255 Bamburgh Cir., Apt. 1210,

Scarborough, ONT M1W 3T6

CANADA

October 1, 1990

Dr. R.D. Bojkov, Secretary,

International Ozone Commission,

c/o World Meteorological Organization

P.O. Box No. 2300

1211 Geneva 2

SWITZERLAND

Dear Rumen:

I was reminded by a letter from Hans Dütsch three weeks ago that IOC had assigned an action item to Bob Hudson and myself to come up with a final recommendation on the new ozone absorption coefficients to be used with the Dobson, and to set an effective date for their activation.

Enclosed herewith is a report with a number of recommendations as well at a Table 4 which lists the new absorption coefficients. I have taken the liberty of also providing new Rayleigh coefficients. By my agreement with Bob Hudson, he has two weeks after you receive this to contact you and violently wave the red flag, if he deems it necessary.

I am sending copies of this letter and the report to a number of members of IOC. I welcome comments, complaints, and suggestions. I am sure there will be some. Unfortunately, I have no telex and no fax and no answering machine on my telephone (416 - 499 - 4134). If no one answers after ten rings, there is no one home. If the line is busy in perpetuity, I am probably hooked up to the AES computer.

Sincerely, and regards to all,

C.L. Mateer

Encl.

Copies: DeLuisi Dütsch Hudson Kerr Komhyr Megie Miller

APPLICATION OF BASS-PAUR 1984 OZONE ABSORPTION COEFFICIENTS

TO OZONE MEASUREMENTS

WITH THE DOBSON OZONE SPECTROPHOTOMETER

IntroductionRecent measurements of ultraviolet ozone absorption cross sections have been reviewed by Hudson (1990). He concludes that the relative accuracy of these cross sections is of the order of one percent and that this is approaching the experimental limit. He finds that the absolute error is higher, with an upper limit of about 2.5 percent. For ozone measurements in the ultraviolet, he recommends the use of the Bass-Paur 1984 cross sections with the additional Barnes and Mauersberger (1987) temperature correction which will be described later in this report.

The Bass-Paur 1984 (BP84) cross sections extend from 245.018 to 339.981 nm, which is only half way through the band pass of the long Dobson D wavelength. The BP84 spectrum was extended by the NASA/SBUV work team out to 342.7 nm using earlier BP data, which Bass and Paur did not consider to be of sufficiently good accuracy to be included in their BP84 data set. For the purposes of this study, the cross sections were further extended to 344.9 nm, also using earlier BP data. The error implications of these extensions will be discussed later.

The BP84 data set consists of two parts, viz., the actual experimental data at various temperatures and a set of coefficients, derived by regression from the experimental data, which express a quadratic relationship between the cross sections and temperature. The "extended" BP84 quadratic coefficients have been used exclusively in this study.

Six subsidiary data sets have also been used, as follows:

1. The slit function data for the World Standard Dobson (No. 83) as determined experimentally by Komhyr (ca. Oct., 1982).

2. The extra-terrestrial solar spectrum compiled by Furukawa

et. al.(1966).3. The Rayleigh scattering cross sections recently calculated by Bates (1984).

4. The atmospheric temperature profiles in the

U.S. Standard Atmosphere Supplements, 1966and theU.S. Standard Atmosphere, 1962. (The latter is, for our purposes here, negligibly different from theU.S. Standard Atmosphere, 1976.)5. The standard ozone profiles compiled by Mateer, DeLuisi and Porco (1980).

6. The standard ozone profiles compiled by Bhartia

et. al.(1984) for use in the re-evaluation of NIMBUS 7 SBUV ozone profile data.The primary problem addressed in this report is the determination of the new set of ozone absorption coefficients that should be recommended for use for total ozone measurements in the WMO/Dobson world total ozone network. A secondary problem is the recommendation of coefficients for use in inversion algorithms for Dobson Umkehr observations.

Methods of CalculationThere are several ways to estimate the appropriate "effective" ozone absorption coefficients to be used for direct sun total ozone measurements.

