011023-1-T THE UNIVERSITY OF MICHIGAN COLLEGE OF ENGINEERING High Altitude Engineering Laboratory Departments of Aerospace Engineering Atmospheric and Oceanic Science Final Report SATELLITE MEASURE MEN' O'P STRATOSPHERIC POLLUTANTS AND MINOR CONSTITUEN.TS' BYY SbLA.R OCCULTATION: *. ' " f *.;by* A -PRELIMINAR. Y REPOR T S.,' D R.. Drayson F. L.::B'artman W. R. Kuhn R. Tallamraju ORA Project 011023 under contract with: NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION GRANT NG-10-72 administered through OFFICE OF RESEARCH ADMINISTRATION ANN ARBOR November 1972

%!\A -(C:~ 52 -

Abstract The feasibility of determining concentrations of pollutants and minor constituents from satellite measurement of solar occultation absorption in the infrared is examined. The molecules considered include water vapor, 03, OH, CO2, CO, CH4, N20, NO, NO2, N205, HNO3, and also aerosols. A survey of experimentally and theoretically determined concentrations and of infrared spectral properties is made for each constituents. Atmospheric applications and estimates of stratospheric absorption are included. iii

TABLE OF CONTENTS Page ABSTRACT 1. INTRODUCTION 2. MOLECULES CONTAINING HYDROGEN AND OXYGEN (03, H20 OH) 2. 1 Ozone 2. 1. 1 Distribution 2. 1.2 Spectral properties of the ozone molecule 2.1.3 Results 2.2 Water Vapor 2.2. 1 Distribution 2. 2. 2 Infrared Spectral Properties 2.2. 3 Preliminary Results for the Infrared Water Vapor Bands 2. 3 Hydroxyl Radical 2. 3. 1 Distribution 2. 3. 2 Spectral Properties 2. 3. 3 Preliminary Results References 3. MOLECULES CONTAINING CARBON 3. 1 Carbon Dioxide 3. 1. 1 Atmospheric Distribution 3. 1.2 Infrared Spectral Properties 3. 1. 3 Atmospheric Application 1 4 4 4 13 15 20 20 26 29 33 33 36 37 39 44 44 44 47 50 v

TABLE OF CONTENTS (cont.) Page 3. 2 Carbon Monoxide 52 3. 2. 1 Atmospheric Distribution 52 3. 2. 2 Infrared Spectral Properties 55 3.2. 3 Atmospheric Application 58 3. 3 Methane 59 3. 3. 1 Atmospheric Distribution 59 3. 3. 2 Infrared Spectral Properties 61 3. 3. 3 Atmospheric Application 62 References 63 4. MOLECULES CONTAINING NITROGEN 71 4.1 Nitrous Oxide (N20) 71 4. 1.1. Atmospheric Distribution 71 4. 1. 2 Infrared Spectral Properties 75 4. 1. 3 Atmospheric Application 78 4. 2 Odd Nitrogen Oxides 79 4.2. 1 Atmospheric Distribution 79 4.2.2 Infrared Spectral Properties 82 4. 3 Nitric Acid Vapor (HNO3) 83 4. 3. 1 Atmospheric Distribution 83 4. 3. 2 Spectral Properties 85 4. 3. 3 Atmospheric Application 86 References 89 vi

TABLE OF CONTENTS (cont. ) p ~a, 5. ATMOSPHERIC AEROSOLS 5. 1 Introduction 5. 2 Survey of Experimental Data on Atmospheric Aerosols 5.2. 1 Sources of Atmospheric Aerosols 5.2.2 Number Densities 5.2. 3 Composition of Stratospheric Aerosols 5.2.4 Size Distribution 5.2.5 Optical Characteristics of Aerosols 5. 3 Survey of Aerosol Remote Sensing Techniques 5. 3. 1 Sunphotometer 5. 3. 2 Laser Radar 5. 3. 3 Twilight Scattering 5. 3.4 Forward Scattered Sunlight 5.4 Remote Sensing of Aerosols from Satellites 5. 4. 1 Aerosol Monitoring by Satellite Horizon Scanning 5.4.2 Identification of Aerosol Consituents Using Emitted Radiation from the Earth's Atmosphere 91 91 91 92 95 98 99 105 122 122 122 123 123 123 124 124 128 129 5. 4. 3 Discussion of Satellite Measurements of Atmospheric Aerosols References vii

TABLE OF CONTENTS (cont.) Page 6. DISCUSSION AND FUTURE PROGRAM 132 6. 1 Summary and Conclusions 132 6.2 Future Program 134 References 135 APPENDIX 136 viii

Chapter 1. Introduction The atmosphere of the earth is a complex medium in which many physical and chemical processes are continually modifying its composition, dynamics, temperature etc. Most of this occurs naturally in the atmosphere but man's activities are also causing change both on a local and a global scale, While there is still a tendency to study the different phenomena on an individual basis, it is apparent that a true picture of the atmosphere can evolve only by considering the interaction of all of the individual components. The minor atmospheric constituents are of great importance. For instance, ozone in the stratosphere prevents lethal solar radiation from reaching the surface. Most of these minor constituents are involved in chemical and photochemical reactions which ultimately effect the distribution of all of them, but are also modified by atmospheric transport phenomena, horizontal and vertical, large scale and eddies of all scales. In the troposphere mixing is good and measurements are comparitively easy to make. In the stratosphere the atmosphere is much more stable so that eddy mixing processes are much less efficient. Much more dissociating radiation can reach these levels so that the chemical processes are different from those in the troposphere. Many of them are not completely understood and measurements of their concentrations are poor or completely missing. Man is modifying his natural environment by polluting it with gasses, dust and aerosols. Many do not reach stratospheric levels, but some do diffuse upwards and others are deposited in the stratosphere by aircraft. The fate of these pollutants is a matter of speculation and concern, particularly as it may modify the climate of the earth. 1

Thus there is a need for the global measurements of the distribution of minor constituents in the stratosphere as an aid to understanding the natural processes in the atmosphere, gauging the possible effect of pollutants and the actual monitoring of pollutants. A satellite would provide an ideal platform provided a suitable experiment could be devised. The purpose of this study is to investigate the feasibility of one method, solar occultation. This report contains the results of the first six months of an investigation into the feasibility of determining concentrations of minor constituents in the stratosphere by satellite measurement of the attenuation of solar radiation along a tangent path through the atmosphere. The aim of the study is to answer such questions as: What constituents may be measured by this technique? How accurately can the concentrations be measured? What is the altitude range? What vertical resolution can be attained? How does the method compare with other techniques, particularly satellite methods? What inversion techniques can be employed? Before these questions can be adequately answered we must know what range of concentrations we can reasonably expect in the stratosphere. A literature search of measurements and theoretical calculations of stratospheric distributions was made for many of the minor constituents. In addition, for each of them a critical evaluation was made of our knowledge of the spectral properties of each molecule. This study was confined to the infrared region of the spectrum as oxygen and ozone may be e expected to absorb most of the radiation in the ultraviolet region. Finally, the findings on the distributions 2

and spectral properties have been combined to calculate the approximate absorption at various tangent heights, using a procedure described in the Appendix. Dust and aerosols are considered in a separate chapter. It should be stressed that this document is not a final report on a completed investigation. It represents an interim summary of progress. It was prepared in the fall of 1972 so that any references that appeared or became available to the authors later than September 1972 are not included. 3

Chapter 2. Molecules Containing Hydrogen and Oxygen (03, H20, OH) 2. 1 Ozone 2. 1. 1 Distribution Ozone, although a minor constituent, plays a significant role in the meteorological processes of the earth's atmosphere as well as preventing solar ultra violet radiation from reaching the earth's surface and terminating biological processes. Were it not for the small amount of ozone in the stratosphere, there would be no temperature inversion near 10 km elevation and one can imagine that circulation systems, cloud coverage etc. would be much different from what we find today. Thus it is important to monitor atmospheric ozone in order to detect any changes - man induced or otherwise - in its amount and distribution. The primary reactions leading to the formation and destruction of ozone in the stratosphere have been known for many years (see, e. g. Craig, (1965)). Ozone is produced in a three body collision including atomic and molecular oxygen, and is primarily destroyed by photo dissociation. These processes produce a maximum ozone concentration at a height of about 25 km (5 x 1012/cm ). At 50 km the concentration has decreased to approximately 10 3 5 x 10 /cm. Recently, other constituents have been found to influence atmospheric ozone, such as water vapor (Hunt, 1966) and nitrogen oxides (Johnston, 1971) and while inclusion of these molecules may alter significantly the total ozone amount, the general profile is essentially the same, i. e., a lower stratospheric maximum, with much smaller amounts in both the troposphere and mesosphere. In this section we briefly review the observations of atmospheric ozone, with particular emphasis in the height range of approximately 10 to 50 km. 4

Total atmospheric ozone as a function of season is shown in Fig.2. 1. 1 and is taken from Dopplick (1970). This distribution was compiled from monthly mean Umkehr measurements (Ozone Data for the World [Meteorological Service of Canada 1960-1968 1 ) and represents 2-9 year averages for longitudes of approximately 75W. Total ozone amount is a maximum at high latitudes in late winter and early spring, reaching values of about 0. 45 and 0. 35 atm cm in the northern and southern hemispheres respectively. These maxima are produced primarily by an increase in ozone in the low stratosphere ( Dutsch, 1970). The poleward ozone gradient is still present in the summertime, but not so much so as in the-winter. The lesser amounts in tropical regions (0. 25-0. 28 atm cm) are due to the "thicker" equatorial troposphere and associated circulation system. Note that there is more ozone in the high latitudes of the northern hemisphere than in the southern hemisphere. The maximum is also reached somewhat later in the southern hemisphere with the high values extending longer into summer (see e. g. Willett, 1968). Significant longitudinal variations in total ozone have also been detected from remote soundings of Nimbus 3 data (Prabhakara, Salmonson, Conrath, Steranda, and Allison, 1971). During June and July of 1969, ozone minima were detected over northeast India (0. 25 atm cm) and North Africa (0. 27 atm cm) and are thought to be associated with upper air high pressure systems causing a net divergence of ozone from the outward flow. Regions of high ozone amount (a' 0. 3 atm cm) are located over Southeast Asia and Africa and are thought to be associated with the easterly jet stream. In addition to the large seasonal variation in total ozone, harmonic analysis of mean monthly ozone data (Shah, 1967) revealed a 26 month quasibiennial oscillation. This oscillation has been confirmed for both the Northern 5

/I, \ I 70 I 60 - 450 350 50 0 -400 t 350 20 _-300 I _ I0w - " 2575 0- 25- \ 265 - 2753 ~30 —^~~~ 265 o2 40 \\ /3 5 0 / /' 50- "' '3 601 350 -300 70 30 302 J F M A M J J A S 0 N D Fig. 2. 1. 1. Total atmospheric ozone (x103 atm-cm) as a and season. From Dopplick (1972). I' 1% 14 function of latitude E w D ''^^ <e/ - -20 ~ 1 / - I0.,. I 200 300 400 Krueger(1971). 0 / Fig. 2.1.2 Standard ozone profiles for lowest and highest total ozone (atm-cm) 6

and Southern hemispheres, with Northern hemisphere data showing the oscillation extending from equatorial to polar latitudes. The oscillation follows a meridional wave pattern with a maximum at the equator and a wavelength of about 30 degrees latitude. There is no systematic progression in phase with latitude, there being a simultaneous occurrence of wave crests at temperate and polar latitudes There have been observed no significant longitudinal variations in amplitude, although there are differences in phase. Typical vertical profiles of atmospheric ozone are shown in Fig. 2. 1. 2 and are taken from the work of Mateer, Heath, and Krueger (1971). These ozone profiles were developed from data from direct soundings at the Canal Zone (Hering and Borden, 1965), Boulder, Colorado (Dutsch, 1966), and Resolute, Canada (advance of publication). Large variations in ozone amount are possible; at high latitudes values may reach 0. 650 atm cm while smaller amounts are found in the lower latitudes with lowest values reaching 0. 180 atm cm. Note also that the peak in the distribution increases in elevation with decreasing latitude. The double maximum in the mid latitude profile with maximum ozone is probably due to an intrusion of ozone rich polar air into the mid latitude troposphere (see following discussion). A number of latitudinal vertical ozone distributions have been constructed (Ramanathan and Kulkarni, 1960; Dopplick, 1970; Dutsch, 1970) and Figs. 2. 1. 3 and 2. 1. 4 show the results of Dopplick and Dutsch respectively. Note that their results (Fig. 2.1. 3 a, c and 2.1. 4) are qualitatively quite similar. Dopplick's global seasonal northern hemisphere results were from data as described earlier. Southern hemisphere profiles are reflected from corresponding Northern hemisphere latitudes and adjusted to the correct Southern hemisphere total ozone amounts. Dutsch implies his profiles are constructed from both the Umkehr method and direct soundings, and are for the Northern hemisphere only. 7

J ** ^ / / / I.3 _ 0_ / /\ \ 6 ^ " " " ^ / \ *2<~~~/ \1 _ /\ I, i I i ~ I. I,,,, 0 n70 6C 5c 4C 30 20 10 0 IG 20 30 40 LATITUDE Fig. 2.1. 3a Mean ozone (/g/gm for December-February. F 15 /,/, --- —--- I 0 2 ^^ ^. — ITO IE 100 / v) 150 N ~.- 200 - 300 _400 500.05 - -- 700 850 — I —1__I____ 1000 60 70 80WS From Dopplick (1972). MAR- MAY,o 30 / - 50 TO 10 C 150 ~v _ LATITUOC Fig. 2.1. 3b Mean ozone (mm/gm) for March-May. From Dopplick (1970). 8

E 3 - 2 E A cc:3 cn cn LU a: Q. Fig. 2.1. 3c Mean ozone (ug/gm) for June-August. From Dopplick (1972). 30 27 24 21 r 15 12 9 6 3 IF I) r \t' ' //'\ I }' /'foZONE ({/9) SEP - NOV 70 TO.5 - 00 _ _ t'^. ~ os _ _. 400 \ \ I -700 I % - - ^, 850 0 w A^ LA A E w a ~ (A O a," O wru w u 30 O 10 0 10 20 30 40 50 60 70 IWS LATITUDE Fig. 2.1. 3d Mean ozone (mg/gm) for September-November. From Dopplick (1970). 9

Morch / April Septemer / October -40 km 50 10 Fig. 2.1.4a Latitudinal cross section of vertical ozone distribution; early spring compared with early fall, northern hemisphere data. From Dutsch (1970). January / February July /August mb 2 3 5 10 20 30 50 100 200 300 500 6 ((r(^= ---==~?5 6 5 t 4 ---.~ t i-. —~01' ^^! 0 9 qt t 9r 9, trm 45 2 3- 40 5 35 30 20, 25 30 i250_ _20,o - 100 15 200 1300 lSOO S 70 60 50 40 30 20 0 0 10 20 30 40 50 60 70 Fig. 2.1. 4b Latitudinal cross section of the vertical distribution of ozone to air (number) mixing ratio; late winter compared with late summer, northern hemisphere data. From Dutsch (1970). 10

In the very low stratosphere, (see Fig. 2. 1.4a), concentrations near the equator are less than near the poles, while in the mid stratosphere the equatorial regions have the larger concentrations of ozone, (e. g., Randhawa, 1971 has found an order of magnitude more ozone in the upper stratosphere near the equator than at the poles) with the maximum gradient occurring in the fall. In mid latitudes, the most rapid increase in concentration occurs near the ozone maximum, and is most prominent in February. The maximum increase in the low stratosphere occurs 1 to 2 months later, while the tropospheric maximum is not reached until summer (Dutsch, 1970). The diurnal variation of ozone appears to be significant only in the mid and upper mesosphere. Hilsenrath (1971) has measured ozone by a chemiluminescent self pumping parachute sonde released from a rocket over Wallops Island in March, and he finds above 60 km the predawn ozone concentration is twice as large as is measured after sunrise. Similarly, Randhawa (1971) employed a rocket-borne. ozonesonde and found a diurnal variation at 50 km of nearly a factor of 3 at the Panama Canal Zone (9~ 20'N) in November. It is well known that at times the ozone profile is highly structured (see, e. g., Breiland 1967; Danielson, 1968), and a well developed secondary ozone maximum appears (see Fig. 2. 1. 5). This structure is thought to occur when high latitude stratospheric air is transported southward and into the low latitude troposphere. This ozone rich stratospheric air produces a lower elevation secondary ozone maximum; one also finds as would be expected, a secondary tropopause produced by the intrusion of the high latitude polar air. The smaller scale structure can be explained in a similar manner. Indeed Breiland has demonstrated that the gradient of ozone partial pressure is related to the vertical gradient of potential temperature, i. e., an increase in thermal stability leads to an increase in the vertical gradient of the partial pressure of ozone. 11

0o 0 0 O 100 200(MB) Fig. 2.1. 5 Vertical distribution of of secondary maxima. 50 -- I I 40 -230 - II w 20 I atmospheric ozone (pmb) showing development From Breiland (1967). OZONE CONCENTRATION (ppm) Fig. 2.1. 6 Model ozone profiles used in estimates of stratospheric tangent mass paths. 12

Strong seasonal transitions in wind and temperature in the upper atmosphere do not seem to influence very much ozone concentration up to 50 km. Hilsenrath (1971) measured ozone mixing ratios at Pt. Barrow, Alaska during a winter to summer transition on January 30th and January 11th in 1969 for which the temperature on the 30th at 60 km was 50~K lower than on the 11th, while the opposite was true at 40 km. The mixing ratios were essentially the same on both days between 30 and 50 km. Above 50 km, the ozone concentr_. tion increased by almost a factor of four. The four representative ozone profiles to be used in this study are shown in Fig. 2. 1. 6. Profile I corresponds to a total ozone amount of 0. 650 atm cm, and is representative of the maximum ozone amount one might find at high latitudes (from Mateer et. at., 1971). Profile II is from the same authors, and represents a high latitude minimum (0. 170 atm cm), while Profile III illustrates large ozone concentrations above the ozone maximum (. 300 atm cm). Profile IV corresponds to a typical mid-latitude distribution. Note that for any height, the extreme variations in the volumetric mixing ratios are about an order of magnitude. 2. 1. 2 Spectral properties of the ozone molecule (ir region) Ozone is a triatomic molecule belonging to the symmetry group C2v. The apex angle is 116~45', and the bond lengths are approximately 1. 26A in the ground state. The dipole moment has been determined from an analysis of the microwave spectrum, and is 0. 150 debye. The fundamental vibrational modes have been identified, and occur at 1103 cm1 (v), 701 cm (v2), and -1 1042 cm (v3). Band strengths of the fundamentals as well as some of the overtone and combination bands have been measured by McCaa and Shaw (1967), and their results are reproduced in Table 2. 1. 1 13

-1 -2 -1 Band (cm ) Strength (cm atm ) 1042(v3) 350. 1103(v1) 10.4 2110 32. 701(v2) 18. 1728 1.35 1792.54 2779.66 3042 3.0 3181.33 Table 2. 1. 1: Band strengths for selected infrared bands of the ozone molecule. From McCaa and Shaw (1967) Note that the v3 band is at least an order of magnitude stronger than the other bands which have been measured. The v3 band strength as determined by McCaa and Shaw is in reasonably good agreement with the earlier measurements of Walshaw (1954, 1957) whose data, if extrapolated to zero mass path, yield -1 -2 - 1 -2 a value of 360 atm cm, while his empirical formula gives 376 atm cm To the authors' knowledge, there are no other measurements for the weaker bands which can be compared with those from McCaa and Shaw. Measurements are made difficult because of the highly reactive nature of ozone, and because large paths are required due to the small band strengths. Theoretical studies of line strengths and positions of the infrared bands of ozone have dealt almost solely with the 9 micron region, primarily because of the importance of the v3 band to atmospheric radiation calculations. Early theoretical studies included those of Kaplan, Migeotte, and Neven (1956), who evaluated line strengths and positions for the spectral interval from 1052. 66 to 1057. 67 cm and compared their results with high resolution measurements to 1057. 67 cm and compared their results with high resolution measurements 14

of the solar spectrum; good agreement was found up to rotational level seventeen. The more recent evaluation of Clough and Kneizys (1965) extended over the spectral region from 995 to 1070 cm, and included the Coriolis effect as well as a second order distortion term. Thesy indicate that their theoretical absorption contours are in satisfactory agreement with experimental absorption spectra taken over the interval from 1000 to 1150 cm; the spectral resolution -1 was 0. 08 and 0. 06 cm for the v3 and vl bands respectively. While these studies may be all right for energy budget calculations, it is doubtful if they are adequate for such applications as remote sensing. Indeed, Drayson and Young (1967) and Goldman et. al. (1970) have found large discrepancies between absorption contours calculated from theoretical parameters and observational results. Also, there has been little study of the half-width variation within the band; estimates for the mean half-width have been made by Kaplan (1959) and Walshaw (1955), who give values of 0. 089 and 0. 076 cm1 (STP) respectively. Recently, Aida (1971) presented a paper at the Symposium on Radiation held in Sendai, Japan, in which he calculated band parameters for the 9 micron region and found good agreement with experimental results. Much of the improvement which he found was apparently due to his inclusion of some of the weaker bands in this region. Any final conclusion concerning his results should await publication. 2. 1. 3 Results The tangent path ozone amounts (atm cm) for profiles I through IV (see Fig. 2. 1. 6) are shown in Fig. 2. 1. 7. The method for computing these profiles is given in Appendix A. Profile III representing the maximum ozone 15

30 Z 5 20 I n 10 o0 I I [...I. I I 1, 11., I _ I i, 0.1 I 10 TANGENT MASS PATH (ATM-CM) Fig. 2.1. 7 Ozone tangent mass paths (atm-cm) based on model ozone profiles from Fig. 2.1. 6. 100 100

amounts found in the mid and upper stratosphere and profile II for the high latitude minimum differ by approximately a factor of six for any given height; these two profiles are used in the following discussion. The absorption contours given by McCaa and Shaw (1967) for the nine bands 701, 1042, 1103, 1728, 1792, 2110, 2779, 3042, and 3181 cm-1 have been used to estimate the absorption for profiles II and III in Fig. 2. 1. 7 and are summarized in Table 2.1.2. For each case, the maximum percent absorption shown on the contour has been used. One should bear in mind that these contours correspond to room temperature which differs significantly from the tangent path temperature. This varies from about 220 K near the tropopause to about 270 K at the stratopause. The temperature effect on the absorption becomes quite significant especially for transitions which do not involve the ground state. In addition, most of the stratospheric pressures and mass paths do not correspond to pressures and mass paths used by McCaa and Shaw; thus only a limited number of comparisons are possible. It is hoped that in future work we shall be able to carry out line by line calculations and produce high resolution absorption profiles which correspond to stratospheric pressures and mass paths as well as temperatures. The vl and v2 bands show similar absorptions as one might expect because of their similar strengths (see Table 2. 1. 1). At 16 km, e. g., both give a maximum absorption greater than 30-35% for Profile II. For Profile III, the maximum absorption has not fallen to the above value until a height of 24 km is reached. At this elevation, Profile II gives an absorption of 15% for the vl band and somewhat larger for the v2 band. Only for Profile III does one get significant absorption from the 1728 and 1792 cm bands in the lower stratosphere. Profile II yields absorp 17

tions of only 10 and 5% respectively for a height of 24 km. Also, even for the largest tangent paths represented by Profile III, maximum absorption from the 2779 and 3181 cm bands is only 4 and 9% respectively at the 24 km height. Maximum absorption for the 3042 cm-l band is somewhat larger, being 18% at 16 km and 10o at 24 km for Profile II; Profile III yields an absorption of 25%o at 24 km. 18

