TB.SM.APJ/Ap14479111ART.15/10/2012 THE ASTROPHYSICAL JOURNAL, 759: I (7pp),2012 ??? doi: 10.1088/0004-637XnS9/1I1 02012. The American Astronomical SocielY. All righls reserved. Printed inlhc U.S.A. ELUSIVE ETHYLENE DETECTED IN SATURN'S NORTHERN STORM REGION B. E. HESMANI,9, G. L. BJORAKER2, P. V. SADA3,IO, R. K. ACHTERBERG I ,9, D. E. lENNI GS4,IO, P. N. ROMANI2, A. W. LUNSFORD5,1O, L. N. FLETCHER6, R. 1. BOYU7,IO, A. A. SIMON-MILLER8, C. A. NIXONJ,9, AND P. G. 1. IRWlN6 I Department of Astronomy, University of Maryland, College Park, MD 20742, USA; [email protected] 2 NASA/GSFC Code 693, Greenbelt, MD 20771, USA 3 Departamento de Fisica y Matemalicas.UniversidaddeMonterrey.GarzaGarda.NL 66238, Mexico 4 NASA/GSFC Code 693 and Code 500, Greenbeh, MD 20771, USA 5 Department of Physics, Catholic University of America, Washington, DC 20064, USA 6 Atmospheric, Oceanic and Planetary Physics, Clarendon Laboratory, Parks Road, Oxford, OXI 3PU, UK 7 Department of Physics and Astronomy, Dickinson College, Carlisle, PA 17013, USA 8 NASAjGSFC Code 690, Greenbelt, MD 20771, USA Received 2012 May 7; accepted 2012 September 28; published 20/2 1?? ABSTRACT The massive eruption at 400N (planetographic latitude) on Saturn in 2010 December has produced significant and lasting effects in the northem hemisphere on temperature and species abundances. The northern storm region was observed on many occasions in 2011 by Cassini's Compo ite Infrared Spectrometer (CIRS). In 2011 May, temperatures in the stratosphere greater than 200 K were derived from CIRS spectra in the regions referred to as "beacons" (warm regions in the stratosphere). Ethylene has been detected in the beacon region in Saturn's northem storm region using CIRS. Ground-based observations u ing the high-resolution spectrometer Celeste on the McMath-Pierce Telescope on 20]1 May 15 were used to confirm the detection and improve the altitude resolution in the retrieved prOfile. The derived ethylene profile from the CIRS data gives a C H mole fraction of 2 4 5.9 ± 4.5 x 10-7 at 0.5 mbar, and from Celeste data it gives 2.7 ± 0.45 x 10-6 at 0.1 mbar. This is two orders of magnitude higher than the amount measured in the ultraviolet at other latitudes pI;or to the storm. It is also much higher than predicted by photochemical models, indicating that perhaps another production mechanism is required or a loss mechanism is being inhibited. Key words: planets and satellites: atmospheres - planets and atellites: individual (Saturn) hemisphere throughout 2011. Due to technology limitations, 1. INTRODUCTIO studie of these massive torm systems in the thermal infrared Ql Storms on Saturn probe the deep atmosphere as material have previously been impossible. Optical studies of these is transported from levels beyond the reach of sunlight up storm provide important information on morphology and cloud to the observable atmosphere. Storms have a dramatic effect top locations. However, thermal infrared studies are essential on the neighborhood, introducing sudden changes that dwarf to measuring the temperature and gas composition of the the effects of seasonal change as Saturn progresses through environment inside and adjacent to storm systems on Saturn. its 29 year revolution about the Sun. In 2010 December, Fortunately, the effects of the massive storm continued Saturn's nOlthern hemisphere was spectacularly disrupted from through at least 2011 July and the Compo ite Infrared Spec its slow springtime warming by a massive storm eruption at trometer (CIRS) on Cassini has produced an extensive record approximately 400N (planetographic latitude; Fletcher et aL of the distw·bance. Thi has been supplemented by ground 2011). This storm is only the sixth of it magnitude to ever based observation at higher spectral resolution. Here we report be observed and the first one at this latitude in over a century. the detection of ethylene in 2011 May u ing both ClRS and Saturn is known to have massive storm ystems that erupt the ground-based pectrometer Cele te on the McMath-Pierce approximately once per aturnian year (29.4 Earth years), the Tele cope. Ethylene was detected in the warm beacon regions in most recent of which happened in 1990 September at 12°N Saturn's u'ato phere. These are localized regions in latitude and (Sanchez-Lavega et aL 1991; Beebe et al. 1992). It is not longitude that are heat~d by 50 K or more compared with pre known what triggers these large storm . The approximate annual storm conditions. These stratospheric features are thought to be periodicity suggests solar forcing, yet sunlight does not reach generated as