on 30 August after review of the References and Notes was made to estimate thickness of cover rev 20 stereoscopic coverage. Detailed 1. H. Masurskyand N. L. Crabill,Science 193,809 based on dune spacing and dune slopes. mosaic of the landing site with the lander (1976). The minimum thickness was estimated dispersion ellipse is given in Fig. 10. 2. C. B. Farmer,personal communication. 3. G. Schaber,J. Boyce, A. Dial, M. Strobell,and as being adequate to cover small crater The crater Mie east of the landing G. Stewartprovided near-real-time crater count- ing and analysesfor these and many other loca- ejecta. Estimating impact crater ejecta point is covered with larger dunes and tions. block size at two crater diameters from deflation hollows (Fig. 11), and the polyg- 4. H. Kieffer,personal communication. 5. The Viking 2 site certificationeffort was a team the crater Mie (100 km in diameter) was onally fractured lava flows west of the effortas it was for the Viking1, except thatinfra- difficult. The estimated 10-m block size site (Fig. 12) are covered with a thin man- red observations rather than radar data were used in conjunctionwith imagesto evaluatethe was based on ejecta sizes measured in the tle of wind-blown material that partially sites. H. H. Kiefferand his associates reduced 7 site and fills the fractures. At the south end of the infraredthermal mapperdata and made them Surveyor landing Apollo 15, availablein near real time to help in site evalua- Apollo 16, and Apollo 17 high-resolution B3 region, a large channel (Fig. 13) dis- tion. C. B. Farmerand his teamsimilarly reduced The block de- sects and crosses the 600 the Mars atmosphericwater data very expedi- photographs. populations area, extending tiously. Hazard analyses were done by J. A. pend on the number of small craters be- km to the south toward the Elysium vol- Cutts, K. W. Farrell,L. Crumpler,T. Spudis, and E. Theilig. Geologicaland terrainanalyses low the resolution limit that may exca- canoes. Some of the large craters (Fig. 14) were performedby J. E. Guest, R. Greeley, vate blocks from below the wind-laid are extensively modified by wind erosion, G. Schaber, J. Boyce, J. A. Cutts, K. W. Farrell, M. H. Carr, and L. Soderblom. mantle and the number exposed by defla- possible water sapping, and dissection, so Imagedata processing proceeded through a com- tion. were deemed as that their blankets are etched out plicatedchain with the help of R. Johansen,M. Slopes acceptable ejecta Martin,and associates at the Mission and Test based on Earth analogs, except on the in negative relief. The area is thus partial- ImagingSystem Facility;G. Traverand associ- inner of craters. mantled aeolian material in the ates at the Mission and Test Photoprocessing margins ly by System; R. Ruiz and associates at the Image The infrared thermal mapper results north, where the landing site is located, ProcessingLaboratory; J. Hewittand associates at the JPL photo laboratory;and R. Tyner and for the B2 region indicated low thermal and stripped in the south. The crater W. Sowers (who madephotographic mosaics as inertia and large amounts of fines (4). No counts previously cited confirm this in- fast as these images were delivered). L. B. Garrett and others organized this image pro- data were available for the specific can- terpretation. cessing system; and D. Roos monitoredits day- didate The thermal inertia at B3 The conclusion of the search for the to-day implementation.Thanks are also due to ellipses. M. J. Alazard, D. L. Anderson,G. A. Briggs, was determined to be approximately sim- Viking 2 site was the selection of the B3 D. M. Davies, J. D. Goodlette,R. Hargraves,S. L. Hess, L. Kingsland,T. Z. Martin, W. H. ilar to that at the Viking 1 site; the re- site in Utopia Planitia. The landing was Michael, H. J. Moore, II, E. C. Morris,T. A. quired noon coverage was not available successfully accomplished at 3:58:20 Mutch,F. T. Nicholson,W. J. O'Neil, T. Owen, F. D. Palluconi,J. D. Porter,R. J. Reichert,C. (4). p.m. P.D.T., earth received time, on 3 Sagan, R. W. Sjostrom, B. A. Smith, C. W. Observations showed more atmospher- September 1976. Snyder,G. A. Soffen,T. E. Thorpe,P. Toulmin III, E. D. Vogt, and the rest of the VikingFlight ic water at B2 than at B3 (2). There was a Studies are under way to compare the Team at Jet PropulsionLaboratory and the U.S. diurnal variation in water content actual conditions encountered at both GeologicalSurvey at Flagstaff.As before, J. S. greater Martin,Jr., A. T. Young,and B. G. Lee provided at B2 and it had an assumed 10? warmer the Viking 1 and Viking 2 landing sites a readyhand at the Vikingtiller. surface temperature, although no data with those expected on the basis of the 2 September1976 were available at the site. These factors were carefully weighed; and the B3 site was selected for the fol- lowing reasons. 