Significant absorption is found for the 2110 cm band well above the 24 km height, where the maximum absorption is 40 and 75% for Profiles II and III respectively; at 16 km, the absorption from II has increased to about 70%. As one would expect, the greatest absorption is from the strong 3band. The band center is effectively saturated at heights of 24 and 16 km for Profiles III and II respectively. Thus this band would produce measurable absorption well into the mid-stratosphere. Even for Profile II, the absorption is still 88% at the 24 km level. Band (cm ) 1042(v,) 1103(x)2110 701CiI)1728 1792 2779 3042 3181 Height (km) 16 II 100 >30-35 70 >30-35 18 88 15 40 >15 10 5 10 100 30-35 75 30-35 4 25 9 24II 24III Table 2. 1. 2: Maximum absorption (%) for Profiles II and III for selected ozone bands. 19

2. 2 Water Vapor 2. 2. 1 Distribution The significance of atmospheric water vapor to biological and atmospheric processes is well known. Water vapor is as important in sustaining biological activity as is oxygen. In addition water vapor is the most significant gas in maintaining the tropospheric temperature structure, not only through its gaseous absorption and emission properties of planetary and solar radiation but also in the liquid and solid state as clouds whereby much of the planetary radiation is prevented from escaping the earth, and also in shielding the earth from solar radiation. Water vapor is not as important in the stratosphere in maintaining the thermal structure although it has been determined that reactions of water vapor with ozone reduce the ozone concentration which, as indicated previously, shields the earth from harmful ultraviolet radiations. In addition, high flying aircraft may induce cloud formation which could alter the radiation balance of certain regions over the earth. Remote soundings of stratospheric water vapor would provide the continuous coverage necessary to monitor such possible effects. Typically, water vapor concentrations are found to decrease exponentially in the troposphere. The source of most of this vapor is the oceans. The tropopause provides a barrier to vertical motions and most of the vapor is "trapped" in the troposphere, with stratospheric concentrations thought to be primarily controlled by the tropopause saturation vapor pressure. Total atmospheric water vapor averages about 2 precipitable cm with maxima (4 cm) and minima ((1 cm) over the equator and poles, respectively. Large longitudinal variations are also evident (Fig. 2. 2. 1) e. g., the yearly average water content over central South America is about three times larger than over 20

Fig. 2. 2. 1 E LII W (n 0C a. Time average of the vertically integrated values of specific humidity (precipitable water) in units of gm/cm2 for yearly data. Isoline spacing (full curves) 1 gm/cm2. From Starr, Peixoto, and McKean, (1969). i I II o-6 10'5 VOLUMETRIC MIXING RATIO Fig. 2. 2. 2 Comparisons of lower stratospheric profiles for atmospheric water vapor. 21

south central Africa. The vertical distribution of water vapor and its variation with latitude, longitude, and season are now reasonably well known at least up to the mid troposphere (see, e. g., Taljaard, Van Loon, Crutcher and Jenne, 1969; and Crutcher and Van Loon, 1970) and we shall here only discuss those observations made in the stratosphere. Early studies of stratospheric water vapor are somewhat suspect because of the high degree of instrumental sophistication required to measure such small amounts of water vapor. In addition, water vapor contamination from the instrument itself is a source of some uncertainty. Studies prior to 1961 have been carefully reviewed by Gutnick (1962)who called attention to "the widely divergent and apparently inexplicable stratospheric humidity data" e. g., the results of Japanese investigators indicated a much higher concentration (mixing ratio of 3 x 10-4 gm/gm) than those of the British group (4 x 105 gm/gm). Gutnick (1962) also gives a mean annual vertical profile to 31 km, which as pointed out by Sissenwine, Grantham, and Salmala (1968) would give noctilucent clouds over most of the earth and water vapor partial pressures exceeding the total atmospheric pressure if his mean profile is extrapolated upward. More recent observations indicate smaller stratospheric concentrations; Houghton (1963) measured the absorption of solar radiation in the 6. 3Aum region and found mixing ratios of 1. 5 x 10-6 and 5 x 10- gm/gm in the low and mid stratosphere respectively. Williamson and Houghton (1965) measured the downward planetary radiation in the 6. 3Mm band, and found a similar value for the lower stratosphere (3 x 10-6 gm/gm). Recently there have been a number of studies of stratospheric water vapor over particular regions, the most extensive being that of Mastenbrook (1968, 1971) over Washington, D. C. during the 6 year period from 1964-1969. These measurements were made from descent balloon soundings with a frost point hygrometer. The median of Mastenbrook's values is plotted in Fig. 2. 2. 2. 22

The volumetric mixing ratio varies from 1 to 4 ppm between 40 and 80 mb, with a mode of 2. 5 ppm. At higher and lower elevations the mode is also 2. 5 but the distributions are skewed toward higher concentrations which is thought due to water vapor contamination from the instrument at the higher levels, and at lower levels the increase is probably due to the close proximity to the tropopause. Three soundings were also made at Thule, Greenland (Masteabrook, 1968) in July and August 1965 and stratospheric mixing ratios were not much different from those found in lower latitudes, being about 1. 5 to 3. 5 x 10-6 gm/gm (Fig. 2. 2. 2). Measurements at Trinidad during 1964-1965 also by Mastenbrook (1968) gave the same results as for Washington, D. C. for the 50-60 mb range but were larger at lower elevations than the Washington, D. C. results (Fig. 2. 2. 2). Also plotted in Fig. 2. 2. 2 is a mid latitude yearly average humidity profile for Chico California (Sissenwine et. al. (1968)). This profile is the result of seventeen observations made from a balloon carrying an alpha radiation hygrometer. The dashed portion of the curve is the result of only one sounding. As in the case of Mastenbrook's profiles, the humidity reaches a minimum near the tropopause, but then shows a slight maximum near 25 mb with a gradual decrease upward. Thus in the vicinity of 25 mb, Sissenwine et. al. 's profile yields almost an order of magnitude more water vapor than does that of Masteabrook. Sissenwine et. al. speculate that tropospheric water vapor may pass into the stratosphere in mid latitudes when multiple tropopauses are present, and also that ice crystals from convective clouds which penetrate the stratosphere over continental mid latitude regions in summer could vaporize 23

and produce a high water vapor content. It is however, interesting that Mastenbrook does not find such large stratospheric values over any of the three sites used for his measurements. The only direct measurements of water vapor near the stratopause are those of Scholz, Ehhalt, Heidt, and Martell (1970) who used a hydrogen cooled cryocondenser carried on a rocket. Measurements were made at White Sands in September, 1968 in the height range of 44-62 km. The volumetric mixing ratio was found to be within the range of 3 to 10 ppm. An annual cycle in stratospheric water vapor has been detected by Mastenbrook (1971)1 as well as a linear increase over a 6 year period of about 1 ppm. It can be seen from Fig. 2 2. 3. that the annual cycle is present at least to 100 mb with the amplitude being greatest (50 ppm) at the 300 mb level and decreasing with decreasing pressure. At the 100 mb level the amplitude is only about 0. 5 ppm and less than 0. 2 ppm above 80 mb. There is also some indication of a phase shift; in the upper troposphere, the maximum water vapor content occurs in summer, while at 80 mb the maximum is in winter. The only latitudinal study of stratospheric water vapor is that of McKinnon and Morewood (1970) who measured solar absorption in the v' and V3 bands (2. 59,). These spectra were obtained from jet aircraft from April 1967 through. May 1968 over North and South America (70N to 40S) in the height range of 10. 7 to 18. 3 km. At all latitudes, concentrations of water vapor were found to reach a minimum about 2 km above the tropopause and then remain nearly constant to 18. 3 km. A large seasonal variation was Sissenwine et. al. (1968) point out that their data appear to show an annual cycle but harmonic analysis did not show these cycles to be statistically significant. 24

80o 180 I 20 200MBs 30 40 250 MB 60 70 80 -00MB - 90 - 00 I F M I A M J J A I S I 0 N Fig. 2. 2. 3 Seasonal variability of mixing ratio for selected pressure levels as determined from three-month running medians. From Mastenbrook (1971). 50 1 -' -- -;j,7 —7 — ' - - -.-... ' '..::-::;::,.:.''.:::::;:::::::::::::...... 40- - 30 -I. / VOLUMETRIC MIXING RATIO (ppm) Fig. 2. 2. 4 Model water mass paths. vapor profiles used in estimates of stratospheric tangent 25

detected below the tropopause, the mixing ratios being four to eight times larger in summer than in winter; stratospheric variations were only a few percent. The latitudinal variation in mixing ratio above 17. 7 km obtained during January-March is rather small, representative values being 1. 75 x 10-6 gm/gm at 65N, 1. 45 x 10-6 at 25N, 1. 6 x 10-6 at 7N, and -6 1. 25 x 106 at 30S. It is interesting to note that the minimum concentration does not occur over the equator, as one would expect if the equatorial tropopause were the only controlling region for stratospheric water vapor. Water vapor profiles which were chosen for the tangent path calculations are shown in Fig. 2. 2. 4. The extreme distributions are represented by Sissenwine et. al.'s (1968) results and those of Mastenbrook (1971). Note that the variation is slightly less than an order of magnitude. 2. 2. 2 Infrared spectral properties There have probably been more studies, both theoretical and experimental, devoted to the spectral characteristics of water vapor than any other nonlinear triatomic molecule. H20 belongs to the point group C2v and has three fundamental vibrational bands, v1 and v3, with band centers at 3657. 05 and 3755. 92 cm respectively, and the strong bending mode v2 located at 1594. 78 cm. In addition, there is the strong pure rotational band centered about 125 1 (band strength 1440 cm-2 atm The most centered about 125 cm (band strength,^ 1440 cm atm ). The most abundant isotope (99. 73%) is H2016 which has a dipole moment of 1. 87 debye. In addition to the fundamentals, there are important water vapor bands throughout the visible and infrared spectrum, and because of the quite different moments of inertia, the rotational structure of these bands is very complex. The strengths of many of these bands are given in Table 2. 2. 1. 26

Transition 1Band Center (c1) Band Center (cm ) Strength (cm2 -1 Strength (cm atm 000-100 (v ) 3757.05 11.2 000-010 (v2) 1594. 78 223. 000-001 (v3) 3755.92 167. 000-411 18394.0005 000-203 17495.003 000-401 16899. 008 000-302 16898.0008 000-321 16822. 005 000-113 15832.0005 000-311 15348.005 000-103 14319.03 000-400 14221. 003 000-301 13831. 08 000-202 13828. 0005 000-221 13653.2 000-013 12565.003 000-112 12408. 2 000-211 12151.002 000-210 12140.003 000-131 11813. 05 000-003 11032.05 000-102 10869. 01 000-201 10613.3 000-300 10600.02 000-121 10329. 05 000-220 10284.001 000-041 9834. 002 000-012 9000.008 000-121 8807. 21 000-210 8762.0003 000-130 8274.0002 000-031 8374.0008 000-002 7445.03 000-101 7250 40.0 000-200 7201.40 000-021 6871.3 000-120 6775.005 000-011 5331 5.9 000-110 5235. 2 000-030 4667.008 Table 2. 2. 1: Band intensities for visible and infrared water vapor bands (from Goody, 1964; intensities given by Goody are in cm)

There is still much uncertainty regarding the strengths of many of these weaker bands; indeed, even for the strong v2 band, Penner (1959) gives a value some ten percent smaller than that reported by Goody. Note that the above listings all involve transitions with the ground state and will thus not be strongly temperature dependent. Upper state bands depend critically on temperature, and are not included in the above listing. There have been numerous low and medium resolution measurements of water vapor, notable being those of Palmer (1960) for the rotational band, and Burch, Gryvnak, Singleton, France and Williams (1962) for the fundamentals as well as the rn bands (000-011 and 000-110). Palmer examined the spectral region from 200 to 500 cm and expressed the transmission as averages over 50 cm intervals. Burch et. al. present graphs of total hand absorptance vs. either mass path or pressure, as well as representative absorption curves. The more recent high resolution work includes that of Babrov (1967) and Fridovich and Kinard (1972) (for earlier work, see Goody, 1964). Babrov analyzed 12 lines in the vl and 30 lines in the v3 band by a curve of growth method, while Fridovich and Kinard determined intensities of 7 lines in the v2 band, and compared their results with other investigators. They found, for example, the line strength at 1429. 96 cm to be 981 cm/gm, while Prinz (1968) gives 725, and Krakow and Healy (1969) givelll6 cm/gm. These results indicate that while H20 has been extensively investigated, there is still need for improvement in the band parameters; Goldman and Kyle (1968) reach the same conclusion from an investigation of the 2 7,um band for averages over intervals smaller than about 10 cm. The most detailed compilation of band parameters (line positions, strengths and half-widths) is that of Benedict (unpublished), Benedict and Calfee (1967) and Gates, Calfee, Hansen, and Benedict (1964). Calculations 28

-1 for 6100 lines extend over the spectral range from 0. 74 to 4444 cm. The bands included are the fundamentals, 2v2, 3v2, v2v3, vlv2, as well as the hot bands 021-010, 100-010, 020-010, 001-010, 030-010, 011-010, and 110-010. 17 0 18 The isotopes H2O and H20 are included along with HDO. Results are presented in a form such that strengths corresponding to different temperatures can be computed. Half-widths are based on the earlier results of Benedict and Kaplan (1964, 1959), who used Anderson's theory to calculate the widths of many pure rotational transitions for both self broadened H20, and H20 broadened by nitrogen and oxygen. They found large variations from line to line in both absolute value as well as in the ratio of self to nitrogen broadening. Also there is a good correlation between nitrogen and oxygen broadening. An intensity weighted mean half width for oxygen broadening is 0. 04155 (STP) and -1 for nitrogen broadening 0. 08767 cm. To the author's knowledge, there are no comparable studies for the visible water vapor bands. It should be obvious that we have not attempted to review all, or even most of those papers which have made significant contributions to the spectral properties of water vapor, but only those which relate directly to the present work. 2. 2. 3 Preliminary results for the infrared water vapor bands Tangent path water vapor amounts for the profiles of Fig. 2. 2. 4 are given in Fig. 2. 2. 5. The profile of Sissenwine et. al. (1968) gives the largest tangent paths above the height of 20 km, being some 40 atm cm at 20 km and decreasing to 0. 65 atm cm at the stratopause. The minimum distribution shown is that of Mastenbrook (1970), and is about a factor of six less than that of Sissenwine et. al. These two profiles, I and II, will be used in the following analysis.

TANGENT MASS PATH (ATM-CM) OH *5 W-4 10 10 50 2.o Z: LLJ I-lr 101 _ i I i I i I I I 10' 10~ 101 TANGENT MASS PATH (ATM-CM 102 ) HoO - -. - - - - v - - - --- — d Fig. 2. 2. 5 Water vapor and hydroxyl tangent mass paths (atm-cm) based on model water vapor profiles from Fig. 2. 2. 4 and hydroxyl profile from Fig. 2.3.1.

An estimate of water vapor absorption at various heights for Profiles I and II was made for the bands at 1595 (v2), 5332 (thefn bands), and the v band at 3755 cm. These bands were chosen since absorption contours are readily available from Burch, Gryvnak, Singleton, France, and Williams (1962). As was the case for the ozone profiles the maximum percent absorption from each profile was used. Unfortunately, their choice of pressure and mass path do not, in general, coincide with our tangent path pressures and mass paths, so that only a very approximate maximum absorption estimate is possible. As in the case of the ozone results, the water vapor contours also apply to room temperature, while our tangent path stratospheric temperatures are much lower. For Profile I, corresponding to maximum water vapor, the 5332 cm band gives appreciable absorption at least to 19 km, but certainly not to 38 km. For example, at 19 km, the tangent path pressure is 36. 7 mb while the path is 41 atm cm. From Burch et. al 's absorption profile, for a pressure of 36 mb, and a mass path of only 4. 12 atm cm, the maximum absorption is about 5%. On the other hand, at 38 km, the tangent mass path is about 4 atm cm, but the pressure is only 2. 7 mb, which would make the tangent path absorption much less than 5% at this elevation. A comparison for the 14 km height (pressure of 77 mb and mass path 56 atm-cm as opposed to Burch et al's contour for a pressure and mass path of 66 mb and 58 atm-cm respectively) indicates that the absorption is 27%, while at the tropopause, it is somewhat less than 55%. For Profile I, the 5332 cm-l band will have a small absorption above about 20 km; at this elevation the tangent pressure and mass path are about 36. 7 mb and 9. 5 atm cm respectively, and as indicated above, for this same pressure but 4.12 atm cm, the maximum absorption is only 5%. 31

The same reasoning can be applied to the 3700 cm1 band as was done above for the 5332 cm band, except that the 3700 cm band is stronger and will have appreciable absorption above 19 km, but certainly not extending to 38 km. For Profile II, e. g., rather than an absorption of 5% for a pressure of 36 mb and a mass path of 4. 12 atm cm, the maximum absorption is 20%. An analysis for the v2 band cannot be made above a height of 25 km, since here the tangent pressure is 18 mb, representing the smallest pressures used by Burch et. al. along with a mass path of 2. 12 atm cm; these parameters yield a maximum absorption of 12%; however the mass path associated with the 25 km level is about 33 atm cm for Profile I.and 4. 3 atm cm for Profile II. Thus there will be significant absorption for both profiles at least to 25 km. The absorption is certainly very small at the 42 km height, since here the tangent path for Profile I is 2. 12 atm cm, the same as Burch et. al., but the pressure is only 1. 6 mb; thus the maximum absorption would be much less than 12%. At 17 km, for Profile I, the maximum absorption is about 65% for a pressure of 50 mb and mass path of 45 atm cm (Burch et al's contour is for a pressure and mass path of 50 mb and 44 atm cm respectively). Profile II, at 19 km, yields an absorption of 35 % for a pressure of 45 mb and 11 atm cm while Burch et al's contour has values of 45 mb and 10. 1 atm cm. 32

2. 3 Hydroxyl Radical 2. 3. 1 Distribution The hydroxyl radical is significant in the stratosphere, primarily because of its influence on concentrations of other minor constituents. Much of the literature is concerned with carbon monoxide (primarily troposphere) and ozone (stratosphere and mesosphere), for which the hydroxyl is an important controlling mechanism, and relatively few papers are available which give explicit estimates of hydroxyl concentrations. These are theoretical or model studies; the authors are unaware of any observational results for the troposphere and mid stratosphere. Tropospheric models for which estimates of hydroxyl concentrations were calculated are those of Levy (1971); McConnell, McElroy and Wofsy (1971) and Weinstock and Niki (1971). They find tropospheric concen6> 3 trations of approximately 2 x 10 6/cm (see Fig. 2. 3. 1). Levy constructed a steady state model for large radical and formaldehyde concentrations for 6 3 which the hydroxyl concentration can be estimated at 1. 2 x 106/cm. Both McConnell et. al., and Weinstock and Niki were primarily concerned with the sources and removal mechanisms for carbon monoxide, the former showing that an important source for carbon monoxide below 30 km is the reaction, OH + CH4-HH2O + CH3, while the latter demonstrated that the major removal mechanism is CO + OH —CO2 + H. Weinstock and Niki find that the average concentration of hydroxyl in the troposphere required to maintain a carbon 6 3 monoxide balance is 2. 3 x 106 molecules/cm, with daylight values being approximately twice as large. McConnell et. al. present a vertical profile, with a hydroxyl concentration of 3 x 106/cm up to 5 km and then decreasing to about about 9 x 5/cm3at about 12 km. One should keep in mind that the to about about 9 x10 /cm at about 12 km. One should keep in mind that the 33

1-.^~ W 1 Weinstock a Niki (1971) 5 30- Levy (1971) w ---- McConnell et al (1971) ^ ^^^'4p~~~~ -- Hunt (1966) 20 o<>0< Leovy (1969) ---- Hesstvedt (1968) ^8M Anderson (1971),1_ ""'0 o Approximote Values for 0 - ^' '-.. > ~Mean OH Amount as |:"i~\ -Used in Program I*i 0. I!, _10,,,,,,1,.. 1., I1i 106 107 108 OH CONCENTRATION (MOLECULES/CM3) Fig. 2. 3.1 Comparisons of hydroxyl distributions in the troposphere and stratosphere.