a response to the upwelling of the storm (Fletcher the water cloud where these torms are believed to originate et aJ. 20 II). Previous infrared studie of Saturn's ethylene emis (Dyudina et al. 2010; Hueso & Sanchez-Lavega 2004). Saturn sion have shown it to be difficult to detect on Saturn (Encrenaz emits more radiation than it receives from the Sun, but the et al. 1975; Bezard et aL 2001), but the high temperatures in details of this process are not well understood. Stolm may the beacon region have produced trong thermal emi sion at be a way in which Saturn releases its excess thermal energy 10.5 tLm (950 cm- I) due to ethylene in the stratosphere. A in sudden bursts, rather than gradually. The convective plume recent analysi of tellar occultation data acquired by the Ul that erupted in 2010 December was nearly 10 years earlier than traviolet Imaging Spectrograph (UVIS) on Cassini has yielded expected in the cycle of great eruptions on Saturn (Sanchez a detection and vertical profile of ethylene between 0.1 tLbar Lavega et al. 1991). Thi convective plume was sheared to the and 0.5 mbar in Saturn's atmosphere at 15°2 (Shemansky & north and south and its effects continued to disrupt the northem Liu 2012). Comparisons of ethylene abundances derived in the storm region with other locations on Saturn provide insight into photochemical and dynamical processes occurring on this giant 9 Also at CRESST and NASA/GSFC Code 693, Greenbelt, MD 20771, USA. 10 Also at National Solar Observatory. planet. THE ASTROPHYSICAL JOURNAL, 759: I (7pp) , 2012 ??? HESMAN ET AL. Brightne .. Temperatu'" (K) 1.0 150 160 170 180 190 200 -I 1305 em 60 SO ~ .:..:.l 40 :;; co ...J 30 120 60 0 West Longitude Brightness T" "'perature [K) SO 60 70 80 90 100 110 120 -I 950em 240 180 120 60 o West Longitude Figure L 2~5 em- I eIRS map over the northern storm latitudes from 2011 May 4. Top: this map shows data by plotting all of the FP4 pixels in the methane band at 1305 em I .. Bottom: thiS map shows data by plotting all of the FP3 pixels in the ethylene region at 950 em-I. The cenlroid of the hot stratospheric beacon is at - 29(J'W longttude. Both maps have spatial resolutions of 2° by 2°. 2. OBSERVATIONS map covered almost the full extent, in the north-south direction, The combination of Cassini CIRS data with ground-based ?f~e storm region. Figme 2(a) shows the CIRS ethylene spectra lllslde (red curve) and outside (blue curve) the beacon region. spectroscopy provides a powerful set of capabilities for studying The beacon region spectrum was created by averaging a total hydrocarbons on Saturn. CIRS provides broad-band, absolutely of 304 spectra from 3° in latitude (average planetographic calibrated spectra at high spatial resolution and modest spectral latitude of 38°) and 20° in longitude (average longitude of resolution. It also provides, through measurements of C~ and 2900W- system III). The non-beacon spectrum was created by collision induced H2 opacity, temperature profiles for both the averaging a total of 3504 spectra from 3° in latitude and 190° in stratosphere and upper troposphere of Saturn. The use of a longitude (average longitude of 91 OW-system III). Ethylene is cryogenic grating spectrometer (Celeste) operating at resolving not present above the noise level in the non-beacon spectrum but powers of up to 30,000 on ground-based telescopes permits I appears as a single feature at 950 cm- in the beacon spectrum the detection of multiple emission lines of ethylene and yields with a signal-to-noi e ratio (SIN) of 8. improved altitude resolution in the retrieved vertical profile. Our ground-based instrument, Celeste, is a cryogenic grating CIRS is a dual Fomier transform spectrometer covering spectrometer with an anay detector that can achieve resolving the thermal infrared with three focal planes: the far-infrared powers up to 30,000. This instrument has been described in focal plane, FP 1, which is a single thermocouple detector I Jennings et al. (2009). The instrument setup was similar to that covering 10-600 cm- , and two mid-infrared focal planes, FP3 used in previous observations (Romani et al. 2008; Hesman and FP4, which are arrays of 10 HgCdTe detectors covering I I et al. 2009). The Celeste observations were performed at the 600-1100 cm- and llOO-I500 cm- , respectively (Flasar McMath-Pierce Telescope on 2011 May 15. The spectrometer et aJ. 2(05). The far-infrared detector operates at the instrument was coupled to the telescope with foreoptics which provided temperature of 170 K, and the mid-infrared detectors operate an approximately f/8 beam to the spectrometer