1 The Aeronomy of Ice in the (AIM) Mission: Overview and early science results

2 Corresponding author: James M. Russell III*a

3 Co-authors: Scott M. Baileyb, Larry L. Gordleyc, David W. Ruschd, Mihály Horányid, Mark E.

4 Hervigc, Gary E. Thomasd, Cora E. Randalld, David E. Siskinde, Michael H. Stevense, Michael E.

5 Summersf, Michael I. Taylorg, Christoph R. Englerte, Patrick J. Espyh, William E. McClintockd

6 and Aimee W. Merkeld

7 *a Center for Atmospheric Sciences, Hampton University, Hampton, VA 23668, USA,

8 [email protected], phone (757)728-6893, fax (757)727-5090

9 b Bradley Department of Electrical and Computer Engineering, Virginia Polytechnical Institute

10 and State University, Blacksburg, VA 24061, USA

11 c GATS, Inc., 11864 Cannon Boulevard, Suite 101, Newport News, VA 23606, USA

12 d Laboratory for Atmospheric and Space , 1234 Innovation Drive, Boulder, CO 80303,

13 USA

14 e Space Science Division, Naval Research Laboratory, Washington DC, 20375, USA

15

16 f Department of Physics, George Mason University, Fairfax, VA, 22030, USA

17

18 g Department of Physics, Utah State University, Logan, Utah 84322, USA

19

20 h Department of Physics, Norwegian University of Science and Technology, Høgskoleringen 5,

21 NO-7491, Trondheim, Norway

22

23

1 24 Potential Reviewers:

25 Matthew DeLand ([email protected])

26 Eric Shettle ([email protected])

27 Uwe Berger ([email protected])

28 Svetlana Petelina [email protected]

29 Abstract

30 The Aeronomy of Ice in the Mesosphere (AIM) mission was launched from Vandenberg Air

31 Force Base in California at 1:26:03 PDT on April 25, 2007 becoming the first mission

32 dedicated to the study of polar mesospheric clouds. A Pegasus XL rocket launched the satellite

33 into a near perfectly circular 600 km sun synchronous . AIM carries three instruments

34 specifically selected because of their ability to provide key measurements needed to address the

35 AIM goal which is to determine why these clouds form and vary. The instrument payload

36 includes a nadir imager, a solar occultation instrument and an in-situ cosmic dust detector.

37 Detailed descriptions of the science, instruments and observation scenario are presented. Early

38 science results from the first northern and southern hemisphere seasons show a highly variable

39 cloud morphology, clouds that are ten times brighter than measured by previous space-based

40 instruments, the presence of a previously suspected but never before measured layer of small

41 particles believed to be the cause of summertime radar echoes and a mesospheric ice layer that

42 extends from below the main northern hemisphere peak at 83kmup up to ~90km in one

43 continuous layer. Additionally, complex features were observed that are reminiscent of

44 tropospheric phenomena.

45 Key Words: Polar Mesospheric Clouds, Mesosphere, Remote Sensing, Gases, Particles, Global

46 Change

2 47 1. Introduction

48 Polar Mesospheric Clouds (PMCs) are ’s highest clouds, occupying the atmospheric region

49 near the high-latitude summer mesopause at ~ 83 km. They typically extend from 55° latitude to

50 the geographic pole, occurring in each summer in both hemispheres between about mid-May to

51 mid August in the North and mid-November to mid-March in the South. The polar upper

52 mesosphere becomes the coldest place on earth around summer solstice, when the temperature

53 plummets below 130K. Known as noctilucent, or “night-shining” clouds (NLCs) as seen by

54 ground based observers (Figure 1.), their name derives from the fact that they are seen at evening

55 or morning twilight, when the lower atmosphere is in darkness, but the upper atmosphere is still

56 sunlit. First identified over 120 years ago (Leslie, 1885), their nature as water-ice crystals was

57 not established until relatively recently [Hervig et al., 2001]. There is great interest in PMCs

58 because they are changing in ways that are not understood. They are appearing more frequently,

59 becoming brighter and they are being seen at lower latitudes than ever before [Deland et al.,

60 2006, 2007; Taylor et al., 2002]. This behavior suggests a possible connection with global

61 change occurring at lower altitudes.