hydroxyl concentration depends on the assumed ozone and water vapor concentrations. In constructing their height profile, McConnell et. al. used a typical mid latitude ozone distribution, a stratospheric water vapor concentration of 3 x 106 by volume, and tropospheric water vapor densities equal to half the dew point values. Numerous photochemical studies for an "oxygen-hydrogen atmosphere" have been made (see, e. g., Hunt (1966)) and the significance of the hydrogen-oxygen compounds in reducing the ozone amount to values smaller than given by the pure oxygen atmosphere is well known. In such a study the hydroxyl radical is an important constituent and its variation with height has been estimated by Leovy (1969), who developed an analytic model for the photochemistry to a height of 60 km (see Fig. 2. 3. 1). The earlier, more detailed study of Hunt (1966) is also given. The shaded ragion refers to uncertainty due to the different ways of treating the H202 reactions. The maximum hydroxyl concentration occurs at a height of about 40 km with a value of 7 3 2-3 x 107/cm. These results should only be viewed as very approximate since it was pointed out (see discussion, Leovy, 1969) that the reaction, CO + OH —CO2 + H, was not included in Leovy's treatment of the problem. Results for hydroxyl for Hesstvedt's (1968) oxygen-hydrogen model are also shown in Fig. 2. 3. 1. His results only extended to 45 km since he was primarily concerned with the mesosphere and lower thermosphere. His concentration is nearly two orders of magnitude larger during the daytime than at night. It should also be indicated that the calculations presented in Fig. 2. 3. 1 do not include vertical eddy transport, which is quite significant. The theoretical studies shown in Fig. 2. 3. 1 are in reasonable agreement near the stratopause with the experimental results of Anderson (1971) 35

who made rocket-borne observations of an electronic transition in the hydroxyl molecule, and computed number densities in the height range of 45 to 70 km. The shaded area indicates the range of uncertainty. There appears to be no other measurements for hydroxyl concentrations in the stratosphere. 2. 3. 2 Spectral Properties The spectral properties of the hydroxyl radical have been extensively studied primarily because of its importance to the airglow emission especially in the spectral region of 6700 to 8900 A at heights near the mesopause level (for a review, see Wallace, 1969 or Chamberlain 1961 ). The upper vibrational levels are very important to studies of this emission, and consequently much of the work has been devoted to the band constants of these excited vibrational states. These states are probably not very significant to the present study, since in the stratosphere the population densities of these levels should be given by the Boltzmann distribution and the band intensities and therefore the absorption should be small. Band constants for the lower vibrational levels have been determined from laboratory measurements by Dieke and Crosswhite (1948) and Herman and Hornbeck (1953). The former analyzed vibrational levels up to v=4 from high resolution data for the A2 - x2 bands, while the latter studied the emission spectra from the 4-0, 5-1, and 6-2 bands, and give band constants up to v= 7. This work was further extended by Wallace (1960) who obtained the parameters for vibrational levels 7, 8, and 9, and Bass and Garvin (1962) for the levels 7 through 10; in this latter study, ground state wavenumbers for rotational lines in 26 bands in the visible and near infrared are given. Recernly, Engleman (1972) has more accurately determined rotational 36

wavenumbers of the A2, - X2f (0-0) amd (1-0) bands. In the present study, we shall use the compilation of rotational line positions and relative intensities given by Chamberlain (1961). He includes the first 7 rotational levels for each of the vibrational transitions through v=9. Relative intensities have been calculated for a temperature of 225K. Intensities of a few of the hydroxyl bands are given in Table 2. 3. 1. The 0-1 vibrational transition is much larger than the overtones or upper level transitions for a Boltzmann distribution of energy states. For comparison, we have included the band strength of the electronic transition 2: - 2 T, -1 -2 -1 Transition (cm ) Band strength (cm atm 3568 (0-1) 100 6971 (0-2) 4 3400 (1-2) 3. x 10 9 326002 _ - 2T 29000. -2 -1 Table 2. 3. 1:, Band strengths (cm atm ) for selected transitions in the hydroxyl molecule. From Penner (1959) 2. 3. 3 Preliminary results Since the ground configuration of the hydroxyl radical gives a 2r state, the vibrational-rotational bands will have strong Q branches, and for the 0-1 transition, Chamberlain (1961) estimates that 0. 268 of the total band strength resides in this branch. If one assumes a band strength of -2 -1 -1 100 cm 2atm, and a spectral resolution & of 5 cm, then the absorption for a mass path u can be estimated from the parameter Su/S. From our 37

approximate hydroxyl distribution (see Fig. 2. 3. 1) the tangent paths have been calculated and are shown in Fig. 2. 2.5. Note that because the volumetric mixing ratio of OH increases, there is not a large variation in the stratospheric tangent path. If we assume a tangent path of 10 atm cm, then -4 the above parameter has a value of approximately 5 x 10. The fractional absorption can then be estimated from the weak line approximation and is about 0. 05%. In order to obtain an absorption of 10%, a resolution of 0. 03 cm would be required. Thus it appears doubtful that one could measure hydroxyl concentrations in the stratosphere utilizing the vibrational bands. Possibly above the ozone layer the electronic bands could be used; this will be investigated in future studies. 38

References Ozone Breiland, J. G., (1967), Comparison of the Vertical Distribution of Thermal Stability in the Lower Stratosphere with the Vertical Distribution of Atmospheric Ozone. J. Atmos. Sci., 24, 569-576. Clough, S. A., and F. X. Kneizys, (1965), Ozone Absorption in the 9. 0 Micron Region. AFCRL Rep. No. 65-862 or Physical Sciences Research Paper No. 170, 79.pp. Craig, R. A., (1965), The Upper Atmosphere, Meteorology and Physics. Academic Press, New York. 509 pp. Danielsen, E. F., (1968), Stratospheric-tropospheric Exchange Based on Radioactivity, Ozone, and Potential Vorticity. J. Atmos. Sci., 25, 502-518. Dopplick, T. G., (1970), Global Radiative Heating of the Earth's Atmosphere. MIT Rep. No. 24, Dept. of Meteorology, Planetary Circulations Project, 128pp. Drayson, S. R. and C. Young, (1967), Private Communication. Dutsch, H. U., (1966), Two years of Regular Ozone Soundings Over Boulder, Colorado. NCAR-TN-10, Nat. Center for Atmos. Res., Boulder. 433pp. Dutsch, H U., (1970), Atmospheric Ozone-A Short Review. J. Geophys. Res. 75, 1707-1712. Goldman, A. and T. G. Kyle, (1968), A Comparison Between Statistical Model and Line by Line Calculation With Application to the 9. 6G Ozone and the 2. 7 water vapor bands. Applied Optics, 7, 1167-1177. Hering, W. S., and T. R. Borden, Jr., (1965), Mean Distributions of Ozone Density Over North America, 1963-1964. Environmental Res. Papers. No. 162, AFCRL, Bedford, Mass. 19 pp. Hilsenrath, E., (1971), Ozone Measurements in the Mesosphere and Stratosphere During Two Significant Geophysical Events. J. Atmos. Sci., 28, 295-297. Hunt, B. G., (1966), Photochemistry of Ozone in a Moist Atmosphere. J. Geophys. Res., 71, 1385-1398. Johnston, H., (1971), Reduction of Stratospheric Ozone by Nitrogen Oxide Catalysts from Supersonic Transport Exhaust. Science, 173, 517-522. Kaplan, L. D., (1959), The Atmosphere and Sea in Motion, Rockefeller Inst. Press. p. 170. Kaplan, L. D., M. V. Migeotte, and L. Neven, (1956), 9.6 Micron Band of Telluric Ozone and its Rotational Analysis. J. Chem. Phys., 24, 1183-1186. 39

Mateer, C. L., D. F. Heath, and A. J. Krueger, (1971), Estimation of Total Ozone from Satellite Measurements of Back Scattered Ultraviolet Earth Radiance. J. Atmos. Sci., 28, 1307-1311. McCaa, D. J., and J. H. Shaw, (1967), The Infrared Absorption Bands of Ozone Ohio State Univ., prepared for AFCRL, Rep. No. 67-0237. 94 pp. Prabhakara, C., V. V. Salmonson, B. J. Conrath, J. Steranda, and L. J. Allison, (1971), Nimbus 3 IRIS Ozone Measurements Over Southeast Asia and Africa During June and July 1969. J. Atmos. Sci., 28, 828-831. Ramanathan, K. R. and R. N. Kulkarni, (1960), Mean Meridional Distributions of Ozone in Different Seasons Calculated From Umkehr Observations and Probable Vertical Transport Mechanisms. Quart. J. Roy. Meteor. Soc. 144-155. Randhawa, J. S., (1971), The Vertical Distribution of Ozone Near the Equator. J. Geophys. Res., 76, 8139-8142. Shah, G. M., (1967), Quasi-biennial Oscillation in Ozone. J. Atmos. Sci., 24, 396-401. Walshaw, C. D., (1954), An Experimental Investigation of the 9. 6pm Band of Ozone. Thesis, Cambridge Univ. Walshaw, C D., (1955), Line Widths in the 9. 6p Band of Ozone. Proc. Phys. Soc. A., 6b8, 530. Walshaw, C. D., (1957), Integrated Absorption by the 9. 6M Band of Ozone. Quart. J. Roy. Meteor. Soc., 83, 315-321. Willett, H. C., (1968), Remarks on the Seasonal Changes of Temperature and of Ozone in the Arctic and the Antarctic Stratospheres. J. Atmos. Sci., 25, 341-360. 40

Water Vapor Babrov, H. J., (1967), Strength of Lines in the v and v3 Infrared Bands of H20. Warner and Swasey Co., Flushing, N. V. Prepared for AFCRL, No. 67-0156. Benedict, W. S., and L. D. Kaplan, (1959), Calculation of Line Widths in H20-N2 Collisions. J. Chem. Phys., 30, 388. Benedict, W. S. and L. D. Kaplan, (1964), Calculation of Line Widths in H20-H20 and H20-O2 Collisions. J. Quant. Spectrose. Radiat. Transfer, 4, 453-469. Benedict, W. S., and R. F. Calfee, (1967), Line Parameters for the 1. 9 and 6. 3 Micron Water Vapor Bands. ESSA Professional Paper No. 2. Washington, D. C. Burch, D. E., D. Gryvnak, E. Singleton, W. France, and D. Williams, (1962), Infrared Absorption by Carbon Dioxide Water Vapor, and Minor Atmospheric Constituents. Ohio State Univ. Res. Rep. Optical Physics Lab. No. AFCRL-62-0156. Crutcher, H. L. and H Van Loon, (1970), Selected Level Heights, Temperature, and Dew Points for the Northern Hemisphere, supercedes NAVAIR 50-1C-52, Asheville, N. Carolina. Fridovich, B. and J. R. Kinard, (1972), Intensity Half Width Products for Seven Lines in the 6. 3pm Water Vapor Band. J. Opt. Soc. Amer., 62, 542-544. Gates, D. M., R. F. Calfee, D. W. Hansen, and W. S. Benedict, (1964), Line Parameters and Computed Spectra for Water Vapor Bands at 2. 7~. Nat. Bureau of Standards Monograph 71, Washington, D. C., 126 pp. Goldman, A. and T. G. Kyle, (1968), A Comparison Between Statistical Model and Line by Line Calculations with Application to the 9. 61m Ozone and the 2. 7 Water Vapor Bands. Applied Optics, 7, 1162-1177. Goody, R. M., (1964), Atmospheric Radiation. London, Oxford Univ. Press, 436 pp. Gutnick, M., (1962), Mean Annual Mid Latitude Moisture Profiles to 31 km. Res. Rep. No. 147. AFCRL-62-681, Air Force Cambridge Res. Lab., 30pp. Houghton, J. T., (1963), The Absorption of Solar Infrared Radiation by the Lower Stratosphere. Quart. J. R. Meteor. Soc., 89, 319-331. Krakow, B., and A. R. Healy, j1969), Strengths of 31 Water Vapor Lines Between 1612 and 1429 cm-. J. Opt. Soc. Amer., 59, 1490-1491. Mastenbrook, H. J., (1968), Water Vapor Distribution In the Stratosphere and High Troposphere. J. Atmos. Sci., 25, 299-311. 41

Mastenbrook, H. J., (1971), The Variability of Water Vapor in the Stratosphere. J. Atmos. Sci., 28, 1495-1501. McKinnon, D. and H. W. Morewood, (1970), Water Vapor Distribution in the Lower Stratosphere Over North and South America. J. Atmos. Sci., 27, 483-493. Palmer, C. H., Jr., (1960), Experimental Transmission Functions for the Pure Rotational Band of Water Vapor. J. Opt. Soc. Amer., 50, 1232- 1242. Penner, S. S., (1959), Quantitative Molecular Spectroscopy and Gas Emissivities. Reading, Mass., Addison Wesley. 582 pp. Prinz, D. K., (1968), Strength-half-width Products of Self-broadened Lines in the 6. 3 Micron Band of Water Vapor. No. PB-173429, Nat. Tech. Information Service, Springfield, Va. Scholtz, T. G., D. H. Ehhalt, L. E. Heidt, and E. A. Martell, (1970), Water Vapor, Molecular Hydrogen, Methane, and Tritium Concentrations Near the Stratopause. J. Geophys. Res., 75, 3049-3054. Sissenwine, N., D. D. Grantham, and H. A. Samela, (1968), Mid-latitude Humidity to 32 km. J. Atmos. Sci., 25, 1128-1140. Starr, V., J. P. Peixoto, and R. G. McKean, (1969), Pole to Pole Moisture Conditions for the IGY, 75, 300-331, Pure and Applied Geophys. Taljaard, J. J., H. VanLoon, H. L. Crutcher, and R. L. Jenne, (1969), Climate of the Upper Air, Part I, Southern Hemisphere, Temperatures, Dew Points, and Heights at Selected Pressure Levels. NAVAIR 50-1C-55, Washington, D. C. Williamson, E. J., and J. T. Houghton, (1965), Radiometric Measurements of Emission from Stratospheric Water Vapor. Quart. J. R. Meteor. Soc., 91, 330-338. 42

Hydroxyl Radical Anderson, J. G., (1971), Rocket Measurement of OH in the Mesosphere. J. G. R., 76, 7820-7824. Bass, A. M. and D. Garvin, (1962), Analysis of the Hydroxyl Radical Vibration Rotation Spectrum Between 3900 A and 11500 A. J. Mol. Spectrosc., 9, 114-123. Chamberlain, J. W., (1961), Physics of the Aurora and Airglow. Academic Press, New York, 368 pp. Dieke, G. H. and H. M. Crosswhite, (1948), The Ultraviolet Bands of OH, Bumblebee Series Rep. No. 87 (John Hopkins Univ. ) 2 2 Engleman, R., Jr., (1972), Accurate Wavenumbers of the A 2 - X T (0, 0) and (1, 0) Bands of OH and OD. J. Quant. Spectrosc. Radiat. Transfer., 12, 1347-1350. Herman, R. C. and G. A. Hornback, (1953), Vibration-rotation Bands of OH. Ap. J., 118, 214-227. Hesstvedt, E., (1968), On the Effect of Vertical Eddy Transport on Atmospheric Composition in the Mesosphere and Lower Thermosphere. Geofysiske Publikasjoner, XXVII, 1-35. Hunt, B. G., (1966), Photochemistry of Ozone in a Moist Atmosphere. J. G. R., 71, 1385-1398. Leovy, C.. B., (1969), An Analytical Model for Photochemistry in the Presence of Water Vapor, in Meteorological and Chemical Factors in D-Region. Aeronomy - Record of Third Aeronomy Conference, Report No. 32., Univ. of Illinois, Urbana. Levy, H. II, (1971), Normal Atmosphere: Large Radical and Formaldehyde Concentrations Predicted. Science, 173, 141-143. McConnell, J. C., M. B. McElroy, and S. C. Wofsy, (1971), Natural Sources of Atmospheric CO. Nature, 233, 187-188. Penner, S. S. (1959), Quantitative Molecular Spectroscopy and Gas Emissivities. Addison Wesley, 585 pp. Wallace, L., (1960), The Constants of the 2 T_ - 27T Oh Bands. Ap. J., 132, 894-897. Wallace, L., (1969), OH Bands in the Airglow from International Dictionary of Geophysics, England, and Kitt Peak Nat. Obs. Rep. No. 142. Weinstock, B. and H. Niki, (1971), The Balance of Carbon Monoxide in Nature. Publication Preprint. Ford Motor Co., Scientific Res. Staff. 8pp. 43

Chapter 3. Molecules Containing Carbon 3. 1 Carbon Dioxide Carbon dioxide is present in the earth's atmosphere in a sufficiently large quantity and has many strong absorption bands in the infrared and hence it is a very important contributor of atmospheric radiation. Since 1909 (Gold) when the importance of atmospheric CO2 radiation to the terrestrial heat budget was noted, a vast amount of effort to understand this radiative transfer process of CO2 (Drayson 1966, Plass, 1956) and to measure the CO2 content in the earth's atmosphere has been carried out by many researchers (Gluckauf 1944, Callender 1958, Bolin and Bischof 1970, Bischof 1971). 3. 1. 1 Atmospheric Distribution A systematic program to measure carbon dioxide concentrations in the atmosphere was started during the International Geophysical Year (IGY) 1957-1958 and measurements were routinely made at many ground stations, from aircraft and on ship cruises on the Pacific, Atlantic and Indian Oceans. The United States program was mainly concentrated at Hawaii, Alaska and the Antarctica. It has been noticed that the global atmospheric content of carbon dioxide has been increasing since the dawn of the industrial revolution. The reason for this increase being the consumption of fossil fuels by man. The average atmospheric CO2 concentration at present is considered to be about 320 ppm by volume (Table 3. 1. 1) and its estimated rate of increase is about 0. 7 ppm by volume a year (Robinson and Robbins, 1969). This is an overall figure and the actual concentration of CO2 differs from place to place and time to time. A seasonal variation of atmospheric CO2 has been noticed (Pales and Keeling 1965, Junge and Czeplak 1968). The amplitude of these seasonal variations was found to decrease with altitude being approximately 6 ppm at 44

about 2 km and decreasing to about 3 ppm at 10 km altitude. In the stratosphere there is an abrupt change in that these variations are very small, about ~ 1 ppm. This is due to the considerably smaller vertical mixing in the stratosphere. The maximum CO2 content in the northern hemisphere occurs in April-May and the minimum during the months of September-October. This is due to the cycle of the growth of vegetation on the earth's surface. Close to the ground a diurnal variation of C02 with a maximum in the predawn hours can also be noticed due to the diurnal variation of the respiration by plants. In the stratosphere Bischof (1965, 1971) noticed that during May, the time of tropospheric C02 maximum the C02 content in the lower stratosphere was 4 ppm lower than in the upper troposphere, while in October the time of CO2 minimum in the troposphere, it was 3 ppm higher. The tropopause thus acts as a damping layer for vertical exchange. Hays and Olivero (1970) have considered the photochemical production and loss and diffusion of CO2 and calculated atmospheric CO2 profiles for model atmospheres. They show that CO2 is well mixed throughout the lower stratosphere and troposphere, while deviations from complete mixing are possible in the upper stratosphere and mesosphere. Fig. 3. 1. 1 shows the average distribution of CO2 in the atmosphere. Carbon dioxide is produced naturally by biological decay and in respiration. Outgassing from volcanoes, and other geochemical emmisions are other natural sources. Release of CO2 from the oceans also occurs. Man's pollution by the combustion of fuels is another major source. Estimated emission rates for natural and pollution sources is 1012 and 1. 4 x 1010tons/year (Robinson and Robbins, 1970). CO2 is removed from the atmosphere by absorption in biological processes and in photosynthesis and byhe absorption by the oceans. There is a continuous exchange of CO2 between the oceans and the atmosphere 45

Author Location Altitude Technique Year Cone. Remarks Hagemann et. al. (1959) U. S. Panama Brazil 45000 -100000 ft Balloon flight Air sampling 1957-58 312ppm Avg: 312ppm 31 Oppm Steinberg and Rohrbough(1962) Texas 22 km Balloon flight absorption on Zeolite March 1961 323-338ppm Bolin and Keeling (1963) Global esp. Pacific Ocean, Hawaii Surface 700 mb 500 mb Continuous infrared gas analysis flask sampling 1957-62 315~8 ppm Seasonal and latitudinal variations noticed. Yearly increase; sources in tropical oceanic areas and industrial area in Northern Hemisphere at mid latitudes. Brown and Keeling (1965) Bischof (1971) Ref. Papers (1960, 62, 70) Antarc- Ground tica Level Continuous infrared gas analyzer Air sample collection 1960-63 313. 1~3. 1 ppm 1963-68 Average 322. 9ppm Regular seasonal variations in troposphere, decreasing amplitude with altitude 0. 7 ppm/year increase noted Sweden Ground to lower stratosphere Table 3. 1. 1: Measurements of CO2 in the atmosphere

on a global scale (ReVelle and Suess 1957, Takahasi T. 1961). Speculation that pollution by man is possibly altering the delicate balance between atmospheric CO2 and CO2 in the oceans has been carried out. Eriksson (1963) discusses the effect of changes of surface temperature of the oceans on atmospheric CO2. Photodissociation of CO2 occurs in the thermosphere and mesosphere. The photochemistry of atmospheric CO2 has been treated by Bates and Witherspoon (1952). CO2 is relatively inactive chemically and has a residence time of 2-4 years in the atmosphere. This conservative character makes CO2 suitable as a tracer for studying transport processes in the atmosphere. 3. 1. 2 Infrared Spectral Properties Carbon dioxide has a complex infrared spectrum. The molecule is triatomic and linear. There are various isotopic species in the atmosphere (Keeling 1960) but the main most abundant (approximately 98. 42%) is C120 16 2 The isotopic species C13 16 12016018 The isotopic species C 0216 and C 10101, form about 1. 11% and 0. 41% respectively and the other isotopic species are present in much smaller fractions of the total carbon dioxide. The carbon dioxide infrared spectrum has been thoroughly 12 16 13 16 studied by many authors. The C 1202 and C 02 molecules are symmetric (O-C-O) and the infrared spectrum consists of the v2 (15M m) and v3 (4. 3,um) fundamentals. The v3 fundamental being a parallel hand consists of only P and R branch lines. In the 15Mm region, besides the v2 fundamental, are many overtone and combination bands, the total strength of all these together being approximately one-eighth that of the v2 fundamental. In the 4. 3,m region also there are other overtone and combination bands besides the fundamental v3. Besides these two regions 15,m and 4. 3am where there are strong bands, there are other regions of the spectrum where C02 has other overtone and combination 47

70 60 50 40 30 20 10 -8 10-7 10-6 10-5 1-4 10 Volume Mixing Ratio Fig. 3. 1. 1: Atmospheric distribution of CO2, CO and CH I: (01'0-020) 2: (000o-O 'O) 0.15- 3: (01'0-02*0) 4: (000 -00 o ) 2 2 "//3 qco,=7.18 x Io2esu. '" Cz 3-CO, qN. "32 8x IO"e esu. l 0. 10 " bmcoa-co&- 4.88a - 9^ bin.^^3 Go,- ~ ~bm. IcO-N 4.9 4. mzo. f CO,- N, 0.05 - 0 t0 20 30 40 50 60 70 80 90 100 M Fig. 3. 1. 2: Variation of self- and N2- broadened line widths with rotational quantum index number for CO2 bands. (from Yamamoto et. al., 1969) 48