entrance slit. at 80 K. The fields of view per detector are 3.9 mrad for the The spectrometer uses an 18 x 34 cm2 echelle grating and an far-infrared and 0.3 Imad for the mid-infrared. The apodized spectral resolution is selectable from 0.5 to 15 cm-1 . Si:A detector array with a 128 x 128 pixel format and a spatial The CIRS observations were perfonned on 2011 May 4 using re olution of ~1" pixel-I The entire spectrometer is cooled a specu·al resolution of 2.5 cm-1. The spacecraft was targeted with liquid helium to an operating temperature of 6 K. For the Celeste observations the slit was oriented east-west at 45°N (planetographic latitude) for an 11 hr integration while on the planet and centered on 35°N latitude with a 300 /km slit the planet rotated across the field of view. This was followed by width to capture the spatial extent of the warm storm region. a second pointing at 35°N for an additional planetary rotation. This translates into a slit length of 103/1 and a slit width of 4/1. The result was a cylindrical map of Saturn spanning 31 °N-49°N The long slit was oriented in the east-west direction. Saturn over all longitudes, as shown in Figure 1. Calibration reference spectra were recorded by ob erving deep space and an internal was ~ 19/1 during these observations, which is Significantly smaller than the slit length of Celeste. Each resolution element warm somce (the shutter of the instrument). Figure 1 shows the subtended 33° in latitude (slit width) and 10° in longitude (along radiance (expressed in brightness temperature) in the methane I the slit). In addition, each time step covered 1800 in longitude band (1305 cm- ) as well as in the center of the ethylene I and over a 3.5 hr time period 300° of longitude were sampled. band (950 cm- ). The maps of Figure 1 show that the beacon This technique was used to guarantee continuous coverage centroid is located near 2900W longitude and the mid-infrared 2 __1 THE ASTROPHYSICAL JOURNAL, 759: I (7pp),2012 ??? HESMAN ET AL. ,', Ii .,1'" 'i;' Ii, i " j region using data from May 13th when the beacon was on the opposite side of the planet (blue curve). This spectrum repre sents a total of 4.5 hr of integration time on the non-beacon side of Saturn. A scale factor for performing radiance calibration of the Celeste data was derived from modeling the continuum outside of the beacon region. As no ethylene was visible in the non-beacon data the continuum was modeled using known tem peratures to derive a scale factor. The resulting scale factor was then applied to the beacon region. The uncertainty in the scale factor is estimated at 15%. 3. DATA ANALYSIS AND RESULTS Ethylene emission in the thermal infrared is coupled to both atmospheric temperature and the species abundance. The Ot, I .,1 ".,1 1 ••• ! I. ", ! ,< .,, ,1" observed spectrum near 950 cm- consists of ethylene emission 920 930 940 950 960 970 980 from the stratosphere riding on an optically thick continuum of WAVENUMBER (em") collision-induced hydrogen produced in the troposphere. Thus, 3 it is important to derive both the temperature in the troposphere 35b5~YWll:Northern~lormb~aeon";gIO~(_35N,_3;OW) i ' (b) and in the stratosphere from CIRS data. This temperature profile is then applied to models to retrieve the ethylene abundance ~ profile in analyzing both the CIRS and Celeste spectra. 'E 2Sr The temperature prOfile was retrieved using spectral regions separate from the ethylene region, in both the FP3 and FP4 focal planes. The same averaging used to produce the ethylene spec trum in the beacon region (as described in Section 2) was used to create the spectra lIsed for temperaUlre retrievals. Separate il0~-' \~_,' .:', " "", ' retrievals for upper tropospheric and stratospheric temperatures are performed using the constrained linear inversion algorithm a~: tF... .i "'l:"-'~ Jj,.- ~ . described by Conrath et al. (1998). A detailed summary of the method used in pelforming the temperature inversion can be 5,... found in Achterberg et al. (2008). For the tropospheric retrievals, I r the spectral ranges 600-620 and 640-660 cm- (FP3) are used Oi, I , 948.5 949,0 949.5 where the atmospheric opacity is from the collision-induced S(I) WAVENUMBER (em") line of hydrogen, assuming equilibrium hydrogen. Opacity from Fthieg uhroet 2s.t r(aato) sTphheer i2c. 5b ceamco-nI C(rreRdS) aspnedc torua tsaitd e3 8t°hNe rferogimon 2 0(b1l1u eM). aTy h4e ienslRidse fHro2-mH B2,o rHyzso-Hw ee, t Hal2. -(C1H9845 ,p a1i9r8s8 )is ainndc lBuOdIeyds,o uws i&n g FarolgmOmlithhomlds spectra are shown over a 60 cm-i interval. At this resolution, ethylene is (1986). An