1) Safety. It appears that B3 is ade- Isotopic Composition of the Martian Atmosphere quately mantled, muted, and filled. Site B2 may be as good, but the seeing due to Abstract. Results from the neutral mass spectrometer carried on the aeroshell of clouds and imaging quality diminishes Viking 1 show evidence for NO in the upper atmosphere of Mars and indicate that the confidence in the coverage. Site B3 ap- isotopic composition of carbon and oxygen is similar to that of Earth. Mars is en- pears more homogeneous throughout the riched in 15Nrelative to Earth by about 75 percent, a consequence of escape that area of the 99 percent ellipse. implies an initial abundance of nitrogen equivalent to a partial pressure of at least 2 2) Science. There is a small distinc- millibars. The initial abundance of oxygen present either as CO2 or H20 must be tion between the sites. The warmer tem- equivalent to an exchangeable atmospheric pressure of at least 2 bars in order to in- perature at B2 is in its favor. The water hibit escape-related enrichment of 18O. content difference was not deemed significant. The most significant scientific dis- Viking 1, whichlanded on Mars on 20- the planet's surface (1). A preliminary ac- tinction had already been realized when July 1976, included as part of its scientif- count of the results has been published the northern latitude band was selected. ic payload a mass spectrometer designed (2). The martian atmosphere consists 3) Operations. Implementation is to measure properties of the neutral at- mainly of CO2, with traces of N2, Ar, O2, straightforward at B3. The additional mosphere between about 100 and 200 km CO, and 0. The relative abundances of data analysis and acquisition required to during the descent of the spacecraft to oxygen and carbon isotopes in the mar- 68 SCIENCE, VOL. 194

This content downloaded on Fri, 18 Jan 2013 12:54:12 PM All use subject to JSTOR Terms and Conditions 12cI80

Fig. 1. Mass spectrumshowing various mass peaks (amu)as spectra are scanned. Since time also increases to the right (5 seconds per scan), the spectrum includes a distortion due to the increase in 12c ambient pressure during the time of a spectral scan. Thus, for example, the 12-amupeak due to 12Cfrom CO2is 1.4 times higher,when referredto the 44-amupeak, than it would have been had the spectrum been scanned at constant altitude. In all of the results reported -40 22 here, ion peak magnitudeswithin a particularspectrum are corrected - 2 to the values they would have had at the time of the start of the 40 32

tian atmosphere were observed to be sim- which might result in an enrichment of and N2 fragmentation patterns. After cor- ilar to values measured for the terrestrial the signal due to (13C'602)2+. (This matter rection for CO+, with carbon and oxygen atmosphere, although quantitative re- will be investigated further.) The peaks at isotopic compositions being measured sults could not be presented in the pre- 45 and 44 indicate a 1:1C/12C ratio of independently, as noted above, we de- liminary report. An updated account of 0.0115 ? 0.0003. The peaks at 13 and 12 duce a ratio 5N/14N of magnitude the analysis is given here, with emphasis give a value of 0.0115 ? 0.0004. These 0.0064 + 0.001 for the bulk atmosphere. on isotopic composition. We report re- results may be compared to the average This may be compared to the terrestrial sults for the ratios 80/160, 13C/12C,and terrestrial value of 0.0112 (4). The enrich- value of 0.00368 (4). There can be little 15N/4N. We describe also evidence for a ment in 13C for the martian atmosphere doubt that the martian atmosphere is en- tentative identification of NO and note cannot exceed 5 percent, and is most riched in 15N relative to Earth by about briefly some implications for the past probably less than 2 percent. 75 percent. The enrichment may be evolutionary history of martian volatiles. Peaks at 28 and 29 include contribu- clearly seen in all six spectra adopted for The analysis given here is based on six tions due to both N2+ and CO+, with the critical study in our investigation. spectra taken over the height range 111 latter formed mainly in the instrument by The peak at mass 30 contains a contri- to 157 km. Data at lower altitude are of dissociative ionization of CO2. For the al- bution due to (12C80O)+,whose magni- relatively limited use for present pur- titude range discussed here, approxi- tude may be readily estimated from poses, because of pressure smearing of mately one-third to one-half of the sig- known calibration parameters for the