Region Center,u-m cm1 Upper Lower Intensity at 300~K cm(atm cm)- p STPP Reference 1. 2 8192. 6 8136. 0 8294.0 8276. 8 7593. 5 7460. 4 1. 4 6972.5 6935.0 6679. 7 1.6 6503.0 6347. 8 6227. 9 6057.9 5857. 6 5687. 0 5584. 3 5349. 0 2. 0 5248.0 5099. 6 4977. 8 4853.6 4591.0 2. 7 3723. 2 3714. 8 3612. 8 3580.3 4.3 2349.3 2336.7 4.8 2129.8 2093. 4 2076.9 5. 2 1932. 5 9.4 1061.0 10.4 961.0 (103:0)II (1 13:1)II (103:0)1 (113:1)1 (401:0)III (401:0)IV 003:0 013:1 (112:1)1 (301:0)I (301:O)II (301:O)III (301: O)IV (102:0)II 003:0 003:0 (102:0)I (102:0)II (201:0)1 (201:0)11 (201:0)111 (301:1)111 (1 1 1:1)I (l01:O)I (101:0)11 (111:1)I 001:0 011:1 (200:0)1 (120:2)I (110:1)1 (110:1)II (100:0)1 (100: 0)II 000:0 010:1 000:0 010:1 000:0 000:0 000:0 010:1 000:0 000:0 000:0 000:0 000:0 000:0 (100:0)1 (100:0)1 010:1 010:1 000:0 000:0 000:0 000:0 010:1 000:0 000:0 010:1 000:0 010:1 010:1 010:1 000:0 000:0 001:0 001:0 0. 0012 0. 0001 0. 0017 0. 00015 0.00029 0. 00011 0. 041 0. 0033 0. 0001 0.00161 0.0132 0. 00137 0. 00145 0. 00001 0.00002 0.00002 0. 14 0. 33 0. 33 1.05 0. 23 0. 42 3. 5 44.9 29.0 2. 3 2700. 0 210. 0 0.01 0. 041 0. 66 0. 05 0. 0159 0. 0199 Burch et. al. 1965 Burch et. al. 1967 Burch et. al. June 1965 Schurin & Ellis 1968 Burch et. al. Aug. 1965 Burch et. al. 1964 Burch et. al. 1969 Schurin & Ellis 1970 McCubbin & Mooney 1968 Drayson 1972 15. 0 828. 3 791. 4 757.5 741. 7 738. 6 688. 7 668. 1 667. 8 667.4 647.1 720.8 618.0 597.3 581.7 544.3 (120:2)1 (110:1)I (120:2)1 (110:1)I (200:0)11 (1 10:1)I 030:3 020:2 010:1 (110:1)II (100:0)I (100:0)II (110:1)II (120:2)11 (110:1)II (110:1)1 (100:0)II 030:3 020:2 (110:1)11 (100:0)I 020:2 010:1 000:0 (100:O)H1 010:1 010:1 020:2 030:3 (100:0)I 0.00022 0. 0242 0. 0099 0. 158 0. 0154 0.33 0. 93 16. 5 214.0 0. 77 5. 5 4. 7 0. 154 0. 0092 0. 011 Table 3. 1. 2: 12 ) bands.6 Band centers and intensities of CO2 (C 12 ) bands. 49

bands. There are bands at lOmm, 5um, 2 7pm and also some weaker bands at 2gm, 1. 6Mm, 1. 4~m and 1. 24m regions of the electromagnetic spectrum. Many authors have carried out theoretical calculations and experimental measurements of the CO2 bands (Boese et. al. 1966; Burch et. al., 1962; 1970; Calfee and Benedict, 1966; Drayson and Young, 1965, 1966, 1967; Drayson et. al., 1968; France and Dickey, 1955; Gryvnak et. al., 1966; Oberly et. al., 1968; Plyler et. al., 1962; Tubbs and Williams, 1972; Winters et. al., 1964; References in Table 3. 1. 2). Table 3. 1. 2 gives a list of the important bands of carbon dioxide, the band centers and intensities. The vibrational and rotational constants for the CO2 bands have been determined by many workers and are known fairly accurately. Detailed calculations on the relative intensities of individual lines for various bands of the CO2 molecule have been published by Gray (1967), Gray and Young (1969), Young (1970) and Young et. al. (1970). The variation of line half-width for important bands has been calculated by Yamamoto et. al. (1969), using the Anderson theory. 3. 1. 3 Atmospheric Application Transmittance calculations show that CO2 absorption can be readily measured in all the strong band regions for all tangent heights in the stratosphere and lower mesosphere. Calculations of CO2 absorptance at 655 cm have been made for a model atmosphere, (Table 3. 1. 3) assuming a concentration distribution of CO2 of constant mixing ratio of 320 ppm by volume. 50

Tangent Mixing A verage Height Ratio Abs (km) (ppmv) (atm cm)300K (mb) (OK) 5 320 11100. 372. 239. 1. 10 320 5760. 183. 218. 4 1. 15 320 2650. 84. 217. 6 1. 20 320 1210. 38. 3 219. 5 1. 25 320 553. 17. 7 225. 1. 30 320 257. 8. 3 231. 6 0.84 35 320 121. 4. 242. 6 0. 4 40 320 59. 2. 255. 0. 2 45 320 30. 1. 264. 7 0. 1 50 320 16. 0. 56 266. 1 0.05 55 320 8. 5 0. 3 258.7 0.03 60 320 4. 5 0. 16 247.4 0.02 65 320 2.3 0. 08 231.5 0.01 70 320 1. 1 0. 04 214. 4 0. 0 75 320 0.46 0. 02 198. 3 0. Table 3. 1. 3: Average absorption at 655 cm-l by atmospheric CO2 The above calculations are approximate and calculate the average absorption by CO2 using the isolated line model. Only, Lorentz broadening of the spectral line is incorporated. No variation of the strength or the half-width with temperature has been included. For an accurate calculation of the atmospheric transmittance, the variation of line strengths and line half-width with tem, perature, and Doppler broadening of lines must be included. 51

Measurement of CO2 absorption in the atmosphere is slightly complicated by the fact that other constituents have absorption bands close to those of CO2. In the 15gm region there is absorption by water vapor (especially in the troposphere) and by ozone (especially in the stratosphere) and N2O0 (17gm), in the 4. 3gm there is absorption by water vapor, N20 (4. 5 m), CO (4. 6gm), and in the 2. 7gm region there is absorption by water vapor and methane. However using the ideal spectral channels for measurement, the accurate determination of atmospheric CO2 absorption and thereby the CO2 distribution in the atmosphere appears feasible. 3. 2 Carbon Monoxide 3. 2. 1 Atmospheric Distribution Carbon monoxide is one of the minor constituents in the atmosphere. Its atmospheric presence was first noticed spectroscopically by Migoette (1949) and many subsequent measurements of its atmospheric content have been made (Table 3. 2. 1). Recent measurements by Seiler and Junge 1969, made on airplane flights show that CO is well mixed in the troposphere. The average global concentration of CO in the atmosphere is roughly 0. 12 ppm and the estimated total amount in the atmosphere is 7. 4 x 10 gm. Considerable variations of the atmospheric CO content from the average value of 0. 12 ppm have been noticed depending on the location of measurement. Measurements in the atmosphere over oceans far from pollution sources showed CO concentrations as low as 0. 06 ppm while CO content in the atmosphere over highly polluted areas was found to be as high as 0. 3 ppm. Measurements in the Southern hemisphere indicate lower concentrations of CO of 0. 06 ppm. In the stratosphere Seiler and Warneck (1972) measure a rapid decrease in concentration. They-noticed suddenchanges from tropospheric values of 0. 12 ppm 52

Author Location Altitude Technique Year Cone. Remarks Benesch et. al. (1953) Jungfrauj och Switzerland total am ount Spectroscopically 1950-51 0. 031-0. 14 ppm From Junge (1963) Locke and Herzberg (1953) Ottawa Ohio Jungfrauj och Mt. Wilson total am ount Spectroscopically 1952 0. 11-0. 22 0. 19 0. 15 0. 16 ppm Robbins et. al. (1968) Greenland West U. S. Alaska N. Pacific Near coast S. Pacific near surface Chemical method 1965 0. 2-0. 8 0. 8 0. 09 0. 1 0. 03-0. 06 0. 05 ppm Cin.Seller and CO Junge (1969) Polar FrankfurtAnchorage Tokyo flights 10 km upper troposphere and lower stratosphere Commercial aircraft Lufthansa 1968 0. 1-0. 2ppm sudden decrease to 0. 03ppm in stratosphere Seiler and Junge (1970) Western Europe South Africa Atlantic Ocean 10~S to 60~N at 30~W upper troposphere lower stratosphere surface water upper troposphere and lower stratosphere Commercial aircraft flights ship cruise chartered aircraft continuous recording instrument 1967-69 0. 1-0. 15ppm mid tropospheric 0. 16-0.3ppm conc. show marke, (in polluted differences betwee areas) hemispheres 10 to 40 times the atmospheric equilibrium value Seller and Warneck (1972) Western Europe Jan- Mar 1971 average 0. 12 in tropopause 0. 4 in stratopause Table 3. 2. 1: Measurements of CO in the atmosphere

to stratospheric values of 0. 03 ppm. There are not many measurements at higher altitudes because of the difficulty in measurement. The molecular weight of CO is nearly the same as N2 and this makes identification by mass spectrometry extermely difficult. The main sources of CO in the atmosphere are forest fires (Robinson and Robbins, 1968), release by the oceans by the photochemical oxidation of dissolved organic carbon (Wilson et. al. 1970), photochemical reactions of terpenes and other organic matter, the oxidation of atmospheric methane (Wofsy et. al., 1972), dissociation of atmospheric CO2 by lightning (Swinnerton et. al., 1971) and of considerable importance man's contribution by the combustion of fuels. This last source has been steadily increasing since the industrial revolution, and the present estimated rate of pollution is 3. 4 x 101 gm. of CO a year. Jaffe (1968) and Grenda et. al. (1971) have given a detailed listing of the various sources of atmospheric CO. Recent measurements (Lamontagne et. al., 1971) of CO content in surface waters of oceans show very high concentrations ( 15 to 75 times that of the equilibrium value in the atmosphere) with a diurnal cycle showing a maximum during daylight hours. This seems to indicate a very large production of CO in the surface layers of the oceans. Thus release of CO by the oceans could be a very important source of atmospheric carbon monoxide. The processes by which CO is being removed from the atmosphere are still only slightly understood. It is accepted that there must be such processes, since the atmospheric CO content has not increased considerably in spite of the huge inputs of CO. Chemical reactions in the stratosphere (Hesstvedt 1970, Pressman and Warneck 1970, Levy 1971) and at ground level (Westberg 1971) and biological actions on the ground and in the sea are some 54

of the sinks of atmospheric CO. Plant life absorbs CO from the atmosphere and this is thought to be an important sink for CO (Douglas, 1967). A recent review of the sources and sinks of atmospheric CO has been presented by Bortner et. al. (1972). Wofsy et. al. (1972) have calculated CO concentration profiles in model atmospheres. In the stratosphere the dominant sink mechanism is the reaction CO + OH —CO2 + H. Junge et. al. (1971) have calculated CO residence times and vertical fluxes in the troposphere and at the tropopause. In the upper stratosphere and the mesosphere the photodissociation of CO2 is responsible for the increase in mixing ratio of CO. Fig. 3. 1. 1 shows the average distribution of atmospheric CO. Carbon monoxide is toxic and at high concentrations is fatal to human and animal life (Stern 1968). CO has been listed as a possible serious health hazard by the National Air Pollution Control Administration and is one of the gasses measured in the U. S. Public Health Service Continuous Air Monitoring program as an air quality indicator in cities. CO has absorption bands in the infrared region of the electromagnetic spectrum and hence it contributes to a small extent to the earth's long wave radiative transfer and heat budget. Also the overall CO balance is interrelated to the CO2 balance and the photochemistry of the two are interdependent (Kummler et. al., 1969). Hence a detailed knowledge of the radiative properties and the distribution in the atmosphere of CO are important. 3. 2. 2 Infrared Spectral Properties The infrared spectrum of CO has been extensively studied because of its great simplicity. CO is a diatomic molecule and there are many isotopic species in the atmosphere, but approximately 98. 7% of total CO is C 20 The fundamental 0-1 band of CO is at 4. 6pm and the first and 55

second overtone bands 0-2 at 2. 7pm and 0-3 at 1. 577m. Theoretical calculations of line strengths and line half-widths have been carried out by Kunde (1967). Experimental studies in the 0-1 bands (Benedict et. al. 1962) 0-2 bands (Kostkowski and Bass 1961, Korb et. al. 1968, Hunt et. al. 1968) and 0-3 bands (Plyler and Thibault 1963, Burch and Gryvnak 1967, Bouanich and Haeusler 1972) have been reported. For remote sensing applications these CO bands have some disadvantage in that there is overlapping by absorption lines of other constituents. There is overlapping by CO2, N20, 03 and H20 absorption bands in the 4. 6pm band of CO and the 2. 3pm band is strongly overlapped by water vapor and methane absorption bands. Table 3. 2. 2 lists the strengths of the important bands of CO. Fig. 3. 2. 1 gives the variation of half width with rotational quantum number for CO bands. Region Center Transition Intensity at 300~K M m cm-1 in cm-l(atm cm)S1 STP 1. 57 6350.5 3-0 0. 0130 2. 7 4260.0 2-0 2.06 4. 6 2143.2 1-0 261.0 Table 3. 2. 2: Band centers and intensities of CO (C12016) bands 56

0. 09 I I I I Tl rI E o *r-l "d 0. 07 0. 05 0. 03 I I I 5 10 15 20 1M1 Fig. 3. 2. 1: Variation of N -broadened line half-width with rotational quantum index number for CO bands (from Bouanich and Hauesler, 1972) 57

3. 2. 3 Atmospheric Application Calculations of average absorptance at 2173 cm have been made for a model atmosphere (Appendix) assuming a concentration distribution for CO (Fig. 3. 1. 1). Mixing ratio 0. 1 ppm from 0-12 km, 12-25 km varying linearly, 25-40 km constant 0. 02 ppm, 40-79 km varying linearly 79 km 1 ppm. The calculations are similar to those described for CO2 in section 3.1. 3. Tangent Mixing T Average Height Ratio Abs (km) (ppmv) (atm cm),,,n (mb) (OK) a uur", 5 10 15 20 25 30 35 40 45 50 0. 1 0. 1 0, 08 0. 05 0. 02 0. 02 0. 02 0. 02 0. 14 0. 27 3. 3 1. 6 0. 54 0. 15 0. 045 0. 028 0. 021 0. 022 0. 023 0. 018 387 197 92 40 14 5. 2 2. 1. 2 0. 84 0. 49 240 218 218 221 233 243 253 258 262 263 0. 52 0. 26 0, 1 0. 036 0. 012 0. 006 0. 003 Table 3. 2. 3: Average absorption at 2173 cm1 by atmospheric CO There is measurable absorption in the stratosphere, and by the suitable choice of spectral channels the measurement of atmospheric CO from a satellite appears feasible. 58

3.3 Methane 3. 3. 1 Atmospheric Distribution Methane is present in the earth's atmosphere in a considerable amount, being the eighth most abundant gas in the earth's atmosphere. The main sources of atmospheric methane are the decomposition of organic matter (Koyama 1963, Ehhalt 1967) and the emissions from coal and natural gas fields and petroleum industries. No known sources of methane exist in the atmosphere. The total estimated global emission rate is approximately 8 x 1014 gm/year. The major sink for atmospheric methane is thought to be the oxidation to H20 and CO2 in the stratosphere. Migoette (1948) made the first measurements of atmospheric CH4 content and later measurements followed (Table 3. 3. 1). Most of the measurements have been the determination of total amount of methane in the atmosphere or the determination of methane content of ground level air. Measurements of methane content higher in the atmosphere are very very few, and of doubtful accuracy. The measurements indicate that methane is well mixed up to the tropopause with a mixing ratio varying from 0. 6 to 1. 6 ppm by volume and that the mixing ratio decreases above the tropopause to about 65%7o of its tropospheric average value at 25 km and to 20% of its tropospheric average value at the stratopause (Ehhalt et. al., 1972). Crutzen (1971) has calculated theoretical distributions of atmospheric methane showing a rapid decrease of methane above 25 km to nearly zero value at 50 km. The variation 0. 6 to 1. 6 ppm of the mixing ratio of methane could be latitude, seasonal or proximity to a source or sink variation. However the widely differing techniques, and the inherent difficulty of measurements could be the reason for the uncertainty. Calculations of methane content are also roughubecause of the limited knowledge of the sources and sinks of methane. More detailed measurements are required for the 59

Author L oc ati on Altitude Technique Year Conc. Remn arks Goldberg (1951) Bowman and Shaw (1963) Fink et. al. (1964) Bainbridge and Heidt (1966) Kyle et. al. (1969) Michigan Mt. Wilson Columbus Ohio Pennsylvania Southern U. S. A. East Texas and West Louisiana U. S. Holloman Air Base Total am ount ground level total amount 0-23 km 0-30 km Spectroscopic 1. 67p measurement Spectroscopic 3. 3A measurement Spectroscopic 1. 67,u measurement Air sampling balloon and aircraft flights Balloon flight absorption at 1300 cm-1 vertical amount in atmosphere 1950 1963 1963 MarchJuly 1965 Dec. 1967 1. 2 (atm cm)STP 1. 7~0. 3ppm 1. 11 (atm cm)STP 1. 4-1. 6ppm in troposphere rapidly decreases in stratosphere 1. 1-1. 3ppm at 24 krn. total vertical amount 1. 2 atm cm decreases above tropopause to 0. 3 atm cm at 24 km Scholz et. al. (1970) White Sands New Mexico 44-62 km Air sampling with cryocondenser on rocket flight Sept. 1968 Stratopause average 0. 25~0. 01 ppm between 44 and 62 km Table 3. 3. 1: Measurements of CH4 in the atmosphere

determination of atmospheric methane distribution. Fig. 3. 1. 1 shows the approximate distribution of methane in the atmosphere. 3. 3. 2 Infrared Spectral Properties Methane has a very complex infrared spectrum characteristic of a tetrahedral molecule. Many theoretical and experimental studies have appeared in the literature (Plyler et. al. 1960, Barnes et. al. 1972, Kyle 1968, Margolis and Fox 1968, Hecht 1960, Henry et. al. 1970, Varanasi and Tejwani 1972, McMahon et. al. 1972). More laboratory measurements in the area of determination of line strengths and half-widths of methane bands for atmospheric application are useful. The v3 (3. 3gm) and v4 (7. 7gm) fundamentals are active in the infrared. Besides these two bands there are many overtone and combination bands at 1. lgm, 1. 7gm, 2. 3gm, 3. 3gjm, 3. 6gjm, 3. 9gim, 7. 7gm and other regions of the spectrum. Table 3. 3. 2 gives a few important bands and the band strengths. Region Band Center Intensity at 300~K in cm-l (atm cm)s Region (atm cm ~STP Am cm 1. 67 6005 1. 6 3.31 3019 370.0 3. 85 2600 307. 0 7. 66 1306 204. 0 Table 3. 3. 2: Band centers and intensities of CH4 bands Adapted from Goody (1964) 61

3. 3. 3 Atmospheric Application The absorbpion by atmospheric methane is estimated from Burch et al (1962) and is shown in Table 3. 3. 3. Optical masses were calculated by the method described in the appendix. The methane distribution assumed (Fig. 3. 1. 1. ) is mixing ratio 1. 2 ppm from 0-14 km varying linearly to 0. 8 ppm at 24 km and decreasing to 0. 25 ppm at 50 km. a R r: _ _: _ - Tang He i (ki 1 21 2 3 3 4 4 5 J - sent lmixing ght Ratio m) (ppmv) (atm cm)300. 5 1.2 40.3 0 1. 2 20. 5 1.04 7.8 0 0. 88 3. 5 0. 74 1. 2 0 0. 64 0. 46 5 0. 54 0.18 0 0. 45 0. 07 5 0.35 0. 025 0 0. 25 0. 009 Table 3. 3. 3: Average absorption p (mb) 382 191 87. 6 40 T (OK) Absorption 240 218 217 219 18.5 224 8. 8 230 4. 3 241 2. 2 254 1. 2 266 0. 65 270 at 3020 cm1 by atmospheric CH4.7. 45.37.30. 22.12.08. 03. 01.0 The accurate measurement of absorption by atmospheric methane from a satellite appears feasible. 62

References Bainbridge, A. E. and L. E. Heidt, (1966), Measurements of Methane in the Troposphere and Lower Stratosphere. Tellus, 18, 221. Barnes, W. L., J. Susskind, R. H. Hunt and E. K. Plyler, (1972), Measurement and Analysis of the v3 Band of Methane. J. Chem. Phys., 56, 5160. Bates, D. R. and A. E. Witherspoon, (1952), The Photochemistry of Some Minor Constituents of the Earth's Atmosphere (CO2, CO, CH4, N2O). Mon. Not. Roy. Astron. Soc., 112, 101. Benedict, W. S., R. Herman, G. E. Moore and S. Silverman, (1962), The Strengths, Widths and Shapes of Lines in the Vibration-Rotation Bands of CO. Astrophys. Jour., 135, 277. Benesch, W. M. Migeotte, and L. Nevan, (1953), Investigation of Atmospheric CO at the Jungfraujoch, J. Opt. Soc. of Am., 43, 1119. Bischof, W., (1960), Periodic Variations of the Atmospheric CO Content in Scandinavie. Tellus, 12, 216. Bischof, W., (1962), Variations in Concentration of Carbon Dioxide in the Free Atmosphere. Tellus, 14, 67. Bischof, W., (1965), Carbon Dioxide Concentrations in the Upper Troposphere and Lower Stratosphere I. Tellus, 17, 398. Bischof, W., (1970), Carbon Dioxide Measurements from Aircraft. Tellus, 22, 545. Bischof, W., (1971), Carbon Dioxide Concentration in the Upper Troposphere and Lower Stratosphere II. Tellus, 23, 558 Bischof, W., and B Bolin, (1966), Space and Time Variations of the CO Content of the Troposphere and Lower Stratosphere. Tellus, 18, 155. Boese, R. W., J. H. Miller, and E. C. Y. Inn, (1966), Intensity Measurements of the lpm C02 Bands. J. Q S. R. T., 6, 717. Bolin, B., and W. Bischof, (1970), Variations of the Carbon Dioxide Content of the Atmosphere in the Northern Hemisphere. Tellus, 22, 431. Bolin, B., and C. D. Keeling, (1963), Large Scale Atmospheric Mixing as Deduced from Seasonal and Meridional Variations of Carbon Dioxide. J. G. R., 68, 3899. Bouanich, J. P. and C. Hauesler, (1972), Linewidths of Carbon Monixide Self-broadening and Broadened by Argon and Nitrogen, J. Q. S. R. T., 12, 695. 63

Bortner, M. H., R. H. Kummler, and L. S. Jaffe, (1972), A Review of Carbon Monoxide Sources, Sinks and Concentrations in the Earth's Atmosphere. NASA-CR- 2081. Bowman, R. L., and J. H. Shaw, (1963), The Abundance of Nitrous Oxide, Methane and Carbon Monoxide in Ground Level Air. App. Optics, 2, 176. Brown, C. W., and C. D. Keeling, (1965), The Concentration of Atmospheric Carbon Dioxide in Antarctica. J. G. R., 70, 6077. Burch, D. E., and D.A. Gryvnak, (1967), Strengths, Widths, and Shapes of the Lines of the 3v CO Band. Philco-Ford Corp. Aeronutronic Div. Publ. U-3972. Burch, D. E., D. A. Gryvnak, and R. R. Patty, (1964), Absorption by CO between 4500 and 5400 cm1 Philco (Ford Motor Co. ) Aeronutronic Dv. Publ. No. U-2955. Burch, D. E., D. A. Gryvnak, ard R. R. Patty, (1965), Absorption by CO2 between 8000 and 10, 000 cm (1-1. 25 micron region) Aeronutronic Div. Philco Corp. Publ. No. U-3200. Burch, D. E., D. A. Gryvnak, and R. R. Patty, (June, 1965), Absorption by CO between 6600 and 7125 cm (1. 4 micron region). Aeronutronic Div. Phiico Corp. Scientific Report Publ. No. U-3127. Burch, D. E., D. A. Gryvnak, and R. R. Patty, (Aug., 1965), Absorption by CO between 5400 and 6600 cm (1. 6 micron region). Aeronutronic Div. Phiico Corp. Ford Motor Co. Publ. No. U-3201. Burch, D. E., D. A. Gryvnak, and R. R. Patty, (1967), Absorption by CO2 between 7125 and 8000 cml (1. 25-1. 40 microns) Philco-Ford Corp., Aeronutronic Div. Publ. No. U-3930 Burch, D. E., D. A. Gryvnak, R. R, Patty, and C. E. Bartky, (1969), Absorption of Infrared Radiant Energy by CO2 and H20O. IV. Shapes of Collision Broadened CO2 Lines. J. Opt. Soc. of Am., 59, 267. Burch, D. E., D. A. Gryvnak, and J. D. Pembrook, (1970), Investigation of the Absorption of Infrared Radiation by Atmospheric Gases, AFCRL-70-0373. Air Force Cambridge Res. Labs. Burch, D. E., D. A. Gryvnak, E. B. Singleton, W. L. France, and D. William, (1962), Infrared Absorption by CO2, Water Vapor, and Minor Atmospheric Constituents. AFCRL-62-698 Air Force Cambridge Res. Labs. Calfee, R. F., and W. S. Benedict, (1966), Carbon Dioxide Spectral Line Positions and Intensities Calculated for the 2. 05 and 2. 7 micron regions. Nat. Bur. Std. Tech. Note: 332. Callendar, G. S.,(1958), On the Amount of Carbon Dioxide in the Atmosphere. Tellus, 10, 243. 64