He/H2 ratio of 0.135 is assumed (Conrath & Gautier detected as a single feature. (b) The 0.1 cm-I Celeste spectra at 35°N from 20 I 1 May 15 (McMath,Pierce Telescope) inside the hot stratospheric beacon (red) 2000) along with a pressure dependent CH4 mole fraction profile and outside the region (blue). The Celeste spectra are shown over a 2 cm-I based on the photochemical profile in Moses et al. (2000) scaled interval. Ethylene is detecled as a band of lines at this resolution, to a tropospheric value of 4.5 x 10-3 as given in Flasar et al. (2005). The stratospheric temperatures are retrieved using the of the beacon a it moved across the day ide hemi phere of V4 band of CH4 between 1250 and 1311 cm-I (FP4). Methane Saturn. This slit width gave a spectral resolution of 0.1 cm-I, tran mittances were calculated using the correlated-k method The detector array covered a 2 cm-I interval, centered on (Lacis & Oinas 1991), using line data from the GEISA 2003 949.4 cm-I. Sky subtraction was performed by nodding the line atlas (Jacquinet-Husson et al. 2005) with H2/He broaden telescope and the mOOn was used as a fiat-field reference for ing (Linda Brown, private communication). Figure 3(a) shows calibration, the temperature profile (solid black curve) retrieved from the The Celeste observations were reduced in the standard beacon region in Saturn's northern storm region. The uncer method whereby the spectra from the two nod positions were tainty limits in the retrieved temperature profile are approxi subtracted to remove the offset signal, the response of the ar mately 1 K over the 0.5-5 mbar range in the stratosphere and ray was normalized by using the moon fiat-fields, and the main over the 100-300 mbar range in the troposphere. Between these C2H4 emission feaUlre was shifted spatially between subtracted pressure regions, the inversion algorithm smoothly interpolates pairs (to account for rotation of the planet) and shifted spec temperatures. The temperature profile is not well determined at tralJy (to account for the change in radial velocity as the planet pressure below 0.5 mbar. However, the sen itivity of the ethy rotated). A total of 12 subtracted pairs were averaged to create lene retrievals to different temperatures at pressures less than the C2H4 emission spectrum shown in Figure 2(b) (red curve), 0.5 mbar was extensively tested. There is limited sensitivity to This spectrum represents a total of 3.4 hr of integration time on this altitude range in the ethylene line, and retrieved ethylene the beacon region. In 2011 May, CIRS data from two successive prOfile does not Significantly cbange depending upon the input maps indicate that the beacon was moving at 2° day- I. The bea temperaUlre in this altitude range. con was therefore centered near 3100W longitude On May 15 The retrieved temperature profile in the beacon region was during the Celeste observations. Ethylene is detected as a band used in a photochemical model in order to generate an a of lines in this spectrum with an SIN of 17. This same method priori abundance profile of the hydrocarbons present in the was applied in order to create the spectrum outside the beacon stratosphere (including ethylene). These theoretical profiles are 3 THE ASTROPHYSICAL JOURNAL, 759: 1 (7pp),2012 ??? HESMAN ET AL. 10~ i _ 25 'E .....!..! "e 20 " ;: .s ~ 10' w :e:n> oz en "" t0Ur.J. 10'· "aC":i 10' 10',·· 945 950 955 960 WAVENUMBER (em") 50 100 150 200 250 TEMPERATURE tKl - " I (b) (b) 25 '... ~ 10" .D .5- .~ li! 10° ~ 15 ::> en w en o z ~ 10'f--___ ~ lOt « ------ C,H. CONSTANT PROFILE a: C,H, RETRIEVED PROFILE 5 _____ _ C..H. SCALED PROFILE -------~ R8'RIEVED TEMPERATURE PROFILE 10' o~ ~ 940 945 950 955 960 50 100 150 200 250 WAVENUMBER (em-') TEMPERATURE (K) Figure 3. (a) The temperature profile as retrieved from the 2011 May CIRS data Figure 4. (a) The elRs spectrUm of ethylene (black solid curve) inside the hot (black soljd curve) and the calculated C2H4 photochemical profile produced stratospheric beacon. The colored curves are models based on C2H4 profiles using the temperature profile as input (black dashed curve). (b) The 2.5 cm-I with: a constant abundance of 12 ppb (green); a retrieved profile starting from CIRS spectra at 38°N from 20 II May inside the hot stratospheric beacon (solid the constant abundance profile (blue); and, a scale factor retrieval starting curve). The dashed curve is the model that results when the C2 H4 photochemical from the constant abundance profile (red). (b) The temperature profile used profile and temperature profile are provided as input. Enhanced temperatures in