in- individual mass peaks. Mass peaks in nals at 28 and 29 may be attributed to strument, and concentrations for CO in- spectra taken at higher altitude are too N2+. The exact amounts depend on the dependently set by careful analysis of small for accurate measurement of the altitude and are determined from the the peaks at 46, 44, 28, and 12. After cor- less abundant isotopes. A typical spec- mass peaks 44, 14, and 12 and from the recting for (12Cl8O)+, we find a residual at trum for an altitude of 133 km is shown in laboratory calibrations giving the CO2 mass 30, of magnitude equal to approxi- Fig. 1. The ratio 180/160 may be inferred either from the peaks at 46 and 44, or from the peaks at 23 and 22. The peaks due to singly charged CO2, 46 and 44, indicate an isotopic ratio, 180/160, equal to 0.0020 ? 0.0001 (3). The peaks due to doubly charged CO2 suggest a ratio equal to 0.0021 + 0.0002. These results may be compared to the average terrestrial ra- E u1 tio (4) of 0.00204. It is clear that the mar- D t/ tian atmosphere cannot be enriched to Q- any appreciable extent in O80relative to cL

Earth. The enrichment cannot exceed 13 e- percent and is most probably less than < about 3 percent. 1 I Similar conclusions apply for 13C. The 0.8 ratio 13C/12Cmay be derived from a study of peaks at 45 and 44 or from peaks at 13 0.6 and 12. In principle, the ratio should be given also from peaks at 22.5 and 22. Ra- 0.4 tios derived from the 22.5 and 22 peaks are anomalous, however, both in the martian spectra and in spectra taken with two similar instruments in the laborato- EDDYDIFFUSION COEFFICIENT (cm s Fig. 2. Minimuminitial abundances of N2 to ry. The anomaly may be due to a lack of required supply enrichmentsof 15Nin the present atmosphereof magnitude50, 75, and 100 as functions of the value assumed for the mass resolution in the instrument or to a percent, eddy diffusioncoefficient in the upperatmosphere. A diffusioncoefficient of 108cm2 secel would peculiarity in the ionization process, imply a turbopauselocated at an altitude of about 120 km. 1 OCTOBER1976 69

This content downloaded on Fri, 18 Jan 2013 12:54:12 PM All use subject to JSTOR Terms and Conditions mately 50 percent of the primary peak. some past time must have contained an References and Notes The residual exhibits a scale height simi- abundance of N2 equivalent to a partial 1. A. O. Nier, W. B. Hanson, M. B. McElroy, A. lar to Seiff, N. W. Spencer, Icarus 16, 74 (1972). to N2, and is tentatively attributed pressure of at least 2 millibars. 2. A. O. Nier, W. B. Hanson, A. Seiff, M. B. NO. A similar analysis for the oxygen iso- McElroy, N. W. Spencer, R. J. Duckett, T. C. D. Knight, W. S. Cook, Science 193, 786 (1976). The enrichment in 15N may arise as a topes would imply an exceedingly large 3. Errors given throughout this report are the stan- result of preferential escape of 14N from source of oxygen (6). An enrichment of dard deviations in the averages of the approximately six values from which the averages are the upper atmosphere. Diffusive separa- less than 3 percent in 180 would require calculated in each case. For purposes of compu- tion above the will act to en- between the and a tation, it was assumed that diffusive equilibrium turbopause exchange atmosphere existed above 125 km. Before averaging, the in- rich the upper atmosphere in 14N. Es- subsurface reservoir, either CO2 or H20, dividual ratios were corrected to allow for deple- of fast ni- tion of the heavier molecules with altitude. The cape proceeds by production which must contain an abundance of averages should thus fall near the values expected trogen atoms in the exosphere, either these compounds equivalent to an atmo- in the bulk atmosphere. 4. E. U. Condon and H. Odishaw, Eds., Handbook through dissociative recombination of spheric pressure of at least 2 bars. of Physics (McGraw-Hill, New York, ed. 2, N2+ (5), or through electron impact dis- ALFRED 0. NIER 1967), chap. 2, p. 9-63. 5. M. B. McElroy, Science 175, 443 (1972). sociation of N2 (6). A relatively simple School of Physics and Astronomy, 6. Y. L. Yung, D. F. Strobel, T. Y. Kong, M. B. should suffice to define a lower McElroy, Icarus, in press. analysis University of Minnesota, 7. M. B. McElroy and Y. L. Yung, Planet. Space bound to the initial abundance of N2 (7). Minneapolis 55455 Sci., in press. 