Crutzen, P. J., (1971), Ozone Production Rates in an Oxygen- HydrogenNitrogen Oxide Atmosphere, J. G. R., 76, 7311. Douglas, E., (1967), Carbon Monixide Solubilities in SeaWater, J. Phys. Chem., 71, 1931. Drayson, S. R., (1966), Atmospheric Transmission in the CO2 Bands between 12g and 18/1. App. Optics, 5, 385. Drayson, S. R., (1972), Atmospheric Radiative Transfer by Carbon Dioxide, Conference on Atmospheric Radiation, Aug. 1972, AMS, Fort Collins, Colorado. Drayson, S. R., and C. Young, (1965), Band Strength and Line Half-width of the 10. 44 CO2 Band. J. Q. S. R. T., 7, 993. Drayson, S. R., and C. Young, (1966), Theoretical Investigations of Carbon Dioxide Radiative Transfer. Rep. No. 07349-1-F, Coll. of Eng., Univ. of Michigan. Drayson, S. R., and C. Young, (1967), The Frequencies and Intensities of Carbon Dioxide Absorption Lines bdtween 12 and 18 microns. Rep. No. 08183-1-T, Coll. of Eng., Univ. of Michigan. Drayson, S. R., S. Y. Li, and C. Young, (1968), Atmospheric Absorption by Carbon Dioxide, Water Vapor, and Oxygen. Final Report No. 08183-2-F, Coll. of Eng., Univ. of Michigan. Ehhalt, D. H., (1967), Methane in the Atmosphere. J. Air. Pollution. Cont., 17, 518. Ehhalt, D. H., L. E. Heidt, and E. A. Martell, (1972), The Concentration of Atmospheric Methane between 44 and 62 km Altitude. J. G. R., 77, 2193. Erikson, E., (1963), Possible Fluctuations in Atmospheric Carbon Dioxide Due to Changes in the Properties of the Sea. J. G. R., 68, 3871. Fink, U., D. H. Rank, and T. A. Wiggins, (1964), Abundance of Methane in the Earth's Atmosphere. J. Opt. Soc. of Am., 54, 572. France, W. L., and F. P. Dickey, (1955), Fine Structure of the 2. 7 micron Carbon Dioxide Rotation-Vibration Band. J. Chem. Phys., 23, 471. Gluckauf, E., (1944), Carbon Dioxide Content of Atmospheric Air. Nature, 153, 620. Gold, E., (1909), The Isothermal Layer of the Atmosphere and Atmospheric Radiation. Proc. Roy. Soc. Lond. A 82, 43. Goldberg, L., (1951), The Abundance and Vertical Distribution of Methane in the Earth's Atmosphere. Astrophys. Jour., 113, 567. 65

Goody, R. M., (1964), Atmospheric Radiation I - Theoretical Basis. Oxford at the Clarendon Press. Gray, L. D., (1967), Relative Intensity Calculations for Carbon Dioxide III, Relative Line Intensities of Transitions from the Vibrational Ground State for Temperatures from 1600 to 280~ K. J. Q. S. R. T., 7, 795. Gray, L. D., and A. T. Young, (1969), Relative Intensity Calculations for Carbon Dioxide IV. Calculations of the Partition Function for Isotopes of CO2. J. Q. S. R. T., 9, 569. Grenda, R. N., M. H. Bortner, P. J. LeBel, J. H. Davis, and R. Dick, (1971), Carbon Monixide - Pollution Experiment I. A Solution to the Carbon Monoxide Sink Anomaly. AIAA Paper No. 71-1120. Joint Conference on Sensing of Environmental Pollutants, Palo Alto, California. Gryvnak, D. A., R. R. Patty, D. E. Burch, and E. E. Miller, (1966), Absorption by CO2 between 1800 and 2850 cm-1 (3. 5 - 5. 6 micron). Philco-Ford Corp. Aeronutronic Div. Publ. No. U-3857. Hagemann, F., J. Gray, Jr., L. Machta, and A. Turkevich, (1959), Stratospheric Carbon-14, Carbon Dioxide, and Tritium. Science, 130, 542. Hays, P. B., and J. J. Olivero, (1970), Carbon Dioxide and Monoxide above the Troposphere. Planetary and Space Sci., 18, 1729. Hecht, K. T., (1960), Vibration-Rotation Energies of Tetrahedral XY4 Part II - The Fundamental v3 of CH4 J. Mol. Spectry., 5, 390. Henry, L., N. Husson, R. Andia, and A. Valentin, (1970), Infrared Absorption Spectrum of Methane from 2884 to 3141 cm-1. J. Mol. Spectry., 36, 511. Hesstvedt, E., (1970), Vertical Distribution of CO Near the Tropopause. Nature, 225, 50. Hunt, R. H., R. A. Toth, and E. K. Plyler, (1968), High Resolution Determination of the Widths of Self Broadened Lines of Carbon Monoxide. J. Chem. Phys., 49, 3909. Jaffe, L. S., (1968), Ambient Carbon Monoxide and its Fate in the Atmosphere. J. Air Pollution Cont. Assoc., 18, 534. Junge, C. E., (1963), Air Chemistry and Radioactivity. New York, Academic Press, New York. Junge, C. E., and G. Czeplak, (1968), Some Aspects of the Seasonal Variation of Carbon Dioxide and Ozone. Tellus, 20, 422. Junge C. E., W. Seiler, and P. Warneck, (1971), The Atmospheric 12CO and 1 CO Budget. J. G. R., 76, 2866. 66

Keeling, C. D., (1960), The Concentration and Isotopic Abundances of Carbon Dioxide in the Atmosphere. Tellus, 12, 200. Keeling, C. D, T. B. Harris, and E. M. Wilkins, (1968), The Concentration of Atmospheric Carbon Dioxide at 500 and 700 millibars. J. G. R., 73, 4511. Korb, C. L, R. H. Hunt, and E. K. Plyler, (1968), Measurement of Line Strengths at Low Pressures - Application to the 2-0 Band of CO. J. Chem. Phys., 48, 4252. Kostkowski, H. J., and A M. Bass, (1961), Direct Measurement of Line Intensities and Widths in the First Overtone Band of CO. J. Q. S. R T., 1, 177. Koyama, T., (1963), Gaseous Metabolism in Lake Sediments and Paddy Soils and the Production of Atmospheric Methane and Hydrogen. J. G. R., 68, 3971. Kummler, R. H., R. N. Grenda, T. Baurer, M. H. Bortner, J. H. Davies, and J. MacDonald, (1969), Satellite Solution of the Carbon Monoxide Sink Anamoly. Trans. Amer. Geophys. Union, 50, 174. Kunde, V. G., (1967), Tables of Theoretical Line Positions and Intensities for the V=1, V=2, and V=3 Vibration Rotation Bands of C12016 and C13016 NASA X-622-67-248 Goddard Space Flight Center. Kyle, T. G., (1968), Line Parameters of the Infrared Methand Bands. Sci. Rep. No 1. Air Force Contract AF 19(628)-5706. Dept. of Physics, Univ. of Denver Kyle, T. G., D. G. Murcray, F. H. Murcray, and W. J. Williams, (1969), Abundance of Methane in the Atmosphere Above 20 km. J' G. R., 74, 3421. Lamontagne, R. A., J. W. Swinnerton and V. J Linnenbom, (1971), Nonequilibrium of Carbon Monoxide and Methane at the Air-Sea Interface J. G. R, 76, 5117. Levy, H., (1971), Normal Atmosphere-Large Radical and Formaldehyde Concentrations Predicted. Science, 173, 141. Locke, J. L., and L. Herzberg, (1953), The Absorption Due to Carbon Monoxide in the Infrared Solar Spectrum. Canad. J. Phys., 31, 504. Ludwig, C. B., R. Bartle, and M. Griggs, (1969), Study of Air Pollutant Detection by Remote Sensors. NASA CR-1308. General Dynamics Rep. No. GDC-DBE 68-011. Convair, San Diego, Calif. Margolis, J. S. and K. Fox, (1968), Infrared Absorption Spectrum of CH4 at 9050 cm-1. J. Chem. Phys., 49, 2451. 67

McCubbin, T. K., Jr., and T. R. Mooney, (1968), A Study of the Strengths and Widths of Lines in the 9. 4 and 10. 4m CO2 Bands. McMahon, J., G. J. Troup, and G. Hubbert, (1972), The Effect of Pressure and Temperature on the Half-Width of the Methane Absorption at 3. 39m. Migoette, M., (1949), The Fundamental Band of CO at 4. 7m in the Solar Spectrum. Phys. Rev., 76, 1108. Migoette, M. V., (1948), Spectroscopic Evidence of Methane in the Earth's Atmosphere. Phys., 73, 1948. Oberly, R., K. N. Rao, Y. H. Hahn, and T. K. McCubbin, Jr., (1968), Bands of Carbon Dioxide in the Region of 4. 3 microns. J. Mol. Spectry., 25, 138. Pales, J. C., and R. D. Keeling, (1965), The Concentration of Atmospheric Carbon Dioxide in Hawaii. J. G. R., 70, 6053. Plass, G. N., (1956), The Influence of the 15u Carbon Dioxide Band on the Atmospheric Infrared Cooling. Quart. J. R. Met. Soc., 82, 310. Plyler, E. K., and R. J. Thibault, (1963), Self Broadening of Carbon Monoxide in the 2v and 3v Bands. J. Res. Nat. Bur. Std., 67A, 201 Plyler, E. K., E. D. Tidwell, and L. R. Blaine (1960), Infrared Absorption Spectrum of Methane from 2470 to 3200 cm1. J. Res. Nat. Bur. Std. 64A, 201. Plyler, E. K., E. D. Tidwell, and W. S. Benedict, (1962), Absorption Bands of Carbon Dioxide from 2. 8-4. 2M. J. Opt. Soc. of Am., 52, 1017. Pressman, J. and P. Warneck, (1970), The Stratosphere as a Chemical Sink for Carbon Monoxide. J. Atmo.. Sc., 27, 155. Revelle, R., and H. Suess, (1957), Carbon Dioxide Exchange Between Atmosphere and Ocean and the Question of an Increase of Atmospheric CO2 During Past Decades. Tellus, 9, 18. Robbins, R. C., K. M. Borg, and E. Robinson, (1968), Carbon Monoxide in the Atmosphere. J. Air. Poll. Control Assoc., 18, 106. Robinson, E., and R. C. Robbins, (1968), Sources, Abundance and Fate of Gaseous Atmospheric Pollutants. Final Rep. SRI. Project Pr-6755. Amer. Petr. Inst. Washington, D. C. Robinson, E. and R. C. Robbins, (1970), Global Effects of Environmental Pollution. Ed. S. F. Singer, Springer Verlag, N. Y. Scholz, T. G., D. H. Ehhalt, L. E. Heidt, and E. A. Martell, (1970), Water Vapor, Molecular Hydrogen, Methane and Tritium Concentration Near the Stratopause. J. G. R., 75, 3049. 68

Schurin, B. D., and R. E. Ellis, (1968), Integrated Intensity Measurements of Carbon Dioxide in the 2. 0O, 1. 6Qq, and 1. 43p regions. App. Optics., 7, 467. Schurin, B. D., R. E. Ellis, 1970), Total Intensity Measurements for the C02 Bands in the 961 cm- and 1061 cm-1 regions. App. Optics, 9, 223. Seiler, W. and C. Junge, (1969), Decrease of Carbon Monoxide Mixing Ratio Above the Polar Tropopause. Tellus, 21, 447. Seller, W. and C. Junge, (1970), Carbon Monoxide in the Atmosphere. J. G. R., 75, 2217. Seller, W. and P. Warmeck, (1972), Decrease of the Carbon Monoxide Mixing Ratio at the Tropopause. J. G. R., 77, 3204. Steinberg, S. and S. F. Rohrbough, (1962), The Collection and Measurement of Carbon Dioxide and Water Vapor in the Upper Atmosphere. J. Appl. Meteor., 1, 418. Stern, A. C., (1968), Air Pollution Vol. 1. Second Ed. Academic Press, N. Y. Swinnerton, J. W., V. J. Linnenbom, and R. A. Lamontagne, (1971), Carbon Monoxide in Rainwater. Science, 172, 943. Takahashi, T., (1961), Carbon Dioxide in the Atmosphere and in Atlantic Ocean Water. J. G. R., 66, 477. Tubbs, L. D., and D. Williams, (1972), Broadening of Infrared Absorption Lines at Reduced Temperatures: Carbon Dioxide. J. Opt. Soc. of Am., 62, 284. Varanasi, P. and G. D. T. Tejwani, (1972), Experimental and Theoretical Studies on Collision Broadened Lines in the v4 Fundamental of Methane. J. Q. S. R. T., 12, 849. Westburg, K., N. Cohen, and K. W. Wilson, (1971), Carbon Monoxide. Its Role in Photochemical Smog formation. Science, 171, 1013. Wilson, D. F., J. W. Swinnerton, and R. A. Lamontagne, (1970), Production of Carbon Monoxide and Gaseous Hydrocarbons in Sea Water. Relation to Dissolved Organic Carbon. Science, 168, 1577. Winters, B., S. Silverman, and W. Benedict, (1964), Line Shape in the Wing Beyond the Band Head of the 4. 3m band of CO2. J. Q. S. R. T., 4, 527. Wofsy, S. C., J. C McConnell, and M. B. McElroy, (1972), Atmospheric CH4, CO and CO2. J. G. R., 77, 4477. Yamamoto, G., M. Tanaka, and T. Aoki, (1969), Estimation of Rotational Line Widths of Carbon Dioxide Bands. J. Q. S. R. T., 9, 371. 69

Young, L. D. G., (1970), Relative Intensity Calculations for Carbon Dioxide V., Relative Intensities of the More Abundant Isotopes. J. Q. S. R. T., 10, 99. Young, L. D. G., A. T. Young, and R. A. Schorn, (1970), for the 7820 A and 7883 A Bands of CO2. J. Q. S. R. Improved Constants T., 10, 1291. 70

Chapter 4. Molecules Containing Nitrogen 4. 1 Nitrous oxide (N20) 4. 1. 1 Atmospheric Distribution The first positive identification of nitrous oxide in the atmosphere was reported by Adel (1938) who recognized its absorption bands in the solar spectrum near 7. 8am. Since that time atmospheric absorption spectra of N20 have been detected in other spectral regions and measurements of distribution have.been made using spectral techniques and also a variety of other methods. The measurements prior to 1966 have been summarized by Bates and Hays (1967) and more recently a comprehensive discussion by Schtitz et. al. (1970) has appeared. As is the case with many minor constituents, few measurements of concentration in the stratosphere have been made for N20. These measurements, inherently less reliable than those made near the surface, must be supplemented by estimates based on theoretical consideration involving photochemistry and transport phenonoma. Thus it is important to examine closely the distribution in the troposphere since the stratospheric N20 is transported from below. Measured concentrations are in the range 0. 12 (Seely and Houghton, 1961) to 1. 25 (Shaw et. al., 1948) parts per million by volume. This order of magnitude variation appears not to be a real effect but is probably due to errors and uncertainties in the measurement techniques. There are variations both with time and geographical location but they appear to be much smaller than the extreme figures would indicate. Fig. 4. 1. 1 (from Schuitz et. al. 1970) shows typical variations in the troposphere. These authors measurements at Mainz show an increase for the time period from late 1966 to mid 1968 71

E ZC.C t * \ 'o Mainz (120m) o Schouinsland (1280m) 20 (b) x Garmisch (730m) x Wank (1780m) Zugspitze (296am) \A Izoan (2360m) \. * \ above inversion (2200m) A Tenerife 60m and below nversion (2200m) \ \ * Transpolar flights Frankfurt - Tokyo above and below tropopaue Johannesburg i \ _i * Balloon soundings 10 normalized ~ 18 10) 0 tropopaus standard I t deviation^ ' inve/rsion J/ 18) lzallo '~ ~I" ---- 56) 6) j ~, -, - r-28) 0* 11)~~~~~~9 I I i Q15 Q2 a25 03 Q4 05_ ppm N20 Fig. 4.1.1 Atmospheric N20 concentrations (from Schtitz et. al., 1970) 1966 1967 1968 Fig. 4.1. 2 Atmospheric N20 concentrations at Mainz, Germany and Boston, Mass. (from Schutz et. al., 1970). 72

which is also suggested by the data of Goody (1969) taken at Boston, Mass. (Fig. 4. 1. 2). Goody's concentrations are consistently higher by at least 0. 05 ppmv (20%). It is not clear how much of this difference is in measurement techniques which might lead to a systematic error. The data also indicate that seasonal effects are not marked and that industrial pollution does not contribute significantly to the N20 concentration. Measurements in the tropics show similar results to those made in temperate zones (Lahue et. al., 1970). The source of atmospheric N20 is nitrate reduction due to bacterial activities and apparently is operative in the ocean as well as in soils (Junge and Hahn, 1971). An atmospheric residence time of 10 years or less is indicated. The majorknown atmospheric sink of N 0 is by photodissociation which gives a residence time of about 70 years. Clearly there must be some additional mechanism to destroy N20. Junge and Hahn suggest that arid regions, e. g., the Sahara Desert region, may be sinks since air from these regions has a lower N20 content. No mechanism to accomplish this reduction has yet been advanced. The sink of N20 in the stratosphere has been examined by Bates and Hays (1967). The main mechanism is the photodissociation of N20. N20 + hv — > N2 + O X3370R At shorter wavelengths the production of NO and N is also possible N20+ hv - NO + N X<2500R More recent data seems to exclude this as an important process. In the troposphere the photodissociation is small and the atmospheric mixing is good, so that the concentration stays relatively uniform. In the stratosphere photodissociation becomes more important and at the same time the vertical mixing becomes much less in this stable environment. Thus the vertical 73

profile of N20 depends critically on the mixing in the lower stratosphere, a quantity which is not accurately determined. Following the usual procedures Bates and Hays characterize the vertical mixing by a vertical exchange 2 5 2 -1 coefficient Kz and calculate for a range of values (10 to 10 cm sec ). The results for 10 and 104 cm sec 1are shown in Fig. 4. 1. 1 as the dashed lines. As can be seen the balloon soundings of Schutz et. al. fall between these curves which correspond roughly to the best estimates of vertical eddy mixing. The values are shown in Table 4. 1. la. Information on stratospheric N20 has also been obtained by Goldman et. al. (1970) from solar absorption measurements in the 4. 5/pm spectral region (Table 4. 1. lb). The measurements within the troposphere show much lower mixing ratios than reported by Schutz et. al. and by most of the more recent measurements. There is an indication of a reduction in the mixing ratio at the highest level but it may not be significant in view of the errors in making the measurement and the interpretation of the spectra. Stratospheric N20 concentrations have not yet been determined with sufficient accuracy or frequency to give a comprehensive picture of its stratospheric distribution. The scant measurements that do exist support the general conclusions of the theoretical study by Bates and Hays. Their calculations are of a simple nature involving photodissociation and vertical eddy diffusion, but leave out other influential factors. In particular the role of meridional and vertical transport is completely neglected. The importance of the atmospheric circulation in determining the ozone concentration in the stratosphere is well known, even if the details are as yet imperfectly understood. We should, for example, expect smaller concentrations of N20 at high latitudes in the winter because of the subsidence of air depleted of N20 74

at higher levels. If the stratospheric distributions could be measured comprehensively, then N20 could be used as a tracer to determine atmospheric transport. Global satellite measurements would be particularly suitable. Date Altitude (km) N20 Mixing Ratio ppm volume June 1967 17.3 0. 229* July 1968 18. 0 0. 204 October 1968 23.0 0. 103 * possible contamination Table 4. 4. la: Stratospheric mixing ratios (Schutz et. al. (1970) Date Altitude N20 Mixing Ratio (km) ppm volume 5.03 0.14 March 1968 8. 08 0.17 9.84 0.15 10. 3 0. 15 13.3 0.11 Table 4. 4. lb: Stratospheric mixing ratios (Goldman et. al., 1970) 4. 1. 2 Infrared Spectral Properties Nitrous oxide is a linear asymmetric molecule (N-N-O) whose chemical properties make it almost inert and easy to handle in the laboratory. These factors have contributed to its popularity with experimental and theoretical spectroscopists, so that many experimental and theoretical studies have been made of its spectral properties. Next to carbon dioxide it is the most 75

intensively studied linear triatomic molecule and resembles CO2 in many ways, e. g., laser transitions in the same spectral regions between the same energy levels. The analogy with CO2 becomes almost complete for the less abundant asymmetric isotopic molecules such as 12C1600. Transition o0110 - 0200 _ 10~0 - 1110 - 00~1 - 12~0 - 2000 - 1 011 1 - 02~1 - 1 11 1 - 10~1 - 0002 - 00~0 00 0 00~0 00~0 00~0 00~0 00~0 00~0 0000 0110 00~0 00~0 00 0o 0oco 00 0o 0oo 1oo 01 0o 0oo 0sl Band Center (cm-1) 588. 8 1168. 1 1284.9 1880.3 2223.8 2462. 0 2563.3 2798.3 3364. 0 3473. 2 3480.8 4417.4 Band Intensity cm -l(atm cm -Ok 35 10 230 0. 37 1600 10 40 2. 2 1. 7 4. 3 40 1. 6 Table 4. 1. 2: Infrared bands of N20 (Adapted from Gray Young, 1972) Table 4. 1. 2 shows the more important bands together with estimates of their intensities. For the stronger bands, which are of most interest for remote sensing applications, the intensities are probably known with sufficient accuracy. Laboratory measurements could easily be made on those for which this conclusion does not hold. The intensities in Table 4. 1. 2 76