The BP84 quadratic coefficient relationship is given by

A(L, t) = C0(L) + C1(L) * t + C2(L) * t^{2}(1)where A(L, t) is the absorption coefficient at wavelength L and temperature t (°C) and C0(L), C1(L), C2(L) are the quadratic coefficients at wavelength L.

The simplest approach is to weight the quadratic coefficients with the slit function weights S(L)

C0(L)= ∫S(L) * C0(L) * dl / ∫S(L) * dL (2)and similarly for

C1(L)andC2(L), where the underlining refers to slit average values for C0, C1, and C2 and to a nominal band-center wavelength for L.The second step in this simplest approach is to convolve appropriate temperature and ozone profiles to obtain an ozone-weighted mean temperature, viz.,

t= ∫t(P) * P3(P) * dln(P) / ∫P3(P) * dln(P) (3)where P is atmospheric pressure and P3 is ozone partial pressure. Then we have

A(L)=C0(L)+C1(L)*t+C2(L)*t(4)^{2}where

A(L)is the slit-averaged ozone absorption coefficient. In this simplest of approaches, the effective mean temperature is the same for all of the Dobson wavelength bands.The next level of complexity may be introduced by including the extra-terrestrial solar flux F(L) as an additional weighting factor in eqtn (2). A further level of complexity is to calculate an ozone-weighted effective mean absorption coefficient, with or without the solar flux weighting, as follows

A(L)= ∫∫A(t, L) * F(L) * P3(P) * dln(P) * dL / X∫F(L) * dL (5)where X is the total ozone.

The above might be called "static" approaches, in that transmission through the atmosphere is not considered before the integration across the slit function is performed. In what might be called the "dynamic" approach, we calculate

Q(L)= ∫F(L) *S (L) * exp{-mu * ∫ A(t, L) * P3(P) * dln(P) - mu * B(L)} * dL (6)where mu is the relative slant path of the sun's rays through the atmosphere and B(L) is the Rayleigh scattering coefficient applicable to the entire atmosphere. The effective absorption coefficient may then be defined through the following series of equations

W(L)=Q(L)/F(L)= exp{-mu * [A(L)* X +B(L)]} (7) whereF(L)= ∫S(L) * F(L) * dL / ∫S(L) * dL (8)B(L)= ∫S(L) * B(L) * dL / ∫S(L) * dL (9) This givesA(L)= -{mu*B(L)+ ln[W(L)]} / {mu * X) (10)This last method would appear to most closely approximate what happens in the atmosphere.

Temperature Corrections to the Bass-Paur Cross SectionsBass and Paur normalized their absorption cross sections to the 253.7 nm Mercury line value of Hearn (1961), viz., 114.7E-19 cm

^{2}. It was assumed that there was no temperature dependence of the ozone cross section at this wavelength.Over the temperature range from 195 to 335°K, the Barnes/Mauersberger results may be expressed by the rectangular hyperbola

A(t,253.7) = 114.97 - 78.49/{87.3 - t} (11)in units of 10

^{-19}cm^{2}. The standard deviation of the measured points about this curve is 0.05 percent. Since the Barnes/Mauersberger measurement at room temperature was 113.7E-19 cm^{2}, the relative correction to the BP84 data set is given by

f(t) = 1.0112 - 0.6903/ {87.3-t} (12)All BP84 cross sections should be multiplied by this factor, calculated at the relevant temperature. For example, for total ozone measurements at a mean temperature of -46.3°C, the correction factor is 1.006.

Results of CalculationsCalculations of the weighted quadratic coefficients have been carried out according to eqtn (2) both with and without the solar flux weighting. These results are tabulated in Tables 1A and 1B respectively. Values of the weighted absorption coefficients at -45C (corrected by eqtn (12)) are listed in the second last column of these tables, while the last column gives the percent change in the coefficient for one degree temperature change at -45°C.

