GEOPHYSICAL RESEARCH LETTERS, VOL. 18, NO. 12, PAGES 2261-2263, DECEMBER 1991 THE BLISS MEASUREMENTS OF NO2: SOME NEW INSIGHTS D.J. Laryand J.A.Pyle Department ofChemistry, Cambridge Urdversity, UK. C.R. Webster and R.D.May Jet Propulsion Laboratory, California Institute of Technology Abstract - The diurnal variation of NO 2measured near 30 km by the BLISS in-situ laser spectrometer is compared with calculations from a photochemical model that includes a detailed description of multiple scattering. Introduction In a recent paper, thestratospheric chemistry of NO 2 wasstudied by combining high resolution tunablediode laser measurements from the balloon-borne laser in-situ spectrometer (BLISS)with a numerical model [Webster et al., 1990]. When the modelwas constrained with the measured 03 and temperature, the mainfeatures of the observed diurnal variation of NO 2 were wellreproduced by the model.In general, the study constituted a very successful confirmation of our present understanding of this subset of stratospheric chemistry. However,there were some differences in detail between the model and observations (Figure 1). In particular, the one daytime measurement of NO2 could not be reproduced unless an unrealistically largevalue of the ground albedo wasused, and there was a very rapid,but very brief, decline in the measuredNO2 immediatelyafter the major sunset increase which the modeldidnot reproduce. Finally some minor differences between the modelled and observed nighttime decayswere evident with the model slightly overestimating the values immediately prior to dawn. Thisstudy reconsiders the detailed aspects of themodel data intercomparison. We have used a numerical model with similar chemical kinetics but a muchmore sophisti- cated treatment of scattering thanthat used in the earlier study. The radiative transfer model is a new implementa- tion of the scheme described byMeier et al. (1982). It has been extended after Anderson (1983) to describe correctly the radiation fieldfor solar zenith angles greater than 75 ø . The radiation into any volumeelement of the model atmosphere hasfour contributions: The direct solar flux, the diffuse flux incident from all directions, the ground reflection of the direct solar flux and theground reflection ofthediffuse flux. In this paper clear sky conditions have been assumed. Excellentvalidation of the scheme against various data sets has beenachieved [Laryand Pyle, in press, 1991a,b]. Copyright 1991 by theAmerican Geophysical Union. Paper number 91 GL02732 0094-8534/91/91GL-02732503.00 Results In addition to several other species Webster et al. (1990); May and Webster (1989), made simultaneous in-situ measurements of NO2, 03, temperature, and pressure from the balloon based BLISS instrument launched at Palestine, Texas(32øN, 96øW),on 13 September, 1988. These measurements allowed a critical test of the photochemicalmodel to be made, with very small differences betweentheory and data as discussed above. This new study has reproducedsome of the finer BLISS measurementdetails (Figure 2) with a model/measurement agreement of < 10%, which is comparable to the measurement uncertainties of 8% for sunset andnighttime NO2, and 10% for the daytime NO2 measurement. The model used to produce Figure2 hasa ground albedo of 0.25,a constant temperature of 233 K, a constant ozone concentration of 8.5 ppmv,and a total reactive nitrogen content of 15.24 ppbv. The model pressure is kept at 11.0 mb, exceptfor one hour after sunset, when it is raised to 11.6 mb in accordancewith the observations of Webster et al. (1990). Firstly, as pointed out by Websteret al. (1990), the single day timemeasurement of NO2 could only be expla- ined usingtheir scattering schemebased on Kurzeja (1976), if a groundalbedo of 0.5 was used.This is an unrealistic value for desert like conditions. When the detailedtreatment of multiple scattering is used at all wavelengths, then the single day time measurement of NO 2 can be explained to within the measurement uncertainty by a ground albedoof 0.25 (Figure2). Secondly, the peak in NO2 concentration which occurred justaftersunset canbe reproduced if the slight, but brief, descent of the balloon which occurred then is included in the model. This descent wasaccompanied by an increasein the pressure,and hence, in the total number density.Note that in their model calculations Webster et al. (1990) used a constant valuefor the total number density of 3.6x1047 molecules cm '3. Theobserved peakwould be less obvious if the concentration of NO2 wasexpressed as a volume mixing ratio. Finally, in the Webster et al. (1990) study the modelled NO2 prior to sunrise slightly overestimated the measured values, although the measured and modellednighttime decay certainlyagreed within the experimental error. Webster et al. (1990), discussed thesensitMty of thedecay to temperature. By using a temperature of 233 K, instead of 231K asused in the earlier study, even betteragree- mentis found, especially immediately beforedawn. 2261