are the total intensities in the spectral regions indicated. The listed transitions dominate the absorption spectrum but the hot bands frequently contribute to about 10% of the intensity at room temperature. The exact intensities of these weaker bands are not always well known and would thus contribute to the difficulty in the interpretation of atmospheric spectra. In addition to the hot bands, the isotopic molecules, e. g., 14N14N10, should be considered in very accurate calculations of atmospheric absorption. In order to compute the absorption of N20 under the varying conditions of pressure, temperature etc. found in the atmosphere it is desirable to have available a listing of line parameters, that is the wavenumber, intensity and Lorentz half-width of each line. The parameters needed to calculate the energy levels have been tabulated by Pliva (1968) and these have been used by Gray Young (1971) to compute the partition function for 14N2 60. This may be used to calculate the relative intensities of various transitions (Gray Young, 1972) and the absolute intensities provided that the band inteisities are known. The Lorentz half-widths of N20 are not well known, either for the self-broadened or air-broadened case. The large variation of estimates based on measurements in the laboratory has been discussed by Tejwani and Varanasi (1971). The values vary from 0. 157 to 0. 05 cm at 1 atmosphere. These authors have calculated the half-widths, including the variation with rotational quantum number J, using the Anderson theory, giving a mean value of 0. 06 cm at 1 atmosphere pressure for air-broadened N20. However experience with other molecules has shown that the Anderson theory gives satisfactory results only when the parameters needed by the theory (e. g., the quadrupole moments of the molecules involved) are adjusted to give 77

agreement with the measured values of half-width.. This situation remains unsatisfactory until such time as reliable and consistent measurements of half-width are obtained. Difficulties such as those discussed in the last paragraph make the interpretation of atmospheric spectra hazardous. Part of the disagreement between Goldman et. al., (1970), Goody (1969) and Schiutz et. al. (1970) might be explained in.terms of the infrared properties of N20. The infrared technique is a promising one and should be capable of high reliability once the minor difficulties have been resolved. 4. 1. 3 Atmospheric Application A calculation has been made to determine the optical mass of N20 for several tangent heights in the stratosphere using the technique described in the appendix. The mixing ratio profile used in the computation is shown in Table 4. 1. 3. It was obtained by upward extrapolation from Fig. 4. 1. 1 based on the values of mixing ratio calculated by Bates and Hays. It is also close to the curve given by Crutzen (1971) for a vertical eddy diffusion co3 -1 efficient of 10 cm sec Only a few laboratory measurements were available under the conditions corresponding to these tangent paths, but at the 10 km level almost complete absorption takes place in several spectral regions. It would therefore appear to be feasible to make solar occutation measurements from a satellite to determine N20 concentrations. Essentially the same procedure has already been used from a balloon. The choice of an optimum spectral region has not yet been determined. The 17,m band may have some advantage; although it is not the 78

strongest of the bands it does have an intense Q-branch which gives considerable absorption even under a resolution of a few wavenumbers so that a comparitively simple instrument could be employed. Tangent Altitude (km) 40 35 30 25 20 15 10 Mixing Ratio (ppmv) 0. 0.01.04.10.16.22.25 u (atm cm)300K 0. 0. 002. 020.118. 466.150 4.03 p (mb) 5. 0 10. 0 20. 7 43. 1 91. 8 195 T (OK) 237 227 222 218 217 218 Table 4. 1. 3: Optical mass path parameters for N20 tangent paths. 4. 2 Odd Nitrogen Oxides 4. 2. 1 Atmospheric Distribution This group of nitrogen compounds includes NO, NO2, NO3 and N205. All occur naturally in the stratosphere and have recently been studied in detail as they are thought to play an important role in the ozone photochemistry. Concern has been expressed that the SST would increase concentration of odd nitrogen in the lower stratosphere, leading to a decrease in the ozone concentration.

A natural source of odd nitrogen in the stratosphere is the downward flux of nitric oxide (NO) which is formed by a number of photochemical reactions in the thermosphere (see, e. g., Strobel, 1971; Norton and Barth, 1970; Nicolet, 1970). In the upper mesosphere nearly all the odd nitrogen is in the form of nitric oxide but in the stratosphere substantial amounts are in the other oxides NO2, NO3 and N205 as well as nitrous acid HNO2 and nitric acid HNO3. These components tend to dissociate during the day and also take part in reactions with hydrogen and oxygen compounds (0, 03, OH, H202, etc. ). Calculation of the stratospheric photochemistry of these nitrogen compounds have been made by Johnson (1971) and Crutzen (1971). Such calculations are complex and contain many reactions whose rates are not accurately known. Unfortunately there are few measurements of stratospheric concentrations which can be used to verify the calculated distributions. When Crutzen's values of the nitric acid concentrations are compared with the measurements of Murcray et. al. (1969) or Lazrus et. al. (1972) it is found that the theoretical concentrations are at least an order of magnitude too small (see Section 4. 3. 3 for details). It throws into doubt the validity of the results for any odd nitrogen molecules, although it is possible that the HNO3 number densities are the only ones that are seriously in error and the predicted distributions of the other molecules are substantially correct. It would be extremely helpful to have more measurements, particularly of NO and NO2 concentrations, in the lower stratosphere. The agreement between the calculations of Johnston(1971) and Crutzen (1971) is reasonable. Usually the number densities agree to within a factor of two or three, which can be easily explained by the differences between the models and by the reaction rates and other parameters included in the calculations. The computations of Strobel (1971) extend down only to 30 km but again are close in the region of overlap. 80

Number density measurements of NO in the upper stratosphere and lower mesosphere (40 to 60 km) have been reported by Pontano and Hale (1970). Between 40 and 50 km the average value is close to 10 cm but decreases almost an order of magnitude near 60 km. The measurements were made about a half hour after sunset in September and December 1968 at White Sands, N. M. At the upper level they are about an order of magnitude less than the measurement of Pierce (1969) made at twilight. Those of Meira (1971) extend only to 70 km but in this range are an order of magnitude less than Pierce. The concern over the artificial introduction of odd nitrogen into the stratosphere by the SST is prompted by its ability to catalytically convert odd oxygen into even oxygen, primarily by the following two reactions: NO+ 03->NO2 + 02 NO + O - NO + 02 If sufficient quantities of odd nitrogen were introduced into the lower stratosphere it is conceivable that a marked reduction in the ozone concentration could result in the region of intensive SST flights. This would allow more ultraviolet radiation to reach the surface and possibly upset the energy balance of the stratosphere. Some calculations to determine effect of pollution by the SST have been made (for example, by Johnston, 1971) but they are inconclusive, largely because many of the parameters, such as production rates for odd nitrogen, reaction rates, horizontal and vertical diffusivity etc., are not sufficiently well known. It seems likely that the SST will not influence the ozone concentration greatly but this cannot yet be conclusively demonstrated. From the foregoing discussions it is apparent that there is a critical need for measurements of odd nitrogen concentrations throughout the 81

stratosphere, but especially in the lower stratosphere. They are needed to verify the theoretical models and to encourage the development of new models with more adequate transport phenomena incorporated in them. Only then can we be confident in the reliability of predictions about the polluted stratosphere. Finally, concentrations of nitrogen oxides, ozone etc. would have to be monitored on a continuing basis to verify the predictions and to assure that there were no major modification of the stratosphere. 4. 2. 2 Infrared Spectral Properties Nitric oxide (NO), being a diatomic molecule, has a simple infrared spectra. The fundamental (1-0) band is centered near 5. 3pm (1875 cm 1) the (2-0) near 2. 6um and the (3-0) near 1. 8pm. The band intensities, particularly that of the fundamental, have been measured by a number of workers and are summarized by King and Crawford (1972) and by Michels (1971). Michels describes a calculation of the intensities of the bands, which is in good agreement with experimental values for the fundamental. The averages of the measured intensities for the (1-0), (2-0) and (3-0) bands are 130, 2. 5 and 0. 1 cm (atm cm ) respectively. The Lorentz half-width have been measured by Abels and Shaw (1966) (mean value 0. 055 cm at 1 atmosphere and 300K) and the positions of the lines in the fundamental measured by James and Thilbault (1964). The parameters required to make accurate atmospheric transmittance calculations appear to be known. The survey of the spectral properties of the remaining nitrogen oxides has not yet been completed. 82

4. 3 Nitric Acid Vapor HNO3 4. 3. 1 Atmospheric Distribution Nitric acid vapor was first detected as a stratospheric constituent by Murcray et. al. (1969) who recorded solar absorption spectra with a balloon-borne infrared spectrometer. This constituent has not been observed in the troposphere partly because of interference by other molecules ( e. g., water vapor) and partly due to the lower concentrations in the troposphere. Even in the stratosphere the long paths associated with zenith angles slightly in excess of 900 were required, because of the low mixing ratio. The data of Murcray et. al. was rendered somewhat difficult to interpret because of lack of suitable laboratory spectral data, although this situation has now been largely rectified. The quantitative reduction of the balloon data by Rhine et. al. (1969) showed amounts of nitric acid vapor varying between 2. 0 and 23 x 103 atm cm, along a line of sight and corresponds to a mean value of 4. 41 (~ 0. 75) -3 x 10 (atm cm) per air mass, slightly lower than the value of Murcray et. al. -9 This can be converted to a volume mixing ratio of approximately 5 x 109 which applies to the 20-25 km region A confirmation of the presence of nitric acid has been provided by Lazrus et. al. (1972) who based their conclusion on the analyses of filter samples obtained during stratospheric balloon flights (21 to 27 km) and aircraft flights (17 to 19 km). The balloon data yield mixing ratios ranging from -9 1 to 5. 5 x 10 by volume i. e., comparable to those of Murcray et. al. at about the same altitude. The lower altitude aircraft measurements gave lower mixing ratios, ranging from zero to 0. 7 x 109 by volume 83

These changes of concentration with altitude are in accordance with the theory of the formation of HNO3. It can be produced by the reactions (see, for example, Johnson, 1971). N205 + H20 - 2HNO3 and NO2 + OH + M - HNO3 + M According to Johnstionthe second is much more likely and the maximum production occurs between 25 and 40 km. It can also undergo photodissociation HNO3 + hv - NO + OH X43000o and there may be substantial diurnal variations, at least at some altitudes. Of course, a comprehensive theory on the distribution of HNO3 in the stratosphere should include transport by both eddy processes and by the mean circulation. This has been pointed out by Lazrus et. al. who suggest that the reason for the small concentration in the lower equitorial stratosphere is due to the fact that this air is of recent tropospheric origin It is difficult to extrapolate upwards to estimate the nitric acid mixing ratio above the region where measurements exist. At sufficiently high levels photodissociation will predominate and daytime concentrations will be low. The calculations of Crutzen (1971) indicate the maximum mixing ratio occurs near 25 km at night and a few km lower during the day. However his calculated mixing ratios are lower than the measurements in the lower stratosphere by a factor of 5 to 10 at night and by a factor of 10 or more during the day. Crutzen discusses possible reasons for this discrepancy but does not reach any firm conclusions. There is a need for more measurements of the distribution of this interesting stratospheric constituent. 84

4. 3. 2 Spectral Properties of HNO3 HNO3 is a nearly planar molecule. Few studies of the quantative absorption of this molecule have been made and these have been prompted by the discovery of nitric acid in the stratosphere. The strongest bands are at 5. 9gm, 7. 5gm, 11. 3gm and 22gm (Goldman et. al., 1971). Those at 7. 5gm and 11. 3gm have been investigated by Rhine et. al. and all except the 22gm band by Goldman et. al. Because of the small amounts of the absorber along an atmospheric path it is difficult to make measurements for conditions similar to those in the atmosphere. The band intensities have been estimated by Goldman et. al. (Table 4. 3. 1) and the same authors have derived band parameters for the statistical band model from their data. No calculation of the wavenumbers and intensities of the individual lines have been reported. High resolution spectral studies would presumably be required to give the necessary input parameters. Potentially the most useful feature of the HNO3 bands is the intense Q-branches which give considerable atmospheric absorption for even a low resolution measurement. Band models are not usually very successful in Q-branch regions so that further measurements and development of the theory are required to extend our ability to predict atmospheric absorption at lower pressure and smaller absorber amounts. Band Intensity at 313K (cm-1/(atm cm)313K) 11. 3,m 600 7. 5gm 1150 5. 9,im 1250 Table 4. 3. 1: Band intensities of HNO3 (adapted from Goldman et. al. (1971) 85

4. 3. 3 Atmospheric Application Since nitric acid vapor has been detected from IR solar absorption by balloon-borne spectrometers, there is no reason why the same measurement cannot be made form a satellite. The height to which such a measurement can be extended is uncertain, reflecting our lack of knowledge of its distribution above the level of the present measurements. Absorption by a Q-branch would be the best choice in any spectral region. The result of a calculation using the day and night number densities tabulated by Crutzen (1971) are shown in Tables 4. 3. 2 (night) and 4. 3. 3 (day). These should be lower limits for the optical masses of HNO3 along a tangent path and should be increased by about an order of magnitude to agree with the measured concentrations in the lower stratospheres. This is illustrated in Table 4. 3. 4, which compares the optical mass of HNO3 calculated from Crutzen's day and night profiles with the measured amounts (Rhine et. al. ) at the experimental tangent heights. 86

Tangent Height (km) Number Density (cm-3) 50 45 40 35 30 25 20 15 10 1. 8 7. 0 5. 0 8. 0 9. 5 3. 9 1. 7 2. 0 0. 0 (3) (5) (6) (6) (7) (8) (9) (9) Optical Mass (atm cm)300K 9. 4 (-7) 7. 6 (-6) 1. 7 (-5) 1.4 (-4) 6. 5 (-4) 2. 8 (-3) 5. 0 (-3) 3. 9 (-3) p (mb) 1. 3 2. 4 4. 3 9. 7 20 43 83 108 T (OK) 269 252 241 228 223 218 217 217 Table 4. 3. 2: Optical parameters calculated from Crutzen (1971). Night profile of HNO3 Tangent Number Optical Height Density Mas s (km) (cm-3) (atm cm)300K (mb) 50 1.0 (1) 45 1.0 (2) 1.4 (-10) 1.3 40 2. 1 (3) 2. 9 (-9) 2. 4 35 8. 0 (4) 1. 1 (-7) 4.9 30 5. 0 (6) 6. 8 (-6) 10 25 4. 0 (7) 6. 0 (-5) 21 20 1. 1 (9) 1. 5 (-3) 46 15 1. 9 (9) 3. 9 (-3) 89 10 0. 0 3. 1,(-3) 120 Table 4. 3. 3: Optical parameters calculated from Crutzen Day profile of HNO3 T (OK) 265 251 237 227 222 217 217 217 (1971) 87

Tangent Height (km) 29. 7 28. 6 Calculated day 8.3 (-6) 1.6 (-5) 1.0 (-4) 1.0 (-4) 2.1 (-3) Optical Mass night 1. 6 (-4) 2. 4 (-4) 7. 4 (-4) 7. 4 (-4) 3. 5 (-3) (atm cm)3 OK measured 2. 2 (-3) 3. 5 (-3) 1. 1 (-2) 1. 0 (-2) 2. 5 (-2) 24. 6 24. 6 18. 8 Table 4. 3. 4: Comparison of optical masses of HNO, calculated from the profiles of Crutzen (1971) with the experimental stratospheric values (Rhine et. al., 1969) 88

References Abels, L. L. and J. H. Shaw, (1966), Widths and Strengths of VibrationRotation Lines in the Fundamental Band of Nitric Oxide. J. Molecular Spectros., 20, 11. Adel, A., (1938), Further Detail in the Rock-Salt Prismatic Solar Spectrum Astrophys J., 88, 186. Bates, D. R. and P B. Hays, (1967), Atmospheric Nitrous Oxide. Planet. Space Sci, 15, 189. Crutzen, P. J., (1971), Ozone Production Rates in an Oxygen-Hydrogen-Nitrogen Oxide Atmosphere. J. Geophys. Res., 30, 7311. Goldman, A., D. G. Murcray, F. H Murcray, W. J. Williams, T. G. Kyle, and J. N. Brooks, (1970), Abundance of N20 in the Atmosphere between 4. 5 and 13. 5 km. J. Opt. Soc. Am., 60, 1466. Goldman, A., T. G. Kyle, and F. S. Bonofno, (1971), Statistical Band Model Parameters and Integrated Intensities for the 5. 9p, 7. 5u and 11. 3u Bands of HN03 Vapor. App. Optics, 10, 65. Goody, R., (1969), Time Variations in Atmospheric N20 in Eastern Massachusetts. Planet. Space. Sci., 17, 1319. Gray Young, L. D., (1971), Calculations of the Partition Function for 14N2160. J. Quant. Spectros. Radiat. Transfer, 11, 1265. Gray Young, L. D., (1972), Relative Intensity Calculations for Nitrous Oxide. J. Quant. Spectros. Radiat. '.Transfer, 12, 307. James, T. C. and R. J. Thibault, (1964), Spin-Orbit Coupling Constant of Nitric Oxide. Determination from Fundamental and Satellite Band Origins. J. Chem. Phys., 41, 2806. Johnston H., (1971), Reduction of Stratospheric Ozone by Nitrogen Oxide Catalysts from Supersonic Transport Exhaust. Science, 173, 517. Junge, C. and J. Hahn, (1971), N20 Measurements in the North Atlantic. J. Geophys. Res., 76, 8143. King, W. T. and B. Crawford, (1972), The Integrated Intensity of the Nitric Oxide Fundamental Band. J. Quant. Spectrosc. Radiat. Transfer, 12, 443. LaHue, M. D., J. B. Pate and J. P. Lodge, (1970), Atmospheric Nitrous Oxide in the Humid Tropics. J. Grophys. Res., 75, 2922. Lazrus, A. L., B. Gandrud and R. D. Cadle, (1972), Nitric Acid Vapor in the Stratosphere. J. Appl. Meteor., 11, 389. 89

Meira, L. G., (1971), Rocket Measurements of Upper Atmospheric Nitric Oxide and their Consequence in the Lower Ionosphere. J. Geophys. Res. 76, 202. Michels, H. H., (1971), Calculationof the Integrated Band Intensities of NO. J. Quant. Spectros. Radiat. Transfer, 11, 1735. Murcray, D. G., T. G. Kyle, F. H. Murcray and W. J. Williams, (1969), Presence of HNO3 in the Upper Atmosphere, J. Opt. Soc. Am., 59, 1131. Nicolet, M., (1970), The Origin of Nitric Oxide in the Terrestrial Atmosphere. Planet. Space Sci., 18, 1111. Norton, R. B. and C. A. Barth, (1970), Theory of Nitric Oxide in the Earth's Atmosphere. J. Geophys. Res., 75, 3903. Pierce, J. B., (1969), Rocket Measurement of Nitric Oxide between 60 and 96 Kilometers. J. Geophys. Res., 74, 853. Pliva, J., (1968), Molecular Constants of Nitrous Oxide, 14N 16 J M Spectros, 27, 461. Pontano, B. A., and L. C. Hale, (1970), Measurement of an Ionizable Constituent of the Lower Ionosphere Using a Lyman-Alpha Source and Blunt Probe. Space Research, 10, 208. Rhine, P. E., L. D. Tubbs and D. Williams, (1969), Nitric Acid Vapor Above 19 km in the Earth's Atmosphere. Appl. Optics, 8, 1500. Schiitz, K., C. Junge, R. Beck and B. Albrecht, (1970), Studies of Atmospheric N20. J. Geophys. Res., 75, 2230. Seeley, J. S., and J. T. Houghton, (1961), Spectroscopic Observations of the Vertical Distributions of Some Minor Constituents in the Atmosphere. Infrared Phys., 1, 116. Shaw, J. H., G. B. B. Sutherland and T. W. Wormell, (1948), Nitrous Oxide in the Earth's Atmosphere. Phys. Rev., 74, 978. Strobel, D. F., (1979), Odd Nitrogen in the Mesosphere. J. Geophys. Res., 76, 8384. Tejwari, G. D. T. and P. Veranasi, (1971), Theoretical Line Widths for N20 -N O and N20-Air Collisions. J Quant. Spectrosc. Radiat. Transfer, 1, 1659. 90

Chapter 5. Atmospheric Aerosols 5. 1 Introduction This initial study of the feasibility of determining the characteristics of atmospheric aerosols by satellite remote sensing techniques suffers from our lack of knowledge of the physical (including optical) characteristics of aerosols, from some imperfections in theories of atmospheric radiation processes and from a lack of precision in possible methods of inversion of radiation measurements to aerosol parameters. Nevertheless, the existing data and theories are reviewed in an attempt to indicate spectral regions and techniques which may possibly be used to determine atmospheric aerosol characteristics. Our knowledge of physical and optical properties of aerosols are reviewed in the next section. Some remote sensing techniques that have already been used are then briefly described. Finally two techniques for the remote sensing of stratospheric aerosols from satellites are considered and their feasibility and practicability are discussed. 5. 2 Survey of Experimental Data on Atmospheric Aerosols The fundamental parameters needed for a study of remote sensing of atmospheric aerosols are the physical characteristics of size distribution, number density and index of refraction. This survey is an attempt to provide a suitable but not exhaustive summary of information on these parameters. 91

5. 2. 1 Sources of Atmospheric Aerosols Atmospheric aerosols are both man made and formed by natural processes, i. e.: a) Mechanical processes (including winds) disrupt and disperse matter from the earth's surface. b) Trace gases react chemically in the atmosphere producing aerosol particles. c) Smoke formed by industrial and agricultural combustion conveys aerosols into the atmosphere. d) Vapors produced by condensation and sublimation rise into the atmosphere. e) Extra terrestrial matter falls into the atmosphere. Fig. 5. 1 (Hidy and Brock, 1970) is an interesting summary indicating the origin and nature of particles of various sizes. Fig. 5. 2 (Junge, 1963) indicates nomenclature often used to describe atmospheric aerosols. The smallest particles (Aitken particles), of size r - 0. lpm, are not very different from molecules in their optical properties. The "large particles", from 0. lpm to l1m in radius have optical properties which show strong dependence on size and shape (Rozenberg, 1966). The giant particles, with r >-1, consists mainly of cloud and fog particles and mineral dust. The form of the atmospheric aerosol is modified considerably with the passage of time. Absorption of water vapour, coagulation, sedimentation, rainout, washout, attachment and chemical reactions tend to reduce the number of very small and very large particles. Distribution processes (transport and diffusion) transfer particles from place to place until removal from the atmosphere. Fig. 5. 3 (Junge, 1963) shows theoretical changes in size and volume distribution of continental aerosols due to coagulation. 92