The ozone-weighted mean temperature has also been calculated according to eqtn (3) for 14 different temperature profiles, each convolved with the standard ozone profiles referenced in the Introduction. In addition, ozone-weighted mean absorption coefficients were calculated according to eqtn (5), again with and without solar flux weighting. The results of these calculations were then used to solve eqtn (4) for

t, the implied effective mean temperature for each temperature-ozone profile combination. These temperatures are essentially the same with and without solar flux weighting. For the double pair wavelength combinations, they are some 1 to 3°C warmer than the ozone-weighted mean temperatures calculated from eqtn (3).In order to examine possible effects, calculations of the effective ozone absorption coefficients were made using the "dynamic" method, as described by eqtns (6) through (10), for various combinations of temperature and ozone profiles and for optical slant path ratios from 1 through 5, inclusive. In this exercise, the ozone profiles of Mateer

et. al.(1980) were used. In addition to the temperature effect, these results exhibited two other effects, viz., a relationship between effective absorption coefficient and the product mu*X, and a small residual relationship with total ozone for the AD, BD, and CD double wavelength pairs.To examine the temperature effect in the effective ozone absorption coefficients, the C1 and C2 coefficients from Table 1B were used with the simple ozone-weighted mean temperatures for each temperature-ozone profile combination. The C0 coefficient was set so that the mean absorption coefficient remained unchanged. The ozone path (mu*X), with X in atm-cm, relationship was also established as a quadratic, viz.,

d[ALFA(mu * X)] = CM0 + CM1 * (mu * X) + CM2 * (mu * X)^{2}(13)With CM0 set so that the overall mean absorption coefficient remained unchanged, the quadratic coefficients are listed in the table below:

PAIRCM0CM1CM2AD .0056 -.006517 8.877E-4 BD .0100 -.008506 4.206E-4 CD .0034 -.000306 -9.151E-5 CC' .0050 -.004333 -1.598E-4For each pair combination, the effective ozone absorption coefficient decreases as mu*X increases. For mu*X from 0.2 to 3.25 atm-cm, the total changes were AD: 0.8; BD: 3.8; CD: 2.2; and CC': 1.8 percent. Although these changes are moderately large for the latter three pair combinations, much of the change is contributed by the extreme cases of very large total ozone and optical slant path.

For the total sample of 185 calculations, 5 mu values and 37 temperature-ozone profile combinations, the overall averages and standard deviations of the residuals, after each effect is considered sequentially, are tabulated below:

PAIRABARSDSD(t)SD(mu*X)SD(ln[X])AD 1.4217 .0216 .0050 .0027 .0004 BD 0.7977 .0142 .0081 .0018 .0002 CD 0.4601 .0073 .0035 .0016 .0001 CC' 0.8226 .0120 .0034 .0001 --where ABAR is the overall average absorption coefficient.

For final recommendations on the effective mean absorption coefficients to be used with each wavelength pair combination, "dynamic" calculations were carried out using the Bhartia

et. al.(1984) standard profiles. It was believed these should be better because more recent ozonesonde data were used in their derivation. These calculations were made only for mu=2 which is a good average slant path for the vast majority of total ozone observations. The complete results of these computations are listed in Table 2. In these results, the Barnes/Mauersberger, temperature correction has been applied using the simple ozone-weighted mean temperature, which is also listed in the Table.In Table 2, the column TPROF gives the temperature code (see para. no. 4 in Introduction) as follows: 15 NA is for 15° North annual average; 30 NW is for 30° North winter; S is for summer; 75NWC is for 75° North winter cold; and US 62 is for the US 1962 Standard Atmosphere. The column O3PROF is the ozone profile code: EQ225 is the annual average profile for 15° North and 225 D.U. total ozone (1 D.U. = 1 Dobson Unit 10

^{-3}atm-cm); EM225 is the 225 D.U. profile annual average of 15 and 45° North; NM225 is the 225 D.U. annual average profile for 45° North; MH stands for the profile annual average of 45 and 75° North; and NH stands for the annual average profile at 75° North. The absorption coefficients have units atm-cm^{-1}.The warmest and coldest average temperatures and, consequently, the largest and smallest absorption coefficients, respectively, occur at high latitudes. These extremes are separated by 4.8 percent of their mean. Results for more typical conditions are summarized in Table 3. These variations with temperature are significant.