62 Due to their sensitivity to changes in the environment, PMCs are expected to respond to long-

63 term global change, and may be a modern phenomenon associated with the rise of greenhouse

64 gases in the industrial era [Thomas et al., 1989]. Indeed, an increase of ~5% per decade in PMC

65 brightness has occurred over the past 27 years [DeLand et al., 2007]. Additional interest in

66 mesospheric clouds stems from the unique conditions of cloud formation in the weakly-ionized

67 D-region (a ‘dusty plasma’), and their close relationship with a mesospheric layer of strong radar

68 reflectance (Polar Mesospheric Summer Echoes). At least three factors are believed to control

69 PMC formation: (1) temperature, (2) water vapor, and (3) cosmic dust or some other nucleation

70 source. The detailed role of water vapor has not yet been established, although it is of critical

3 71 importance in the cloud saturation and growth processes. There is now evidence that the clouds

72 redistribute water vapor [Summers et al., 2001], which in turn could affect the ozone

73 concentration and heat balance [Siskind et al., 2005]. There are many open questions regarding

74 cloud nucleation, microphysics (Rapp and Thomas, 2006), chemistry, gravity waves, transport,

75 and cosmic dust influx. Despite the sophistication of present-day modeling [e.g. Berger and

76 Lübken, 2006], these uncertainties hamper our ability to understand the origin of the observed

77 changes, nor can we project with confidence future changes. In short, we do not understand what

78 causes a mesospheric cloud to form, or how it evolves. To establish the basis for new predictive

79 PMC models, high precision measurements of PMCs and key chemical constituents are required

80 for various conditions of temperature, cosmic dust influx, and H2O concentrations.

81 The Aeronomy of Ice in the Mesosphere (AIM) satellite mission was launched from Vandenberg

82 Air Force Base on April 25, 2007 with the overall goal to resolve why PMCs form and why they

83 vary. A Pegasus XL rocket placed the AIM satellite into a near circular (601 km apogee, 595 km

84 perigee), 12:00 AM/PM sun-synchronous orbit. By measuring PMCs and the thermal, chemical

85 and dynamical environment in which they form, AIM will quantify the connection between these

86 clouds and the of the polar mesosphere. In the end, this will provide the basis for

87 study of long-term variability in mesospheric and its relationship to global change. The

88 results of AIM will be a rigorous test and validation of predictive models that then can reliably

89 use past PMC changes and current data to assess trends as indicators of global change. This goal

90 is being achieved by measuring PMC abundances, spatial distribution, particle size distributions,

91 gravity wave activity, cosmic dust influx to the atmosphere and precise vertical profiles of

92 temperature, H2O, CH4, O3, CO2, NO, and aerosols.

93 The AIM science team has established a set of six objectives designed to address the mission

94 goal. The first objective is to understand the cloud microphysics, i.e. the relationship between

4 95 temperature, water vapor and size distribution under a variety of conditions and times in the

96 PMC season, and for different cloud types. The second objective is focused on assessing the role

97 of atmospheric gravity waves in PMC formation. It is clear that waves are important because

98 their presence is obvious in the many ground-based photographs of the clouds. The third

99 objective addresses the question of the influence of temperature in controlling the length of the

100 PMC season. Clearly, extremely cold temperatures (on the order of 140K or less) are essential

101 because where the clouds, form fifty miles above the surface, the pressure is one hundred

102 thousand times less than at the surface and the air is one hundred thousand to a million times

103 drier than surface desert air. The fourth objective addresses upper mesosphere water vapor

104 chemistry and how it changes during the PMC season. As noted earlier, the clouds are known to

105 be made up of water ice crystals and interactions between H2O and O3 and the role of solar

106 Lyman-α radiation which controls water vapor at the cloud altitude needs to be understood. The

107 fifth objective is to asses the sources of seed particles needed for PMC formation by providing a

108 nucleation site to begin the particle growth process. The last objective combines results from the

109 first five to validate coupled chemistry, dynamics and PMC microphysical models and to assess

110 the ability to quantify long-term changes, and predict future changes in PMCs based on these

111 models and historical data sets.

112 The AIM experiment includes three instruments: SOFIE (Solar Occultation For Ice Experiment),

113 an infrared (IR) solar occultation differential absorption radiometer; CIPS (Cloud Imaging and

114 Particle Size experiment), a panoramic UV imager; and CDE (Cosmic Dust Experiment), an in-

115 situ cosmic dust detector. In this paper we describe the AIM mission including the instruments

116 and their data products and the overall mission implementation.