I I I - - - I Rain Po Tobe cco *oSke Ground stone i "i250 concentrated iut I Spray dried dtergent I [ F2kNtNSZCarbon blacki;STS igments I S N ^H4Clt S"NFU — i *cular r 1,i I Pulverized coal Spray dried milk ' Puvrzdca Oil s1oke Rf Stoker f ly ash > fyI I I I Pulverized coal fly ash SS\\sSS\^S Fopundry dusts:>Z S\^ZZ '~N. Zinc Oxide fum II Ceaent dust i Alkali fume i I T I _1 - -- _ Metallurgical dust I I I Pollens [ i \ \, fiMet-allurg-ical fuwe "IZ$ Bacteria I Virus and protein I I Metallic Debbris in rocktr Household nd agricultural sprays ^ s -I msete qri te P.s a -- -a- M etoi I I I I I 10l 102 10 1 10-1 10-2 micrometers Fig. 5.1 Size range of atmospheric particles. (From Hidy and Brock, 1970). NOMENCLATURE A'Aitken Large Giant ~~~~NOMENCLATURE _ particles particles particles Air r smaoll Large o Electricity ion ions 1 Atmospheric /Haze aCE Optics _ortc o3 -.- u..- - --- -—. -Act rei o L Cloud < ondensot n-o Physics Condensation et: ~ Physics Nuclei / S. E.Particles Which A Contain Main Chemistry "_Aerosola Moss iO-3 10-2 10 ',t radius 100 ~ 0' o0' Fig. 5. 2 Nomenelature of nautral acrosols and the improtance of particle sizes for various fields of meteorology (from Junge, 1963). 93

10-' r10,i radius -- Fig. 5. 3 Calculated change in the size and volume distribution of continental acrosols due to coagulation. h=hours, d=days (from Junge, 1963). 94

5. 2. 2 Number Densities Surface number densities of aerosols are highly variable and 3 usually in the range of 100 to 500 particles per cm. Concentrations as high 3 as 20, 000/cm may be found in industrial areas. Concentrations are assumed to diminish approximately exponentially with height for the first few kilometers. At the tropopause (at mid latitudes) number densities of a few particles per cm are generally found. In the lower stratosphere (at about 20 km) there is a maximum in the aerosol particle distribution. Fig. 5. 4 (Chagnon and Junge, 1961) shows this increase for particles larger than about 0. 1Am radius. This stratospheric aerosol layer has a latitude dependent number density as indicated in Table 5. 1 from Rosen (1968). Number Density Location -3 25 cm Panama 8 Minneapolis 2 Ft. Churchill Table 5. 1: Latitude dependence of stratospheric number densities (Rosen, 1968). Since the altitude of the altitude of the tropopause decreases with latitude, the altitude of the aerosol layers decreases accordingly as shown in Fig. 5. 5 (Newell and Gray, 1972). Additional well defined layers exist at the level of about 25 km (nacreous clouds) and at the 80 km level (noctilucent clouds). Aerosol layers have been observed near the 40-50 km level also. They often form within temperature inversions. 95

2 V a 0 B.6. 30 r^a, 0-n t' l *21 Nov 1959 * 9 Dec 1959 25- 0 27 June 1960 2(?) 0U^ I O 25 Sept 1960 A 23 Oct 1960 20 15 __ _...._ 25 Sept 60 15 --- — \ etC — 27 June 60 Tropopauses,- -_- 9 Dec 59 '~p 1' ---- 21 Nov 59 10.............. 5 t I. I Ii1ls1 I I, I I1l,, I I,.... \' \~~~~~~4 - 10-1 100 101 Particle concentration, cm3 Fig. 5. 4 Vertical distribution of stratospheric particles collected by inertial impactors, according to Chagnon and Junge (1961). Tropospheric measurements of other investigators are included for comparison. 96

E Volz I -Rosen Kent w ~C Elterman M 7oyen Grams Rose r Grams *Rosen 10 0 10 20 30 40 50 60 70 LATITUDE (ON) Fig. 5. 5 Height variation of tropopausal aerosol layer with latitude from(Newell and Gray, 1972). 97

5. 2. 3 Composition of Stratospheric Aerosols The stratospheric aerosol layer contains a large portion of sulfate particles. In addition, it contains nitric acid vapor, chlorine, bromine, silicates and other substances. It is believed that sulfate particles are formed within the stratosphere itself by oxidation of the trace gaseous hydrogen sulfide (H2S) and sulfur dioxide (SO2) that enter the atmosphere by vertical mixing. During volcanic eruptions gaseous and solid material is thrown up into the atmosphere, sometimes as high as the stratosphere. Some of this material is sulfurous. The following summary (Table 5. 2) of possible stratospheric aerosol constituents is taken from Remsberg (1971). (NH4)2SO4 A H20(ice) B H202 A Na2SO4 B H2S04 A NaHSO4 B HN03 A NH4HS04 B NOHS04 A NH2SO3NH4 B HNO3-H2SO4-H20 A NH2S03H B NaCl B S02 B NH4C1 B H2S B H2C03 B C12 B H2S03 B NOCi B HC1 B (NH4)2S208 B HN02 B SiO2 B NH40H A abundant or more probable constituents B minor or unlikely constituents Table 5. 2: Possible constituents of stratospheric aerosol (Remsberg, 1971) 98

5. 2. 4 Size Distribution A variety of aerosol size distribution forms have been used in the literature. Among these are: a) The power law distribution: n(r) = ar (5-1) b) The logarithmic Gaussian or log normal form: dn(r) = No exp -A.log ( ) dr r (5-2) c) Diermendjian's (1964) modified / distribution: n(r) = ar exp(-br ) (5-3) where r, ro = particle radius, cm 3 n(r), No = no/cm a, A, b,, r = constants A convenient summary of experimental data on aerosol size distributions is that of Newell and Gray (1972). Distribution types commonly found in the literature include power law, bimodal, and Gaussian logarithmic forms. The power law size distribution in the form m(r) = cr, where r is particle radius, is ubiquitous. Values of the exponent 4o range from 1. 5 (Rosen, 1968) to 7 at 51 km altitude (Elliot, 1970). More common values seen in theliterature are in the 3-to-4 range, as found, for example, by Newkirk (1964), Rosen (1968) (under certain conditions), Mossop (1965), Carnuth (1970) and Pueshel (1967). It has been suggested (Newkirk, 1964) that aerosol layers actually represent the addition of large numbers of small particles to a normal background of larger particles. If this is true, one would expect to see an increase in the effective of within layers, and such an effect has been observed by Rowen (1968). 99

Friend (1966) and Quenzel (1970) have proposed that the lognormal distribution is more representative of true populations. Others, notably Mossop (1965), Sherwood (1967) and Pilipowskyj (1968) have found that the correct distribution form is bimodal, with one peak in the 0. 01 region and the second peak near 0. 2W. Storebo (1970) has derived such a distribution on a theoretical basis. Mean particle radius estimates for stratospheric aerosols have been made by Bigg (1970), who found F = 0. 1 to 0. 4X, and Mossop (1965), who found r = 0. 1/. The diversity of these results indicates a need for extensive further study of size distributions. Each of these measured distributions may have been representative of the aerosol population at the particular time and location of the measurement. " Recent results are those of Blifford and Ringer (1969). A total of 329 size distributions, obtained from measurements of atmospheric aerosols collected at altitudes from the surface- to- 9 km were considered. The authors speculate that the global aerosol may be described by three overlapping log normal distributions which have means: a) in the submicron or pollution aerosol range b) in the Mie scattering range where atmospheric modification processes are relatively inefficient c) in the large particle (rl. Opm) range which is dominated principally by soil dust and biological particles and near the ocean surface by sea salt. In another paper De Luisi, et. al. (1972) fitted data to Diermendjian's / distribution function with =1. The initial data are given in Fig. 5. 6. Table 5. 3 contains the derived constants and Fig. 5. 7 shows curves of n(r) using the constants of Table 5. 3 100

TABLE 5. 3 Constants for Deirmendjian's (1964) Modified Distribution Function (Equation 1) for a = 1. 0 ( from de Luis i, et. al., 1972) Volume y a b Concentra- tion, cm-' S PO DV All S PO DV All S PO DV All 0.1 0.6512 0.3840 0.6289 0.4081 4.440 9.662 (+1) 2.721 8.099 (+1) 4.677 7.963 4.093 7.821 0.2 0.3432 0.3729 0.3543 0.3153 1.423 (+3) 3.118(+2) 5.481 (+2) 2.783 (+3) 1.024 (+1) 8.499 9.125 1.090(+41 1.0 0.0908 0.3840 0.2393 0.2458 2.431 (+17) 1.504 (+3) 1.677 (+6) 5.488 (+5) 4.206 (+1) 8.504 1.616 (+1) 1.483 (4-1, 3.0 0.0502 0.3877 0.2504 0.2346 8.271 (+32) 4.372 (+3) 4.649 (+6) 3.727 (+6) 7.688 (+1) 8.479 1.623 (+1) 1.570 (+1 10.0 0.0501 0.3988 0.2652 0.2309 3.586 (+33) 1.153 (+4) 8.440 (+6) 1.670 (+7) 7.718 (+1) 8.235 1.568 (+1) 1.602 (+l1 The following legend defines the locations where the measurements were made: S is Scottsbluff, Nebraska (125 measurements); PO is the Pacific Ocea:. (82 measurements); DV is Death Valley, California (81 measurements); and All is a combination of the three locations (288 measurements). The number.n in parentheses are the powers of 10 by which the associated constants are multiplied. 0

SCOTTSBLUFF OCEAN O. 0. cI - c c DEATH VALLEY ALL LOCATIONS 0 E\D 6 -c 10 r(/) 10' a Fig. 5. 6 Size distribution curves generated by linear regression for particle concentrations 0.1 (lowest curve), 0. 2, 1. 0, 3. 0, and 10. 0 cm(from de Luisi, et. al., 1972). IO 1l1 10o-2 10-1 loo r() r(I) 101 Fig. 5. 7 Curves of n(r) using the constants from Table 1. The dashed curve represents Deirmendiian's haze M(a=5. 333 X 104, b=8. 9443, y=l. 0) (from de Luisi, et. al., 1972).

Remsberg (1971) in his dissertation on the "Radiative Properties of Several Probable Constituents of Atmospheric Aerosols" used three models of aerosol distributions for stratospheric particles: a) The bimodal size distribution used by Pilipowsky, et. al. (1968) to explain lidar and radiometersonde observations: dNl(r) = 2.31 vNl - d(log r) \rlmin 0.1- r —lOtum (5-4) where v r1 min N1 = 2.5 =. l1m 5. 2 particles/cm dn2(r) 2. 5 108 d (log r) 0. 003-. r- 0. Olm (5-5) b) The bimodal distribution that deBary and Roessler (1970) used to explain scattered radiation from a dust layer. The distribution is defined by three functions: _dN(r)I = 10-0. 474 -3. 82 d (log r d (log r) 0. 3 'r lzAm (5-6) dN(r) d(log r) +3. 719 +4. 193 = 10 r 0. 1- r 0. 3a m (5-7) dN(r) = 10+5 d (log r) O. 03_r-0. 05,m (5-8) c) A modified dN(r) d(log r) The three distributions Dermendjian "haze M" model: 0. 55 8 2 -22.9r = 5.63 ~ 10 r e are plotted in Fig. 5. 8. 103 (5-9)

1o'~ I "A41fs (imc'ole ) Fig. 5. 8 Aerosol size distributions WP (Pilipowakyj, et. al., 1968), DR (de Bary and Roosaler, 1970). R (modified Deirmendjian haze "M" distribution ) (from Remsberg, 1971). 104

5. 2. 5 Optical Characteristics of Aerosols a) Index of Refraction of Aerosols Knowledge of the complex index of refraction of aerosol particles is necessary for investigating the visible and near infrared radiation effects of aerosols. In the visible and near infrared investigators have usually assumed that aerosols are non absorptive and have real indices in the range of 1. 33-1. 5. The value of 1. 33 is associated with water spheres, 1. 42 is the real part of the index of refraction of hygroscopic sulfate. The complex part of the refractive index is very important for the remote sensing of aerosols. Although, as noted above, it has generally been assumed that the absorptive part of the index is zero for aerosols in the visible and near infrared, recent measurements indicate that this may not be true. Bullrich (1969) has found a wavelength dependent absorptive index for hygroscopic sulfates, varying from 0. 033 at 4000R to 0. 055 at 9200R. Data on the infrared refractive index of atmospheric aerosol substances have recently been presented by Volz (1972). The real part nand the imaginary (absorptive) index n' are given for dust, aerosol water solubles, sea salt and water for the wavelength range 2. 5 to 40m. The results are shown in Figs. 5. 9, 5. 10 and 5. 11. Remsberg (1971) has published absorption and refractive index data for aqueous solutions of ammonium sulfate and nitric acid for the spectral range of 7. 5 to 14um, within an instrument resolution of -1 1-3 cm. The data are shown in Figs. 5. 12, 5. 13, 5. 14 and 5 15. Data on crystalline ammonium sulfate are given in Figs. 5.16. (Chermack, 1970). 105

SMw 300 0 2000 uO cm0 0 o. 0 S0o 300 PROPANE SOOT' ---- - - - - - -, -o CHIMNEY SOOTY.003:."~ —1-'' - - - 0 2 4 6 8 10 12 14 20 30 z z. I FROM PRECIPITATION 0 DUS^m.01 -PR. Fig. 5. 9 Absorption index vs wavelength of dust and soot as calculated from transmittance measured by the potassium bromide disk technique. (from Volz, 1972),a 500 200D 1U 0 0 ODD 5000 1.0 I —. --- t *,C.I --- h -- r- --.' '.'.' --- 1 --- f -- r — -, ---. *, SOD.03 *3 /~ AEROSOL..- " wc /B t cWATER SOLUBLES (/ Z2 000 2000 150 20 00 80 003.003 0 2 4 6 8 10 12 14 20 40 A X(pm) Fig. 5. 10 Absorption index of water soluble aerosol substance, typical for continental rain (sample M) and continental and arctic snow as well as tropical rain (sample 270), and sea salt(from Volz, 1972) 2.2 v (cm" 1.8 - X cooarse T"thick _ ] S )'Coarse DUST. Q 1.0 ^ ez - 7 ~ t I 15 40 r 2.6 t1 b 2.2 I 4< ',,ze2. Jl/~.2 z W 1.8 - I 0 1.4 c. sR WATER SOLUBLE.03 ~ nR) AEROSOL.01 1.0 I * ~.- 02 Na~l. c) - NoCI+,,..9 1TER..-.... - c) 1.0 3 5 7 9 II 13 15 3040 X (/m) Fig. 5.11 Real part of the refractive index of dust from precititation (a), water soluble aerosol substance (b), and water and sea salt (c). Dotted line line on curve M near 9 gm: only reflection considered (from Volz, 1972). 106

Wavelength (microns)? T.5 Wavelength (microns) It I 8.s S 10 Ia 13 14 2.0 1.6 1.4 0 c 1.4, c o c 1.2 0 4,0 -t.0 c o U 0.4 0.4 0.2 0.0 I a n a I 2.0 I.e 1.6 X 1.4 o C 1.2 o c 1.0 0 0.4 0.t AO \ 75 % HASO4 k -. a I. 75% HSO, / ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ k - 1100 i6oo 1800 1400 Frequency (cm'l) 0soo 1200 1000 too Frequency (cm*t) too 700 Fig. 5.12 Optical constants for 75% H2S04 (Remsberg, 1971)

Wavetength (microns) Wavelength (microns) 11 7?7.5 8.5 * 10 12 13 14 2.0 c 0 c WI C 0 I-'c 0 O. 0 U 0 'c C 0 c 0 c C 0 a U 0 0 _.0 0.0 ~ _ ~ _......... 1600 1500 1400 1300 1200 Frequency (cm' ) 1150 1100 1000 00 oo 00 Frequency (cm'1) 700 (a) (b) Fig. 5.13 Optical constants for 90% H2S04 (from Rensberg, 1971)

601 *,, o ~ N 0 0 CO N ~. N C1 Hf C 3 _ 3 0 0 N - D s cr o Optical Constant (n and k ) o 0 o o o.- - -.? N 0o tN: 0 w 0 F o a o II I p 0 9* J 3 0' " 3 N o 4u* 4

Wavelength (microns) 7 7.5 Wovelength (microns) 10 II 6.5 S 6.5 12 13 14 0JS 0 C c 0 r,c 0. 0 4,0 -o O 0 ic la c c t. C 0 c 0 U 0 2.0 1.e 1.6 1.4 1.2 1.0 0.e 0.6 0.4 0.2 0.0 II11 n 39 % (NH4)SO, k _ 50 1600 1500 1400 Frequency (cm') 1300 1200 1100 1000 t00 Frequency (cm' ) 800 700 Fig. 5.14 continued.

Wavelength (microns) 10 10.5 11 11.5 9.5 12 12.5 o c c a a C,, C c 0 0 -U3 c 0 0.) (a. 0 2.0 - 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 4 0.2 I 0.0L n 68 % HNOa k - 1100 1000 900 800 Frequency (cm'1) Fig. 5.15 Optical constants for 68% HNO3 (from Remsberg, 1971) 111

0 0 0 1.4 o 1.2 1.0.2 8 4D0 60 7.0 8.0 9.0 10.0 X(~) Fig. 5.16 Optical constants for crystalline ammonium sulfate, Chermack (1970). 112

b) Extinction Coefficients The basic equation of radiative transfer which describes the attenuation of a beam of radiation of intensity IX is: dIX = -Iext(X) ds (5-10) with r2 2 A dnkr' ex / r2 Q ex (r AXn) dn(r) Pext Q/ ^cextlr, dr (5-11) r Q,=Q Qab ext sca abs where Qext Qsca Qabs are efficiency factors determined by Mie scattering theory as a function of particle size r, wavelength x and complex index of refraction n. Techniques for calculating these factors are given in Deirmendjian (1961) In the UV and visible, it is usually assumed that extinction by aerosols is only due to scattering and that absorption is negligible. However, as noted above Bullrich has found absorption effects for hygroscopic sulfates. A typical curve of aerosol extinction at 5500R (Newell and Gray, 1972) is shown in Fig. 5. 17. In the infrared Qscat is normally negligible compared to Qabs. The equation for the attenuation of a beam of radiation could then be written: dI = - I g (X)ds (5-13) Emitted ratiation would be given by: dIX = abs() B(T, ) ds (5-14) where B(T, X) is the Planck function at temperature T. Curves of jext= abs by Remsberg are shown in Figs. 5. 18, 5. 19, 5. 20 and 5. 21. His calculated emitted intensities for a 1 km path are shown in Figs. 5. 22, 5. 23, 5. 24, 5. 25 and 5. 26. 113

0. - - 4 i i i 10 10 -. 10 -2 W —4 -7 k() -J 0 cn 0 0o LI 10 10 -4 XO 10 -6 10 -7 0-5 km Elterman (1970) 5-35 km E1terman (1966) 35-80 km Exponential (0.57 x Rayleigh @5500AY 9.,,r~,99. ~9~~~ V 9% 9 *9 10 -* -r 0 10 20 30 40 50 60 70 0 ---r sO ALTITUDE (KM) Fig. 5.17 Reference and Gray, Aerosol Extinction Model k=5500 R (from Newell 1972). 114

Wovelength (microns) 0 9 7 10 II 12 5 ~ _I~ I,~.~.~ I.14.12 E,ic JO (WP)max disbrlbution 90% H2SO, 9O~/ I a I I I.08.06 I.04.02 t 1800 1600 1400 1200 Frequenty -(cm") 1000 100 00G Fig. 5.18 Extinction coefficient for 75% H2SO4 (from Remsberg,1971) 115

Wavelength (microns) e 9 ~ 7 10 II e2.12 E c.10 F x.-.08 en (W-P)max Distribution 75% H2SO4! I a.06 -.04.02.00 c IO00 1600 1400 1200 Frequency (cm") 1000 800 600 Fig. 5.19 Extinction coefficient for 90% H2S04 (from Remsberg, 1971) 116

Wavelangth 10 (microns) 11 9 12.10 E 0#4...08.06 X.04 caL.02.00 1100 1000 900 800 Frequency (cm-') Fig. 5. 20 Extinction coefficient for 68% HNO3 (from Remsberg, 1971) 117

Wovelength (microns) 6 7.80.70.60 '.50 T se.40 *".30 20 CO..10.00 2100 1900 1700 1500 1300 1100 900 Frequency (cm"') Fig. 5. 21 Extinction coefficient for crystalline (NH4)2 SO4 (from Chermack, 1970)

J | I u, 2.5 00 Fig. 5. 22 9 X(,) one kilometer (from Remsberg, 1971) Radiances for an ice aerosol path of 8 IE 0 60 \11 t3 o ----- II Fig. 5. 23 Rediances for a 75% 2S04 Remsberg, 1971) aerosol path of one kilometer (from 119

15,IO I10 a E 5 0 40 NI o 20:L 0 0.,2 90% H2SO4 (V-P) max T=200~K 1 --- ~~~ ---I A 10 II 12 X(Zp) 13 14 Radiances for a 90% H2SO4 aerosol path (from Remsberg, 1971) of one kilometer Fig. 5. 25 Radiances for a crystalline (NH4)2SO4 aerosol kilometer (from Remsberg, 1971). path of one 120

8 6 I I F. I I-I /^~68 % (W-P) T/ \= 20( I HNO3 max 3~0 K 4, 2 0 I I I 004.00- - L - -- - 10 II X (u ) 12 Fig. 5. 26 Radiances for a 68% HN03 aerosol path of one kilometer (from Remsberg, 1971). 121

5. 3 Survey of Aerosol Remote Sensing Techniques Some information about atmospheric aerosols has already been obtained by remote sensing experiments. Generally speaking, these experiments could not provide quantitative information about particle size distributions, number density vs. altitude, or index of refraction. However qualitative information has been obtained. 5. 3. 1 Sunphotometer Monochromatic solar radiation measurements can be used to determine the extinction coefficient of the vertical atmosphere due to aerosols (the atmospheric turbidity). Many years of measurements have provided a climatology of turbidity over the U. S. These data have been used to estimate particle loading in the lower atmosphere (Flowers, McCormak and Kurfis, 1969). 5. 3. 2 Laser Radar Laser radars have been used to observe aerosol layer altitudes and relative scattering strength. With this technique, the intensity of the backscattered light from a laser beam is determined as a function of range. The presence of particulate matter causes significantly different returns than a dust free atmosphere does. Aerosol particles can therefore be detected. The scattering properties of the atmosphere are determined as a function of range (altitude) with a distance resolution as low as 100 meters. Results include: a) The determination of the dependence of the stratospheric layer on tropopause height. b) The time variations in aerosol scattering following the Mt. Agung eruption. 122