DiscussionThe emphasis today is on a high degree of accuracy for total ozone observations in order that valid results will be obtained from trend analysis of the total ozone data. Although there are significant temperature-dependent differences in the effective ozone absorption coefficients between stations at different latitudes and, during the course of each year, at single stations, will these differences have a serious impact on derived trends? If we are looking for trends of annual means, we have to consider the trend and variability of annual mean (ozone-weighted) temperatures for the days on which ozone observations are made. These temperature trends or inter-annual variability will have an impact of about 0.13 percent per degree Celsius change (see Table 1B).

Considering the logistic difficulties of establishing an ozone-weighted average temperature for correction of total ozone on a real-time basis, it would appear more reasonable to use "standard" ozone absorption coefficients for immediate publication data. In that way the data user would know what coefficient was used and to what ozone-weighted, mean temperature it applied. In short, temperature corrections, if deemed necessary, could be applied

a posterioriby the individual user, or by a designated World Center.Similar considerations may be applied to the smaller ozone path (mu*X) effect on the effective ozone absorption coefficient. In this case, however, if an individual user wished to make this correction, the mu values for each observation are not published. It seems unlikely that the "mix" of mu values used over a specific calendar month would vary very much from year to year. Therefore, for monthly average total ozone (

X), the main contributor to the variability of the average mu*X effect would be the variability ofXitself. If there is believed to be a need to apply a correction for this effect, Members of WMO could be requested to provide monthly mean values of mu for each of their stations. These means should be for direct sun observations taken during the month and submitted for publication. Such means could be included in the "remarks" section of the total ozone data form submitted to the World Ozone Data Centre.The Rayleigh scattering cross sections of Bates (1984) are generally considered to be the best currently available. Slit-weighted average Rayleigh coefficients (atm

^{-1}) derived from Bates' data are listed in Table 4 for each Dobson band pass.

Recommendations1. The effective ozone absorption coefficients listed in Table 4 should be adopted as the standard for total ozone measurements with the Dobson ozone spectrophotometer.

2. The Rayleigh scattering coefficients listed in Table 4 should be adopted as the standard for total ozone measurements with the Dobson instrument.

3. The standard effective ozone absorption coefficients apply to the US Standard Atmosphere, 1962 and the Bhartia

et. al.standard profile for 45° North and 325 D.U. total ozone. The values mu=2 and ozone-weighted mean temperature -46.3°C are also inherent to these coefficients.4. For temperature and/or mu*X corrections to published direct sun total ozone data on this new scale of observation, equations (4), (12) and (13), along with the quadratic coefficients in Tables 1B and the first unnumbered table on page 5 should be used. Before making corrections, these equations must first be normalized to the standard conditions described in Rec. 3.

5. Zenith Blue and Zenith Cloud total ozone observations should be evaluated using existing empirical cloud charts and then multiplying by the factor (old coefficient/new coefficient), viz., AD: 1.338/1.439 = 0.965; CD: .440/.466 = 0.944; and CC': .8/.833 = 0.960.

6. Umkehr evaluations should use the quadratic temperature coefficients to calculate a separate ozone absorption coefficient for each sub layer of the numerical integration system, using the mean temperature for the sub layer. The additional Barnes/Mauersberger temperature correction should also be applied for each layer.

7. These recommendations should be put into effect Jan. 1, 1991, or, if this is impracticable, July 1, 1991, or Jan. 1, 1992.

Discussion of ErrorsAs noted in the Introduction, the relative accuracy of the BP84 ozone cross sections is of the order of one percent, with an upper limit of about 2.5 percent for the absolute accuracy. Our effective absorption coefficients are essentially weighted means, which should improve their relative accuracy. In addition, there is a weighting over the temperature profile, which should reduce the errors introduced by the quadratic temperature regression. At wavelengths near 330 nm and longer, the percent random experimental error increases because the coefficients are becoming small. In addition, the extensions (mentioned earlier) to the formal BP84 data set, increase the uncertainty at the long D wavelength. However, this coefficient is quite small and even a 10 percent error does not have a large impact an the pair combination values. Finally, the single and double wavelength pair combinations involve the linear combination of as many as four coefficients, each with random error. Taking all these factors into account, it is estimated that the relative accuracy of the double pair and pair combinations is within the 2 to 3 percent range.