5 117 2. The AIM Science Instruments

118 The Solar Occultation for Ice Experiment (SOFIE)

119 SOFIE (Fig 2.) is performing satellite solar occultation measurements to determine vertical

120 profiles of temperature, pressure, the abundance of five gaseous species (H2O, O3, CO2, CH4, and

121 NO) and PMC particle extinction at eleven wavelengths. Occultation measurements are

122 accomplished by monitoring solar intensity as the satellite enters or exits the Earth’s sunlit side.

123 The ratio of solar intensity measured through the atmosphere (V) to the exoatmospheric solar

124 intensity (Vexo) yields atmospheric transmissions, which are the basis for retrieving the desired

125 geophysical parameters. SOFIE uses differential absorption radiometry, with eight channel pairs

126 covering wavelengths (λ) from 0.29 to 5.32 µm (see Table 1). Gordley et al., [2008] and Hervig

127 et al., 2008, provide details of the SOFIE instrument design, characteristics, calibration, retrieval

128 methods and on-orbit performance in companion papers in this special issue. Specific gases are

129 targeted by measuring solar intensity in two wavelength regions; one where the gas is strongly

130 absorbing and an adjacent region where the gas is weakly absorbing. SOFIE measures the strong

131 and weak band radiometer signals (VS and VW) and the difference of these which is amplified by

132 an electronic gain G∆V to give the relation:

133 ∆V = (Vw – Vs) G∆V (1)

134 While a single band measurement alone can be sufficient to retrieve gas mixing ratios in the

135 lower mesosphere and , SOFIE will characterize the tenuous regions where PMCs

136 form extending into the lower . At these altitudes, atmospheric and gaseous

137 abundances are low and the corresponding signals can approach typical noise levels.

138 While a single band measurement alone can be sufficient to retrieve gas mixing ratios in the

139 lower mesosphere and stratosphere, SOFIE will characterize the tenuous regions where PMCs

140 form extending into the lower thermosphere. At these altitudes, atmospheric and gaseous

6 141 abundances are low and the corresponding signals can approach the noise levels. However,

142 common mode optics errors and electronics errors can be greatly reduced by measuring the

143 difference in intensity between adjacent strong and weak absorption bands. Benefits are realized

144 because a variety of atmospheric and instrumental effects are nearly identical in the strong and

145 weak bands, and are therefore are greatly reduced by differencing band pairs. For example, Fig.

146 3 shows the SOFIE water vapor channel band pair locations indicated by the orange horizontal

147 bars, along with calculated transmission spectra for relevant gaseous absorbers and PMCs. Note

148 that PMC absorption is nearly identical in the H2O weak and strong bands, and thus will be

149 almost completely removed in the difference signal measurement. Another benefit is that

150 electronic gain can be applied to the difference signals, allowing the measurement precision to be

151 enhanced to a level consistent with the detector noise. Six SOFIE channels are designed to

152 measure gaseous signals, and two are dedicated to particle measurements. Particle measurements

153 are also obtained from the gas channel weak bands. Measurements in two CO2 bands will be

154 used to simultaneously retrieve temperature and CO2 mixing ratio [Gordley et al., 2008]. The

155 SOFIE sun sensor provides measurements of atmospheric refraction angle with sub-arc second

156 knowledge precision that is being used to retrieve temperature profiles at altitudes below about

157 50 km.

158 SOFIE Measurement Geometry

159 SOFIE provides spacecraft sunset measurements at latitudes between about 65° and 85°S and

160 spacecraft sunrise measurements at latitudes between about 65° and 85°N [See Gordley et al.,

161 2008]. SOFIE observes 15 sunrise and 15 sunset occultations per day, and consecutive sunrises

162 or sunsets are separated by ∼96 minutes in time or ∼24° in longitude. The SOFIE field of view

163 (FOV) is 1.8 arc minutes vertical by 6.0 arc minutes horizontal which corresponds to about 1.5 x

164 5 km, respectively, at the tangent point. The SOFIE measurements, consisting of 16 radiometer

7 165 and 8 difference signal measurements, are sampled at 20 Hz, which corresponds to a vertical

166 distance of 150 m in the atmosphere. The vertical resolution of 1.5 km combined with the 3

167 km/sec solar sink or rise rate sets the natural frequency of the data set at ∼2 Hz, which is the rate

168 at which the FOV vertical dimension is swept through the atmosphere.