producing the model spectra shown in (a) (black), the colored curves are the alone do nOt produce a measurable C2I-I4 line. C2H4 abundance profiles that produce the model spectra shown in (a). generated using a one-dimensional photochemical model. The communication). Inputs into the model were the temperature Q2 model takes into account the photolysis and chemical reactions profiles derived from the data inside and outside the beacon that interlink the hydrocarbons with each other and atomic region and the photochemical profiles of C H calculated based 2 4 hydrogen. It solves their coupled continuity equations assuming on the derived temperature profiles. steady state conditions. The net flux of the species includes terms In a first test, a forward model was run using the temperature for both transport (eddy mixing) and molecular diffusion. For a profile retrieved from the CIRS data (Figure 3 top; olid curve) more detailed model description of the model ee Romani et al. and a photochemical model of ethylene that wa calculated using (2008). Ethylene has a photochemical lifetime of about 20 days the retrieved temperature profile (Figure 3 top; dashed curve). in the 2-0.2 mbar region. The photochemical model predicts In (Figure 3 bottom) we compare the observed CIRS pectrum C H abundance profiles for the warmest (220 K) regions of of the beacon region with a synthetic spectrum calculated using 2 4 Saturn's stratosphere that are enhanced by a factor of two in the heated temperature profile and C H profile shown in the 2 4 the 2-0.2 mbar altitude range over the ethylene abundances top figure. It is clear that enhanced temperatmes alone do not calculated for the unperturbed (140 K) atmosphere. produce sufficient ethylene emission to match the observations. The C H abundance profile retrievals were performed using Thus, increasing the abundance of this species is needed in order 2 4 the Non-Linear Optimal Estimator for Multivariate Spectral to fit the CIRS data (Figure 3 bottom; solid curve). Analysis (NEMESIS) code as described in Irwin. et al. (2008). In order to assess the information content in the CIRS and Absorption of the contributing species was calculated using the Celeste ethylene spectra an a priori profile that is constant correlated-k method (Lacis & Oinas 1991). The k-tables for with altitude until 20 mbar was used. This technique was C2~ were calculated using line parameters based on data from used to determine how sensitive the model retrievals are to the GEISA 2003 line atlas (Jacquinet-Husson et al. 2005) with the starting conditions. In Figw'e 4, an ethylene mole fraction modifications to the temperature exponent (set to 0.73) modified of 1.2 x 10-8 was assumed for the constant profile (green to use H2 pressure broadening rather than N2 (B. Bezard, private curves). This profile produces a good fit to the CIRS data. In 4 THE ASTROPHYSICAL JOURNAL, 759: 1 (7pp),2012 ??? HESMAN ET AL. 35 (a) (a) 40 30 'E !d ~ 30- ~.. .. 25 CIl E u .~.s w t) Z <t 0 <a:t: 948.5 949.0 949.5 950.0 WAVENUMBER (em") 940 945 950 955 960 WAVENUMBER (em") 1(y'0 MOLE ~~~CTION 10" 10" 35 10'" (b) (b) 30 ' .. ........ ...... , ... 'E !d ~ 10" .~ 25 ..Ds CIl N' E :i! 100 u 20 ::> .=.=s '" i'"f w 10'~ _____- ==,_ _~ ~ Zt) <t ------ C,H. CONSTANT PROFILE 0 10' C,H, RETRIEVED PROFILE <a:t: -.,._ ..._ . ~~R~~kg~:~~:~L:TURE PRO ALE 10' 50 100 150 200 250 948.5 949,0 949.5 950.0 TEMPERATURE (K) WAVENUMBER (em") Figure S. (a) The Celeste spectrum of ethylene (black solid curve) inside Figure 6. (a) The CIRS spectrum of ethylene (black solid curve) inside the hot the hot stratospheric beacon. The colored curves are models based on C2H4 stratospheric beacon. The dashed curve is the best-fit model to the data. (b) The profiles with: a constant abundance of 3 ppb (green); a retrieved profile starting Celeste spectrum of ethylene (black solid curve) inside the hot stratospheric from the constant abundance profile (blue); and, a scale factor retrieval starting beacon. The dashed curve is the best-fit model to the data. from the constant abundance profile (rcd). (b) '[be temperature profile used ill producing the model spectra shown in (a) (black), the colored curves are the C2~ abundance profiles that produce the model spectra shown in (a). adju t the photochemical profiles between 0.03 and 3 mbar in order to retrieve the ethylene abundance in this pressure range. me other two cases me constant profile was allowed to vary Figure 6 shows the ethylene spectra from CIRS and Celeste continuously (blue) and with a scale factor (red) in order to along with their best-fit model spectra ..I