ReferencesBarnes, J., and K. Mauersberger: Temperature dependence of the ozone absorption cross section at the 253.7 nm Mercury line,

J. Geophys. Res., 92, pp. 14861-14864, 1987.Bass, A.M., and R.J. Paur: The ultraviolet cross-sections of ozone: I. The measurements in

Atmospheric Ozone(Ed. C.S. Zerefos and A. Ghazi), Reidel, Dordrecht, Boston, Lancaster, pp. 606-610, 1984.Bates, D.R.: Rayleigh scattering by air,

Planet. Space Sci., 32(6), pp. 785-790, 1984.Bhartia, P.K., D. Silberstein, B. Monosmith, and A.J. Fleig: Standard profiles of ozone from ground to 60 km obtained by combining satellite and ground based measurements, in

Atmospheric Ozone(Ed. C.S. Zerefos and A. Ghazi), Reidel, Dordrecht, Boston, Lancaster, pp. 243-247, 1984.Furukawa, P.M., P.L. Haagenson, and M.J. Scharberg: A composite, high-resolution solar spectrum from 208O to 3600 Å, NCAR TN-26, 55pp., 1967.

Hearns, A.G.: The absorption of ozone in the ultraviolet and visible regions,

Proc. Phys. Soc. London, 78, p. 932, 1961.Hudson, R.D.: A recommended set of ultraviolet cross sections for aeronomic purposes,

to be published, 1990.Mateer, C.L., J.J. DeLuisi, and C.C. Porco: The Short Umkehr method, Part 1: Standard ozone profiles for use in the estimation of ozone profiles by the inversion of Short Umkehr observations, NOAA Tech. Memo. ERL ARL-86, 20 pp., 1980.

Paur, R.J., and A.M. Bass: The ultraviolet cross-sections of ozone: II. Results and temperature dependence, in

Atmospheric Ozone(Ed. C.S. Zerefos and A. Ghazi), Reidel, Dordrecht, Boston, Lancaster, pp. 611-616, 1984.

US Standard Atmosphere, 1962, U.S. Govt. Printing Office, Washington, 1962.

US Standard Atmosphere Supplements, 1966, U.S. Govt. Printing Office, Washington, 1966.

US Standard Atmosphere, 1976, U.S. Govt. Printing Office, Washington, 1976.

TABLE 1A

TEMPERATURE COEFFICIENTS FOR AVERAGED DOBSON ALFA'S: SLIT WEIGHTS ONLYLAMDA C0(LAMDA) C1(LAMDA) C2(LAMDA) ALFA(-45) DA/DT(%) ***** ********* ********* ********* ********* ******** 3055 2.07270 4.4720E-03 1.8513E-05 1.92040 0.15 3089 1.35270 3.2464E-03 1.5941E-05 1.24633 0.15 3115 0.95116 2.4548E-03 1.2895E-05 0.87201 0.15 3175 0.42081 1.3262E-03 8.2263E-06 0.38006 0.15 3250 0.13774 6.5914E-04 3.1921E-06 0.11523 0.32 3291 0.07870 4.0956E-04 2.0823E-06 0.06487 0.34 3324 0.04941 2.9544E-04 1.6447E-06 0.03968 0.37 3399 0.01467 1.4108E-04 1.0291E-06 0.01047 0.46 A-PR 1.93496 3.8129E-03 1.5321E-05 1.80517 0.13 B-PR 1.27400 2.8368E-03 1.3859E-05 1.18146 0.13 C-PR 0.90176 2.1594E-03 1.1250E-05 0.83233 0.14 D-PR 0.40614 1.1851E-03 7.1972E-06 0.36958 0.15 AD 1.52882 2.6277E-03 8.1237E-06 1.43559 0.13 BD 0.86787 1.6517E-03 6.6615E-06 0.81187 0.13 CD 0.49562 9.7424E-04 4.0531E-06 0.46275 0.13