8. M. B. McElroy and J. C McConnell, J. Atm. We introduce a parameter R to in- MICHAEL B. MCELROY Sci. 28, 879 (1971). dicate the extent to which the exosphere YUK LING YUNG 9. Work at the University of Minnesota and at Harvard University was supported by NASA may be depleted in 15N owing to diffusive Center for Earth and Planetary Physics, under contracts NAS-1-9697 and NAS-1-10492, separation at lower altitudes: Harvard respectively. A.O.N. is indebted to Ward John- University, son for help in making computations. Massachusetts 02138 Cambridge, 2 September 1976 R =fc(t)lfo(t) (1) where fc(t) denotes the mixing ratio of 15N relative to 14N at the exobase in the of present atmosphere-that is, at time t, Isotopic Composition Nitrogen: with fo(t) the analogous quantity for the Implications for the Past History of Mars' Atmosphere bulk atmosphere. The parameter R is a function of the value assumed for the ed- Abstract. Models are presentedfor the past history of nitrogen on Mars based on dy diffusion coefficient in the upper atmo- Viking measurements showing that the atmosphere is enriched in 15N. The enrich- sphere, as discussed, for example, by ment is attributed to selective escape, with fast atoms formed in the exosphere by McElroy and Yung (7). The enrichment electron impact dissociation of N2 and by dissociative recombination of N2. The ini- of the present atmosphere in 15Nwith re- tial partial pressure of N2 should have been at least as large as several millibars and spect to its initial condition is given by could have been as large as 30 millibars if surface processes were to represent an important sink for atmospheric HNO2 and HNO3. 6(t) = fo(t)/fo(0) (2) and it may be readily shown that: Nitrogen accounts for about 2.5 per- It appears that the source should be cent of the present martian atmosphere dominated by a combination of reactions (t) = [.[(O)/. '(t)]'- R (3) (1). It is clear, however, that the concen- 1 and 2. The escape flux due to reaction 1 where JA(0) denotes the initial abundance tration of nitrogen on Mars must have will be a fairly sensitive function of the of N2, and a'(t) indicates the present been much higher in the past. The heavy value assumed for the partial pressure of abundance. The relation given by Eq. 3 isotope, 15N, is enriched in the present at- CO2, since N2+ ions are removed mainly assumes an initial reservoir of N2 which mosphere by about 75 percent relative to by charge transfer to CO2. Loss of nitro- is modified by subsequent escape. It as- a terrestrial standard (2). In contrast, the gen due to reaction 2 is proportional to sumes, explicitly, a passive role for the relative abundances of oxygen and car- the mixing ratio for N2, at least for small surface and ignores therefore the possi- bon isotopes on Mars appear to be simi- values of this parameter. bility that nitrogen might be incorporated lar to values observed for the earth (1, 2). According to Michels (6), reaction 1 in surface minerals, as discussed, for ex- It is hard to escape the conclusion that involves production of both N(4S) and ample, by Yung et al. (6). A surface sink Mars must have lost an appreciable N(2D) in approximately similar amounts. would lead to a smaller value for f(t) or, amount of N2 to space over the past If we assume that ion and electron tem- equivalently, it would imply a larger val- 4.5 x 109 years. The initial abundance of peratures in Mars' exosphere should ue for ,V(0) corresponding to any particu- N2 may have been large enough to pro- have values near 400?K, we may com- lar choice of 6(t). vide an atmospheric partial pressure of pute an average speed for atoms formed Values for ,/V(0) are shown in Fig. 2, at least several millibars (2). by reaction 1 of magnitude 4.96 km for several values of the enrichment fac- Escape of nitrogen from Mars may pro- sec-1. The initial velocities for 15N tor 6(t), as a function of the magnitude ceed by production of fast atoms in the should be adjusted by a factor of 0.95 in for the eddy diffusion coefficient in Mars' exosphere, by either dissociative recom- order to allow for the heavier mass of the upper atmosphere. The Viking results ap- bination of N2+ (3) less abundant isotope. The velocity re- pear to indicate a diffusion coefficient quired to escape Mars' gravitational field + e -> N + N near 108 cm2 sec-1, consistent with pre- N2+ (1) from an altitude of about 210 km above the surface is 4.68 km It vious analyses (8). We assume that f(0) or electron impact lissociation of N2 (4) planetary sec-l. should have a value equal to that for would appear that the escape rate for N + N Earth, a reasonable assumption in view e + N2--e + (2) reaction 1 should have magnitude ap- for the of to the integrated rate of results noted earlier isotopes or predissociation of N2 (5) proximately equal carbon and oxygen. It would be difficult for recombination of N2+ above the exo- to escape the conclusion that Mars at hv + N2-> N + N (3) base (7). 70 SCIENCE, VOL. 194

This content downloaded on Fri, 18 Jan 2013 12:54:12 PM All use subject to JSTOR Terms and Conditions