n both cases the data produce the best-fit to the CrRS line. The profiles that result are fit very well with the model. In Figure 6(b) there is an excel are not significantly different from the constant abundance with lent fit of the model to the peak emission in the Celeste spectrum. I height profile. In Figure 5, a constant with height profile with However, the poor fit around the ethylene feature at 950.1 cm- an abundance of 3 x 10-9 is used as the a priori profile (green may be the result of incomplete laboratory data for ethylene at curves) to fit the Celeste data. This profile does not produce this wavenwnber. a spectrum that fits any of the ethylene lines in the Celeste Figure 7 (top) shows the retrieved ethylene profile from the spectrum (Figure 5(a)). A best-fit profile was retrieved using CIRS data (solid green curve). The dashed green curve is the a scale factor adjustment (red curve) which fits the ethylene photochemical profile computed from the retrieved temperature lines at 949.65 em-I and 950.05 cm-1, but does not fit the profile shown in black. This profile was used as the a priori input main ethylene line at 949.35 em-I. When a best-fit profile into NEMESIS. The error bars are shown only for me pressures was derived by allowing the a priori constant profile to vary over which the retJieval is sensitive based on the altitude probed continuously in altitude the blue curve in Figure 5(b) results. by the C2~ line present in the CIRS data. The peak abundance This profile produces the best-fit overall to the entire Celeste found from the CIRS data is 5.9 ± 4.5 x 10-7 at 0.5 mbar. spectrum. These results have motivated continuous profiles to The uncertainties were calculated using NEMESIS which takes be retrieved from both data sets using the photochemical profile into account both the measured error in the spectrum and an as an input. appropriate amount of forward modeling error. This includes, In performing the retrievals from the CIRS data the tempera for example, uncertainties based on the line parameters and in ture profiles were fixed but the calculated photochemical profiles the correlated-k model (see Irwin et al. 2008). were used as the initial prediction and allowed to vary in order In performing the retrievals from the Celeste data the tem to fit the emission in the CIRS spectra. NEMESIS was used to perature profiles were also held fixed but in this case the C2H4 5 __ J THE ASTROPHYSICAL JOURNAL, 759: 1 (7pp), 2012 ??? HESMAN ET AL. MOLE ~~,g.CT10N 4. DISCUSSION 10'" The detection of ethylene in the northern storm region 10" is significant for two reasons. First of all, thiS short-lived hydrocarbon is a tracer of chemistry occurring in an unusual, 10.2- " dynamical region of Saturn's stratosphere. In addition, prior ground-based searches for ethylene in the thermal infrared ~A 10"-- ,, have resulted only in tentative detections on Saturn (Encrenaz S et al. 1975; Bezard et al. 2001). Ethylene has never been aW: 100- detected by CIRS prior to the eru ption of the northern storm. A ::;) ~ significant finding from these measurements is that calculations Q3 ·.a0W.:. 10'- C?H~ l based on photochemistry do not produce an ethylene profile ------ INPUT PHOTOCHEMICAL PROALE -l that when combined with the increased beacon temperatures 10"- ---- RREETTRRIIEEVVEEDD TC~ HM,P CEIRRSA TPURROEA PLREO FILE i produces a measurable ethylene line. Emission from aJl of the I j hydrocarbons is expected to increase in the beacon regions 10"· because of the extreme temperatures found there (these results 5l0 -.-... -~. 10~ .0. - L- 150 200 2j5 0 isnh oewth yalpepnreo xoifm aaptperlyo x2im20a tKel ya tt w2o mtibmaer)s. tAheb uqnudiaesncceen ti nacmreoausenst TEMPERATURE IKl MOLE Wi-CTlON are expected from photochemical modeling. This, however, 10".2 10.,0 10" 10~ I would have produced no measurable ethylene line in either 10",- the CIRS or Celeste data, as shown in Figure 3(b) for CIRS. (b) In addition, the recent analysis of stellar occultation data Hr'f- acquired by UVIS yielded vertical profiles of ethylene at two different latitudes (15°2 and 42~7S) in the upper atmosphere i 10"- of Saturn (Shemansky & Liu 2012). The UV measurements .Q .