TABLE 1B

TEMPERATURE COEFFICIENTS FOR AVERAGED DOBSON ALFA'S: SLIT + FNOT WEIGHTSLAMDA C0(LAMDA) C1(LAMDA) C2(LAMDA) ALFA(-45) DA/DT(%) ***** ********* ********* ********* ********* ******** 3055 2.07660 4.4859E-03 1.8434E-05 1.92354 0.15 3089 1.35300 3.2493E-03 1.5942E-05 1.24650 0.15 3115 0.95464 2.4487E-03 1.2812E-05 0.87562 0.15 3175 0.42405 1.3226E-03 7.9408E-06 0.38290 0.16 3250 0.13530 6.4735E-04 3.1037E-06 0.11313 0.33 3291 0.07712 4.1051E-04 2.O915E-06 0.06325 0.35 3324 0.04976 2.9675E-04 1.6544E-06 0.03999 0.37 3399 0.01603 1.5126E-04 1.0739E-06 0.01147 0.48 A-PR 1.94130 3.8386E-03 1.5330E-05 1.81041 0.14 B-PR 1.27589 2.8388E-03 1.3851E-05 1.18324 0.13 C-PR 0.90488 2.1519E-03 1.1158E-05 0.83562 0.14 D-PR 0.40802 1.1713E-03 6.8669E-06 0.37143 0.15 AD 1.53328 2.6672E-03 8.4634E-06 1.43898 0.13 BD 0.86787 1.6675E-03 6.9836E-06 0.81182 0.13 CD 0.49687 9.8061E-04 4.2907E-06 0.46420 0.13

TABLE 2DETAILED OUTPUT FOR DYNAMIC DATA SET: MU=2TPROF 03PROF TEMP ALFA(AD) ALFA(BD) ALFA(CD) ALFA(CC') 15 NA EQ225 -39.6 1.4579 0.8228 0.4736 0.8441 15 NA EQ275 -43.2 1.4485 0.8168 0.4701 0.8394 15 NA EQ325 -46.2 1.4411 0.8121 0.4674 0.8358 30 NW EM225 -43.2 1.4495 0.8160 0.4707 0.8386 30 NS EM225 -38.6 1.4589 0.8232 0.4739 0.8445 30 NW EM275 -46.4 1.4416 0.8128 0.4678 0.8348 30 NS EM275 -41.8 1.4499 0.8175 0.4705 0.8402 30 NW EM325 -48.7 1.4353 0.8087 0.4655 0.8319 30 NS EM325 -44.4 1.4432 0.8131 0.4679 0.8368 30 NW NM375 -50.7 1.4303 0.8054 0.4636 0.8296 30 NS NM375 -46.7 1.4375 0.8094 0.4658 0.8340 30 NW NM425 -52.4 1.4267 0.8027 0.4621 0.8277 30 NS NM425 -48.6 1.4335 0.8O65 0.4642 0.8319 US 62 NM225 -42.7 1.4501 0.8180 0.4709 0.8383 45 NW NM225 -50.1 1.4374 0.8111 0.4671 0.8307 45 NS NM225 -35.7 1.4633 0.8256 0.4752 0.8470 US 62 NM325 -46.5 1.4389 0.8103 0.4663 0.8332 45 NW NM325 -52.4 1.4289 0.8049 0.4633 0.8273 45 NS NM325 -40.6 1.4494 0.8162 0.4696 0.8400 US 62 NM425 -48.5 1.4323 0.8053 0.4635 0.8300 45 NW NM425 -53.5 1.4243 0.8009 0.4611 0.8252 45 NS NM425 -43.9 1.4411 0.8103 0.4663 0.8357 US 62 NM525 -56.1 1.4278 0.8014 0.4615 0.8276 45 NW NM525 -54.0 1.4213 0.7979 0.4595 0.8237 45 NS NM525 -46.5 1.4353 0.8057 0.4639 0.8324 60 NW MH225 -48.9 1.4381 0.8114 0.4672 0.8309 60 NS MH225 -34.4 1.4644 0.8260 0.4754 0.8473 60 NW MH325 -51.6 1.4289 0.8049 0.4632 0.8272 60 NS MH325 -38.1 1.4532 0.8182 0.4707 0.8419 60 NW MH375 -52.3 1.4260 0.8026 0.4620 0.8260 60 NS MH375 -39.9 1.4497 0.8154 0.4691 0.8401 60 NW MH425 -52.9 1.4238 0.8O07 0.4609 0.8249 60 NS MH425 -40.2 1.4467 0.8131 0.4678 0.8385 60 NW MH525 -53.6 1.4206 0.7975 0.4595 0.8233 60 NS MH525 -41.7 1.4423 0.8092 0.4658 0.8360 75 NW NH225 -59.0 1.4208 0.8025 0.4623 0.8216 75 NS NH225 -31.3 1.4664 0.8269 0.4759 0.8482 75 NW NH325 -60.4 1.4140 0.7973 0.4591 0.8194 75 NS NH325 -34.6 1.4572 0.8202 0.4718 0.8440 75 NW NH375 -60.7 1.4121 0.7955 0.4581 0.8187 75 NS NH375 -35.7 1.4541 0.8178 0.4704 0.8425 75NWC NH375 -68.6 1.3971 0.7796 0.4517 0.8090 75 NW NH425 -60.9 1.4106 0.7940 0.4573 0.8181 75 NS NH425 -36.6 1.4516 0.8157 0.4693 0.8412 75 NW NH525 -61.0 1.4085 0.7914 0.4560 0.8171 75 NS NH525 -37.9 1.4477 0.8121 0.4674 0.8391