5. 3. 3 Twilight Scattering Measurements of twilight sky radiation yield a qualitative picture of the vertical distributions of aerosol particles (Volz, 1969, Shah, 1970). Vertical dust profiles generally agree with results by more direct methods. The effect of stratospheric dust from the Agung volcano was observed in Germany 2 months later. Highest amounts were observed about 9 months after the explosion, abnormally high amounts were observed even 5 years later. 5. 3. 4 Forward Scattered Sunlight The measurement of sunlight in the sun's aureole as a function of wavelength can yield information about the altitude and size distribution of aerosols (Green, et. al., 1971). Theoretical calculations of forward scattered sunlight are carried out assuming single scattering and analytical models of the altitude-size distribution of aerosols. Optimum fits to experimental data at several wavelengths are achieved by adjusting parameters of the aerosol model. 5. 4 The Remote Sensing of Aerosols from Satellites Two experiments which have been suggested for the monitoring of aerosol layers and the possible identification of aerosol constitutions are described below. In the first (Gray, et. al., 1971), a method of determining the height of aerosol layers from the inversion of horizon-intensity profiles of the earth's sunlit horizon is described. In the second (Remsberg, 1971), the possibility of the identification of aerosol constituents from infrared emission spectra is discussed. 123

5. 4. 1 Aerosol Monitoring by Satellite Horizon Scanning In the MIT experiment (Gray, et. al., 1971) the earth's horizon is scanned by a satellite multispectral sensor. The sensor signals are compared to a theoretically derived intensity profile based on a reference model of aerosol extinction vs. altitude. The computer program corrects the reference profile according to the difference between measurement and prediction considering uncertainties in the reference model and the measurements. In the work carried out so far, simulated horizon profiles have been inverted with results shown in Figs. 5. 27 and 5. 28. Also experimental data obtained with a multispectral photometer on an x-15-1 aircraft were inverted to produce the aerosol attenuation curve shown in Fig. 5. 29. It is estimated that aerosol layers may be determined with an altitude accuracy approaching 1 km. 5. 4. 2 Identification of Aerosol Constituents Using Emitted Radiation from the Earth's Atmosphere Remsberg has compared his calculated emitted radiation curves with data obtained by Murcray, et. al. (1969). He has computed the emitted intensity that would be obtained from a 50 km horizontal path and compared this data (Figs 5. 30, 5. 31) with Murcray's data measured at zenith angles of 80 and 85~ (Fig. 5. 32). He concludes that the emission feature at 10. 8,m observed in Murcray's data can possibly be explained by the 11. Opm feature in the H2S04 data in Figs. 5. 30 and 5. 31 if a wavelength correction 4X=0. 24m is applied. The wavelength correction is necessary since the index of refraction of H2SO4 was measured at room temperature, whereas the atmospheric aerosols are at -60 to -70 C. Such low temperatures could easily cause such a wavelength shift in the absorption coefficient and emission spectra. 124

0 10" z 0 10'4 0 1(-4C3 S. IO o ib 0 4 So so0 40 ALTITUDE (KM) Fig. 5. 27 Simulation Results - Inversion of 20-km, y=3 Aerosol Layer, Zenith 300 (from Gray, et. al., 1971). to 10" ' U ' \ -~~~~~~J~ ~".,,10 O.,, a: 10 0 o0 10 0 30 40 so SO 70 sC ALTITUDE (KM) Fig. 5. 29 Experimental Results -Inversion of X-15 Data, aerosol extinction (from Gray, et. al., 1971). l, Iz X Li 0 co < ALTITUDE (KM) Fig. 5. 28 Simulation Results - Inversion of 50-km, 7=7 Aerosol Layer, Zenith 60 (from Gray, et. al., 1971). 125

40 30 90% H2S04, r\D-R E / \ T=200~ K 20 z \ 50 km I0 9 10 11 12 9 10 11 12 Fig. 5. 30 Emitted intensigy for 90% H2S04 aerosol horizontal path (from Remsberg, 1971). 13 from a 50Km 40 30 I E ' 20 i10 0 I- - I r 75% H2SO4 D-R T-200~ K,, I.,I.......a - 50km 9 10 II x (,) 12 13 Fig. 5. 31 Emitted intensity for 75% H SO aerosol from a 50Km. horizontal path (from Remsberg, 1971). 126

-*'~ " I It l} Il )1^ Szs12 3 20 23.2 km I 24.6 km. 0 10 I1 12 13 WevIneelgth (/) Fig. 5. 32 Radiances observed at various heights and at two zenith angles for the 25 Feb. 1969 flight of Murcray (1970). Dashed line represents the position of the emission feature at 10. 8'. 127

5. 4. 3 Discussion of Satellite Measurements of Atmospheric Aerosols The technique of horizon scanning in the visible with a multispectral scanner to determine the altitude of aerosol layers appears to be a feasible and practical experiment to carry out. The simulated horizon profile data already include consideration of earth albedo, sun angle and a correction factor for multiple scattering. The multiple scattering calcularions will, no doubt, be improved in the near future. The reference aerosol extinction model will also be improved as our knowledge of atmospheric aerosols improves. Thus the experiment is certain to be able to determine the height of aerosol layers. It is not certain that the visible horizon scanning data can be used to determine the physical characteristics of aerosols, however, it is possible that infrared horizon emission or absorption spectra might contain identifiable spectral features as illustrated by the work of Remsberg. Improved know-ledge of the composition of aerosols and the accurate determination of the optical characteristics, of aerosols at low temperatures are a necessity for the successful operation of this experiment. Initial versions of both of these experiments can be flown on high altitude balloons to provide initial tests of their feasibility. 128

References Bigg, E. K., A. Ono and W. J. Thompson, (1970), Aerosols at Altitudes between 20 and 37 Kilometers, " Tellus, 22, 550-563. Blifford, I. H., and L. D. Ringer, (1969), The Size and Number Distribution of Aerosols in the Continental Troposphere, J. Atmos. Sci, 26, no. 4, 716-726. Bullrich, K., (1969), Determining the Vertical. Distribution of Aerosol Particles in the Atmosphere with Radiative Optical Measureing Methods, Proceedings of the NATO Advanced Study Institute: Effects of Atmospheric Water on Electromagnetic Wave Propagation, D. R. May, ed. Univ. of Western Ontario, London, Canada, Aug. -Sept., 198-215. Carnuth, W., (1970), Aerosol Size Distribution at 700, 1800 and 3000 Meter Altitudes, J. Geophys. Res., 75, No. 15, 2999-3006. Chagnon, C. W., and C. E. Junge, (1961), The Vertical Distribution of SubMicron Particles in the Stratosphere, J. Meteorol., 18, 746-752. Chermack, E., (1970), Optical Constants of Ammonium Sulfate in the Infrared, PhD. Thesis, Graduate School of Arts and Sciences, New York University at Oswego. de Bary, E., and F. Rossler, (1966), Size Distributions of Atmospheric Aerosols Derived from Scattered Radiation Measurements Aloft, J. Geophys. Res., 71, 1011. De Luisi, J. J., I. H. Blifford, Jr., and J. A. Takamine, (1972), MIodels of Tropospheric Aerosol Size Distribution Derived from Measurements at Three Locations, J. Geophys. Res., 77, No. 24, 4529-4538. Diermendjian, D., R. Clasen, and W. Vieze, (1961), Mie Scattering with Complex Index of Refraction, J. Opt. Soc., Am., 51, 620-633. Diermendjian, D., (1964), Scattering and Polarization Properties of Water Clouds and Hazes in the Visible and Infrared, Appl. Opt., 3, 187-196. Elliot, D. D., (1970), Effect of a High Altitude (50 km) Aerosol Layer on Topside Ozone Sounding, Space Research XI, Proc. of the 13th Plenary Meeting of Cospar, Leningrad, 1970, Akademie-Verlag, Berlin, 1971. Friend, J. P., (1966), Properties of the Stratospheric Aerosol, Tellus, 18, No. 2, 465-473. Gray, C. R., H. L. Malchow, D. C. Merritt and R. E. Var, (1971), Aerosol Monitoring by Satellite Horizon Scanning, Paper presented at Joint Conference on Environmental Pollutants, Palo Alto, Calif., Nov. 8-10, 1971. 129

Green, A. E. S., A. Deepak and B. J. Lipofsky, (1971), Interpretation of the Sun's Aureole Based on Atmospheric Aerosol Models, Applied Optics, 10, No. 6, 1263-1279. Hidy, G. M., and Brock, J. R., (1970), The dynamics of Aerocolloidal Systems, Pergamon Press. Junge, C. E., (1963), Air Chemistry and Radioactivity, Academic Press. McCormick, R. A., and K. R. Kurfis, (1966), Vertical Diffusion of Aerosols over a City, Quart, J. Roy. Meteor. Soc., 92, 392-396. Mossop, S C., (1965), Geochimica et Cosmochimica Acta, 29, 201. Murcray, D. G., F. H. Murcray, W. J. Williams, T. G. Kyle, and A. Goldman, (1969), Vtriation of the Infrared Solar Spectrum Between 700 cm and 2240 cm- with Altitude, Appl. Opt., 8, 2519-2536. Newell, R. E., and C. R. Gray, (1972), Meteorological and Ecological Monitoring of the Stratosphere and Mesosphere, NASA Contractor Report CR-2094, Prepared by M. I. T. for Langley Research Center. Newkirk, G. and J. A. Eddy, (1964), Light Scattering by Particles in the Upper Atmosphere, 21, No. 1, 35-60. Pilipowskyj, S., J. A.. Weinman, B. R. Blemesha, G. S. Kent, and R. W. Wright, (1968), Investigation of the Stratospheric Aerosol by Infrared and Lidar Techniques, J. Geophys. Res., 73, 7553. Pueschel, F. R. and K. E. Noll, (1967), Visibility and Aerosol Size Frequency Distribution, J. App. Meteor., 6, No. 6, 1045-1052. Quenzel, H., (1970), Determination of Size Distribution of Atmospheric Aerosol Particles from Spectral Solar Radiation Measurements, J. Geophys. Res., 75, No. 15, 2915-2921. Remsberg, E. E., (1971), Radiative Properties of Several Probable Constituents of Atmospheric Aerosols, PhD. Dissertation, University of Wisconsin, Dept. of Meteorology. Rosen, J., (1968), Simultaneous Dust and Ozone Soundings over North and Central America, J. Geophys. Res., 73, No. 2, 479-486. Rozenberg, G. V., (1966), Twilight, a Study in Atmospheric Optics, Plenum Press, New York. Shah, G. M., (1970), Study of Aerosols in the Atmosphere by Twilight Scattering, Tellus, XVII, No. 1, 82-93. Sherwood, R. D. and J. P. Friend, (1967), Project Strap Final Report, NY00-3615-1. 130

Volz, F. E., (1969), Stratospheric Dust Striations, Bull. Amer. Meteor. Soc., 50, 16. Volz, F. E., (1972), Infrared Absorption by Atmospheric Aerosol Substances, J. Geophys. Res., 77, No. 6, 1017-1031. 131

Chapter 6. Discussion and Future Program 6. 1 Summary and Conclusions The results documented in this report have been obtained during the first six months of a study of the feasibility of determining stratospheric distributions of minor constituents and pollutants from satellite measurements of absorption during solar occultation. A major part of the effort has been expended in searching the literature and consolidating the information on the distributions and spectral properties of the molecules and aerosols. Only the infrared spectral region has been considered because absorption due to molecular oxygen and ozone in the ultraviolet is almost complete for stratospheric tangent paths. Ozone (and possibly also some other constituents) can be measured in the highest regions of the stratosphere and in the mesosphere. Such measurements using stellar occultation have already been made (Hays, et. al., 1972). The molecules considered and discussed in this report are 03, OH, H20, N20, NO, NO2, N205, HNO3, CO2, CO and CH4. The distribution and optical properties of stratospheric aerosols are also discussed. For some of the constituents the stratospheric distributions are well known (e. g., ozone) while for others virtually no measurements exist (e. g., oxides of nitrogen) so that estimates have to be made based on incomplete theoretical models incorporating chemical and transport effects. Despite their shortcomings, these models serve a useful purpose and provide an order of magnitude estimate of the concentrations of minor constituents. Similarly the spectral properties of some molecules are well known, particularly the diatomic and linear triatomic molecules, while for others measurements are few and the theoretical analysis is incomplete. Having estimated the distribution of a constituent in the stratosphere the optical mass along stratospheric tangent paths was determined using 132

the procedure outlined in the appendix. Using low or medium resolution laboratory measurements, the absorption for different tangent heights was determined for some of the molecules. In the most strongly absorbing regions the absorption was complete before the tropopause was reached, showing that for some molecules distributional information may be obtained using a relatively simple instrument. However, other constituents absorb only to a small degree (less than one percent) and more sophisticated instrumentation would be require' to detect their presence. This stage of the investigation is not complete and the reader is referred to the individual sections of the report for details on specific molecules. The use of laboratory absorption spectra to determine atmospheric transmittance requires interpolation over pressure and optical mass, which limits its accuracy. A more serious problem is the error introduced by temperature effects. Most laboratory measurements are made at room temperature while the mean temperatures along stratospheric absorption paths are as much as 800C or more lower. The intensities of lines, particularly those of 'hot' bands or with high rotational quantum numbers, may change radically and thus change the absorption considerably. The values of absorption given in the tables are not highly accurate but are sufficiently good for their intended purpose, to examine the feasibility of determining concentrations from solar occultation m easurements. For the reduction of satellite data more accurate absorptions would be required. These could be obtained by the 'line-by-line' technique or perhaps by band models, provided the relevant parameters, line wavenumbers, intensity and half-width, are known. This information is not yet available for all molecules. Even for the better studied molecules the temperature dependence of Lorentz half-width is uncertain, as recent low temperature measurements 133

indicated variations which are not in accord with either the Anderson theory or the classical inverse square root temperature dependence. The effect on atmospheric transmittances has been discussed by Drayson (1972). Further low temperature absorption measurements are required. 6. 2 Future Program The contents of this report indicate that the minor constituents of the stratosphere can be detected and measured from a satellite by examining the absorption of solar infrared radiation along a tangent path. The more abundant molecules can be detected using comparitively simple instrumentation while others would require a more sophisticated design. This work must now be extended to put these results on a quantitative basis to estimate accuracy and altitude ranges for the various constituents. Only when this has been accomplished will it be possible to objectively estimate the usefulness of thetechnique and compare it with other measurement schemes. Many practical details will have to be considered in this analysis. These include, but are not limited to, horizontal inhomogeneities in concentrations, fine structure in the vertical distribution profile, uncertainties in the temperature profile and in the spectral absorption of the constituents, errors introduced by approximations in the transmittance calculations, the effects of refraction and errors introduced by aerosol layers. It will be necessary to study inversion techniques, that is methods of determining atmospheric distributions from the solar absorption measurements. The results of the investigations will hopefully lead to specifications for satellite instrumentation and will point out any deficiencies in our knowledge of spectral properties of the constituents, so that laboratory investigations can be made to remedy the situation. Only with a careful analysis and attention to detail can the full potential of this type of satellite experiment be achieved. 134

References Drayson, S. R., (1972), Atmospheric Radiative Transfer by Carbon Dioxide. A. M. S. Conference on Atmospheric Radiation, Fort Collins, Colorado. Hays, P. B., R. G. Roble and A. N. Shah, (1972), Terrestrial Atmospheric Composition from Stellar Occultations. Science, 176, 793. 135

Appendix Given the vertical distribution of a minor constituent in the stratosphere, we wish to determine the optical mass along tangent paths at various altitudes in the stratosphere, as well as 'equivalent' pressures and temperatures. It was convenient to write a flexible computer routine to make such calculations. The main application was to determine the optical masses of the constituents considered in this report so that the feasibility of stratospheric measurements of solar absorption could be evaluated. Refraction effects are of some importance but they should not effect a feasibility study and were ignored, but for the reduction of satellite measurements they should be included. The atmosphere between 1 and 80 km was divided into 1 km thick spherical shells. The temperature, pressure, absorber concentration etc. in each shell was assumed to be the mean of the values at its boundaries. The geometry is illustrated in Fig. Al. Consider a tangent height of h km above the surface and consider the shell between K and (K+1) km above ho. The distance xK to the tangent point is given by 2 2 2 K = (a + h + K) -(a + h) K o o = K(K + 2(a + h )) where a is the radius of the earth. The path length in the layer, a xK = 2(XK+1 - xK) The optical mass U(ho) for the entire path is given by the sum over all layers U(ho) =. N(K) xK 0 K K where N(K) is the concentration in the K th layer. 136

SUN ~~~~~~~~~~, l Fig. A.1. Geometry of the solar occultation. 137

An 'equivalent' or mean pressure is given by 2 P(K) N(K) AxK P(h) = K,j N(K) A xK K and similarly a mean temperature is defined by Z T(K) N(K) A xK K T(h ) - Z N(K) A x K K The last two parameters are useful when applying the Curtis-Godson or similar approximation. A Fortran subroutine OMASS has been written to make the calculations. The arguments are N the concentration in molecules cm for IS = 1 or volume mixing ratio for IS = 2 UN the computed optical mass in molecules cm2 UA TM CM optical mass in (atm cm)300K PBAR mean pressure in mb TBAR mean temperature in OK The calculations are made for the U. S. standard stmosphere (1962). This may be varied by altering the data cards of AP and AT (pressure and temperature respectively). For each vector the subscript I refers to an altitude of I km for I=1, 2, 3,..., 80. To make the use of OMASS easier an interpolative subroutine INTERP was written. It interpolates linearly the quantities A(1), A(2),..., A(L) defined at heights Z(1), Z(2),.., Z(L) to give the values X(1), X(2),... X(80) at heights 1, 2,..., 80 km. If the interpolated value X becomes negative 138

at any height, it is set equal to zero. The purpose of INTERP is be able to specify the concentration at a few atmospheric levels rather than all 80 levels required by OMASS. A Fortran listing of the subroutines and a short calling program is now given. The input data is L, IS, Z, and A. If IS = 1 then the units of A -3 are molecules cm. If IS = 2 then A is volume mixing ratio. The program has been tested and run on an IBM 360/67 and has been used in the main body of this report. 139

REAL N DIMENSION N(80),UN(80),UATMCM( 80 ), PBAR(AR (80)T ( 80), Z ( 80 ) DIMENSICN A(10) FORMAT (2I20/(I8F10.3 ) READ (5,900) L, I S( Z(I).I=1 L), (A( I ) 1=1.L) CALL 'INTERP( Z,tAL,N) CALL OMASS(NtUNUATMCM,P AR TBAR, IS) WRITE(6,550) (I,N(I), UN( I ),tATMCM( 1),PBAR( I),TBAR( I ), 1=179) FORMAT( ' 1'( I5,4E14*4,F8S1 ) GO TO 100 END 140

SUBROUTINE OMASS(NUN,UATMCM,PBAR,T8AR,IS) REAL N DIMENSICN N(80),UN(80),UATMCM(80),PBAR(80),TBAR(80) DIMENSICN Z Z(80), AN(80), AP(80),AT(80),X(80) DATA Z/ 1t 2.,3,4.,5.,6, 7 8,t9. 1 t,11, 12. 13* 14.,1 5,16, 1 17*tl 8,t19,? 20.,21., 22.,23.,24.,25*,26 t27*.,28.,29.,30.,31.,32., 2 33.,34 t 35., 36., 37*. 38. 39*,40., 41., 42*.,43. 44.,45.,46.t 47*.48., 3 49.,50.,51.,52.,53.,54.,55.,56.,57.,58.,59.,60.,61.,62.,63. 4 64.,65. f,66.,67* 68,69.,70 71.,72. 73.,74, 75.,76.,77.,78.,79., 5 80 / DATA AP/ E45.6, 746o, 65 7 8, 577.5 5,505.4440. 8383. 033 5,285. 8 1 245*4, 20 8, 179, 3,153 3,131.O,112.0,t 95 92,8 1 82,69 95,59, 80, 251.13,4 3.75, 37.46,32.C9,27.52,23.62,20.28,17*43,14*99,12*90, 3. 11, 9 574,8.26,7. 13,6.17,535,4.65,4*04,3.52,3.07, 4 2.68,2.3 5,2. C6,1.81,15 8, 1.40,123, 1.09,.961,. 849, 5.750,. 662,.58 5,.516,. 455,.401,. 353. 311,. 273,.240, 6.210,.184,.161,.141,.123,.107,.0926,.0801,.0692,.0595, 7 * 0511, *0438.C374t*0319,.*C271,. 0229,. 0193,.01 62,. 0136,.0114/ DATA AT/ 278.4,271.9,265.4,258.9,252.4,245.9,t2395,233.0,226.5, 1 220.0,10*216,5, 2 217.1,218*.1219. 1,220.1,221.1,222.0,223.0,224*0,225*0,226.0, 3 227.0,228., 229,6,232. 3,235.1,237. 9,240* 7,243.4,t 246.2,248*9, 4 251. 7,2 545,257. 3260*,262.8,265 5,268.3, 270.7,27C.7,270 7, 5 270.7,270.7 270 5,t268*5,t266*6,6264.6,262.26,260.7,258*7, 2 56.8, 6 254,8,252. 8,249* 1,245.2,241.2,237 3, 233.4,229 5,225.6,221.7, 7 217*7,213.8,209.9,t206.0,202.1,198.2,194*3,190*4,1865t,182*6 / DATA RZ/6378* /,BOLlZ/1 38054E-19/ AN(80)=00 D0 100 1=1,79 100 AN(I )=(N( I)+N(I+1 ))*,5 IF (IS*EQ.l) GO TO 110 DO 102 1=1,79 102 AN (I )=AN( I )*API /(AT( I )*BOLTZ) 110 CONTINUE DO 200 1=1,79 A=2*1(RZ+Z( I)) K= 0 X( I )=0. DO 201 J=I,79 K=K+l 201 X( J+l )=SQRT(K*(A+K) PBAR (I )=0. TBAR( I )=0. UN( I )=O. DO 202 J=1,79 U=AN(J)*(X(J+l)-X(J) I UN1(I)=UN(I)+U PBAR(I )=PPAR(I)+AP(J) *J 202 TBAR( I )=TBAR( I )+AT(J )*U IF (UN( I).EQ*0.) GO TO 2110 PBAR( I )=PPAR(I)/UN(I ) TBAR(I )=TRAR( I)/UN() I GO TO 211 21C PBAR (I)=PPAR( 1-1 TBAR(I )=TBAR( I-1) 211 UN( I )=UN( I )*2.E5 200 UATMCM( I )=UN(I )*40873E-20 TBAR(80)=TBAR(79) 141

PBAR(80)=P AR (79) UN( 80)=C. UATMCM(80)=Oo RETURN END 142

SUBROUTINE INTERP(Z A, L, X) DIMENS ICN Z(l ),A( 1) X( 1 IJ=2 00 100 I=1,80 Y= I nO 110 J=IJL IF (Z(J)*GT.Y) GO TO l 1 CONT I NUE J=L IJ=J X(I )=( (Y-Z IJ -. ) *A( IJ)+( ( IJ)-Y )A( IJ -1 ) /( (IJ-Z( IJ- ) ) IF (X(I).LT.O.) X(I)=0. CONTINUE RETURN END 143

UNIVERSITY OF MICHIGAN 3 911111 15 02844 0496 3 9015 02844 0496