§. of ethylene show good agreement with photochemical models IUIJ: ,00,· .. near the 1 J,tbar level where C2~ is produced from photolysi of :::l (!1f)l C~. Only the profile at 15°N extended to deep enough levels .Ua.J.: 10' ... (0.5 mbar) for comparison with CIRS data. Our results show C2H4 abundances in the storm at 400N two orders-of-magnitude 10" larger than that derived from UVIS observations at 15°N in 2006, well before the northern storm. The significant departure of the ethylene profile found in this study from the photochemical prediction required a thorough 50 100 150 200 250 inve tigation. eIRS data are less sensitive to temperatures at TEMPERATURE (J<) pressures less than approximately 0.1 mbar. Therefore as an Figure 7. (a) Retrieval of ethylene (green solid curve) from 2.5 cm-I CIRS additional te t, temperature profiles with significant deviations spectra from 20J I May inside the hot stratospheric beacon. The dashed curve (±20 K at pressw'es less than 0.1 mbar) were used as input indicates the abundance profile of ethylene as detemuned by photochemistry alone. The black curve shows the temperature profile retrieved from these data. when deriving the ethylene profile from the CIRS data. In (b) Retrieval of ethylene (green solid curve) from 0.1 cm-i Celeste data from all situations, it was found that there was less than a 10% 2011 May (McMath-Pierce Telescope) inside the hot stratospheric beacon. The difference between the derived ethylene profiles indicating that I· retrieved ethylene profile from the eIRS data was used as the starting point in the 950 cm-I region is not probing altitudes where the pressure this case (green dashed curve). The higher spectral resolution of the Celeste data is less than 0.1 mbar. In another test, ynthetic pectl'a were and the combination of it with the ClRS retrieved ethylene profile allow more altitude information to be reD';eved about the ethylene profile. Where there is flO calculated using a scaling factor to the ethylene photochemical I information in altitude the profiles return to the input photochemical profiles. profile. Here the hape of the pboLOcherrucal profile is pre erved, In bOlh retrievals, the uncertainty limits indicate the altitude over which the but the C2~ mole fraction is enhanced at all altitudes. In this retrieval is statistically reliable. case, the overall fit produced was of poorer quality than the fit that allowed the profile to vary cootinuously. A scaling factor of 91 was needed in order to fit the ob erved ethylene emission profile retrieved from the eIRS data was used as the a pri using the pbotochemical profile. ori profile (Figure 7 top; solid green curve). NEMESIS was The hot beacon regions in the stratosphere are a reaction to used to adjust the a priori profile between 0.03 and 3 mbar the forceful eruption of the storm into the upper tropo phere. in order to retrieve the ethylene abundance in this pressure The tropospheric storm occuned at a latitude in which quasi range. Figure 7 (bottom) shows the retrieved ethylene profile stationary planetary waves are able to propagate upward from from the Celeste data (solid green curve). Ethylene still shows the tropopause into the stratosphere (Achterberg & Flasar 1996). a greatly enhanced abundance over the photocberrucal profile Adiabatic heating in the downward phase of a wave would cause with a peak abundance of 2.7 ± 0.45 x 10-6 at 0.1 mbar. The heating of the observed magnitude (~60 K) with downward band of ethylene lines present in the Celeste data and the ini mass motion of approximately one-half of a pressure scale tial starting condition of the retrieved CIRS profile resulted in height, wbich is insufficient to cause the observed 100-fold the abundance profile being retrieved over a broader range of increase in ethylene. Radiative cooling during the descent would altitudes. The uncertainties in the profile derived from the Ce allow transport from higher altitudes, but because the radiative leste data were calculated using the same procedure as used timescale is roughly an order of magnitude larger than the in determining the uncertainties in the profile derived from the photochemical timescale for ethylene (Conrath et al. 1990), CIRS data. vertical velocities consistent with the observed temperature i 6 I --~-.-~ THE ASTROPHYSICAL.loURNAL, 759: 1 (7pp), 2012 ??? HESMAN ET AL. perturbation would produce minimal deviation of the ethylene the year following the appearance of the first storm clouds. profile from photochemical equilibrium. Thus, the enhanced These observations are key to unlocking the mystery of why ethylene emission in the beacon cannot be explained simply by ethylene is greatly enhanced in the storm region compared with transport of ethylene from higher altitudes. photochemical predictions. Ethylene is known to be primarily formed through two reaction schemes. The first being The authors thank the Cassini/CIRS calibration team for their assistance in the calibration of these data sets. The authors also CRt + hv --+ CH + H2 + H, (1) thank the McMath-Pierce Telescope staff for their assistance during these observations. This research was supported by the NASA Cassini/CIRS project, by the NASA Planetary (2) Astronomy (PAST) Program grant number NNXllAJ47G, and the NASA Cassini Data Analysis Participating Scientists where CH is the intermediate in this two-step process. The (CDAPS) Program grant number NNX12AC24G. The National second is a three-step process involving the ground-state (triplet) 3CH2: Solar Observatory is operated by the Association of Universities for Research in Astronomy under contract for the National (3) Science Foundation. Facilities: McMath-Pierce, Cassini (4) Q4 REFERENCES Achterberg, R. K., Conrath, B. J., Gierasch, P. J., Flasar, F M., & Nixon, C. A. CH + CRt --+ C2~ + H. (5) 2008, Icarus, 194,263 Achterberg, R. K., & Flasar, EM. 1996, lc:lrus, 119,350 These processes indicate that the production rate of ethylene is Allen, M., Yung, Y. L., & Gladstone, G. R. 1992, Jcarus, 100,527 directly propoltional to C~ photolysis rate at Lya (Romani Beebe, R. F, Barnel, c., Sada, P. v., & Murrell, A. S. 1992, !carus, 95, 163 et a!. 2008). Potentially, there could be a third mechanism Bezard, B., Moses, J. I., Lacy, J., et al. 2001, BAAS, 33, 1079 Borysow, A., & Frommhold, L. 1986, ApJ, 304, 849 producing ethylene: Borysow, J., Frommhold, L., & Birnbaum, G. 1988, ApI, 326, 509 Borysow, 1., Trafton, L., Frommhold, L., & Birnbaum, G. 1985, ApJ, 296, 644 (6) Conralh, B. J., & Gautier, D. 2000, Icuru~, 144.124 Conrath, B. J., Gierasch, P. J., & Leroy, S. S. 1990, icarus, 83, 255 (Allen et a!. 1992). Current laboratory measurements indicate Conrath, B. J., Gierasch, P. J., & Ustinov, E. A. 1998, [cams, 135,501 Dyudina, U. A., Ingersoll, A. P., Ewald, S. P., el a1. 2010, Geophys. R.:s. L.ett., that the energy for this reaction is too large to become significant 37, L09205 by the increase in temperature observed in the beacon. Encrenaz, T., Combes, M., Zeau, Y, VapiUon, L., & Berezne, J. 1975, A&A, The altitude range where the increased ethylene is measured 42, 355 from these data sets is primarily a sink region via the following FJasar, F M., Achlerberg, R. ~., Conrath, B. J., el a1. 2005, Sciencc, 307.1247 reaction: Fletcher, L. N., Hesman, B. E., Irwin, P. G. J., et a1. 2011, Sci~nc~, 332, [413 Hesman, B. E., Jennings, D. E., Sada, P. V., et al. 2009, learu" 202. 249 (7) Hueso, R., & Sanchez-La vega, A. 2004, lcaf\l~, 172.255 Irwin, P. G. J., Teanby, N. A., de Kok, R., el al. 2008, J. Quant. SpectrOSl·. (Romani et a\. 2008). It may be possible that this reaction is Radial. Transfer, 109, 1 136 being inhibited in the region where the significantly higher Jacquinel-Husson, N., Scott, . A., Chectin, A., et al. 2005, 1. Quam. SpcclroSlC. ethylene is observed. Because ethylene leads directly to the for R:ldiat. Transfer, 95, 429 mation of other more prominent hydrocarbons, such as ethane, Jennings, D. E., Romani, P. N., Bjoraker, G. L., et al. 2009, 1. Phy;. Chern. A, under tanding how it is created and destroyed is impol1ant to 113, 1110[ understanding the photochemistry in Giant Planet atmospheres. Lads, A. A., & Oinas, V. 199J, J. Geoph.vs. R~s., 96. 9027 Moses, J. I., Bezard, B., Lellouch, E., et a1. 2000, 1curu~, 143,244 Additional CIRS and Celeste observations acquired later Romani, P. N., Jennings, D. E., Bjoraker, G. L., et a1. 2008, kaniS, 19 ,420 in 2011 and 2012 will permit us to observe the evolution Sanchez-Lavega, A., Colas, E, Lecacheux, J., et al. 1991, Nature, 353. 397 of the emission from ethylene and other hydrocarbons in Shemansky, D., & Lill, X. 2012, Can. J. Phys., 90, 817 I I 7 ... __ 1 ~-- Queries Page 1 QI Author: The "expansion" of affiliation footnote "10" has been moved to the acknowledgments. Please check for the correct ness. Page 4 Q2 Author: Please provide year in reference citation "B. Bezard, private communication." Page 6 Q3 Author: Please confirm if the edit made to the sentence "A significant finding from these ... " retains the intended sense. Page 7 Q4 Author: Please check the details for any joumal references that do not have a pale purple link (CrossRef doi) or a blue link (NASA ADS or arXiv e-print) in the two-column proof (article style layout). A jow-nal reference with no links may contain some incorrect information. Online-only colour figures This proof PDF is identical in specification to the PDF file that will be published in the online journal. To view any online-only color figures as they will appear in the printed journal, we recommend that this color PDF file be printed on a black & white printer. _J