TABLE 3AVERAGE TEMPERATURE AND OZONE ABSORPTION COEFFICIENTS FOR TYPICAL CONDITIONSTEMP OZONE MEAN ALPHA ALPHA ALPHA ALPHA PROFILE PROFILE TEMP AD BD CD CC' 15 NA EQ225 -39.6 1.458 0.823 0.474 0.844 15 NA EQ275 -43.2 1.448 0.817 0.470 0.839 15 NA EQ325 -46.2 1.441 0.812 0.467 0.836 30 NS EM225 -38.6 1.459 0.823 0.474 0.845 3ONWS EM300 -45.4 1.443 0.813 0.468 0.836 30 NW NM425 -52.4 1.427 0.803 0.462 0.828 45 NS NM225 -35.7 1.463 0.826 0.475 0.847 US 62 NM325 -46.3 1.439 0.810 0.466 0.833 45 NW NM525 -54.0 1.421 0.798 0.460 0.824 60 NS MH225 -34.4 1.464 0.826 0.475 0.847 60 NWS MH375 -46.1 1.438 0.809 0.466 0.833 60 NW MH525 -53.6 1.421 0.798 0.459 0.823 75 NS NH225 -31.3 1.466 0.827 0.476 0.848 75 NWS NH375 -48.2 1.433 0.807 0.464 0.831 75 NW NH525 -61.0 1.409 0.791 0.456 0.817

DOBSON TOTAL OZONE MEASUREMENTSSTANDARD-OZONE ABSORPTION ANDRAYLEIGH SCATTERING COEFFICIENTSWAVELENGTH ALFA BETA d(BETA)/d(ALFA) (nm) (atm-cm)-l (atm)-l (atm-cm/atm) ========================================================== 305.5 1.915 0.489 325.0 0.109 0.375 A 1.806 0.114 0.063 308.9 1.239 0.466 329.1 0.062 0.355 B 1.177 0.111 0.094 311.5 0.873 0.450 332.4 0.040 0.341 C 0.833 0.109 0.131 317.5 0.384 0.414 339.9 0.017 0.310 D 0.367 0.104 0.283 AD 1.439 0.010 0.007 BD 0.810 0.007 0.009 CD 0.466 0.005 0.011 ==========================================================