169 SOFIE Instrument Performance

170 SOFIE performance was characterized in laboratory calibration before launch and now detailed

171 characterizations have been done in orbit. In-orbit performance is excellent with noise levels at

172 or below laboratory values. Ground radiative calibration sources included a solar emulator

173 blackbody (SEB) that achieved an emitting temperature of ∼3000K. Radiance from the SEB was

174 equivalent to 28% - 46% of the exoatmospheric solar signal in bands 5 – 16, and less than 8% of

175 the exoatmospheric signal in bands 1 - 4. Bands 1 - 4 were stimulated with a xenon lamp. The

176 Sun was viewed in the laboratory by directing the solar image into the SOFIE aperture using

177 external pointing mirrors. Calibration sequences addressed important instrument characteristics

178 including measurement background and noise, FOV, response linearity, relative spectral

179 response, time response, absolute gain, and difference signal gain. Because the measurements

180 are used to determine transmissions (τ = V / Vexo), absolute radiometric calibration is not

181 important, except to ensure that the exoatmospheric solar view generates signals near the upper

182 limit of the data acquisition system. In all cases the measured instrument characteristics were

183 within design specifications based on AIM science requirements. Expected on-orbit

184 performance of the SOFIE measurement/retrieval system was characterized by simulating the

185 SOFIE measurements and then retrieving the desired geophysical parameters from these signals.

186 These exercises included key SOFIE performance characteristics measured during calibration

187 and testing, and were used to determine salient characteristics of the retrieved parameters. The

188 retrieved quantities, measurements used, and expected altitude range of reliable retrievals are

8 189 summarized in Table 2. Gordley et al. [2008] give more detail on the bands used and their

190 properties.

191 SOFIE Physical PMC Properties

192 SOFIE multi-wavelength PMC extinction measurements are used to infer a variety of physical

193 cloud properties including ice mass density, particle shape, and particle size distribution [Hervig

194 et al., 2008a]. PMCs are identified using 3.19 µm extinction profiles based on a threshold

195 approach considering the combined measurement and retrieval noise floor (10-7 km-1). The

196 infrared (IR) measurements are due entirely to absorption and are thus proportional to the

197 particle radius cubed. As a result, the IR measurements are directly proportional to particle

198 volume density (or ice mass density). The conversion from extinction to ice mass density is

199 based on theoretical relationships that have total uncertainties (due to uncertainties in the

200 refractive index and particle shape and size) of between 2 and 10% for λ > 2.7 µm. Particle

201 shape is determined by comparing measured extinction ratios to modeled signals from clouds

202 comprised of oblate and prolate spheroids with varying axial ratios (AR). Certain SOFIE spectral

203 band extinction ratios within the IR are sensitive to AR and nearly independent of particle size.

204 The SOFIE measurements yield both oblate (AR > 1) and prolate (AR < 1) solutions for AR < ∼4,

205 and a unique solution when AR > ∼4. The unique combination of UV thru IR PMC

206 measurements provides excellent information concerning the PMC size distribution. The size

207 distribution retrievals assume the unimodal Gaussian form, which describes the concentration

208 versus radius in terms of the total particle concentration (N), mode radius (rm), and distribution

209 width (∆ r). Theoretical uncertainties (assuming perfect measurements) in N, rm, and ∆ r are less

210 than a few percent for 20 < rm < 180 nm and 5 < ∆ r < 25 nm. SOFIE results are discussed later.

9 211 The Cloud Imaging and Particle Size Experiment (CIPS)

212 The CIPS instrument is a panoramic UV imager. CIPS provide images of the atmosphere at a

213 variety of scattering angles in order to remove the Rayleigh scattering component of the signal,

214 determine cloud presence, infer particle sizes, provide the spatial morphology of the clouds, and

215 constrain the parameters of cloud particle size and size distribution. CIPS is providing:

216 • Panoramic nadir imaging at 265 nm with a 120º x 80º field-of-view and 5 km X 5 km

217 spatial resolution

218 • Scattered radiances from Polar Mesospheric Clouds near 83 km altitude to derive PMC

219 morphology and cloud particle size information.

220 • Multiple exposures of individual clouds at various scattering angles to derive PMC

221 scattering phase functions.

222 • High-spatial resolution measurements of the variation of the Rayleigh scattered

223 background that provides information on gravity wave activity in the 55-km region and

224 small scale variations in lower mesospheric ozone.

225 CIPS Instrument Design

226 The CIPS instrument consists of an array of four cameras (Figure 4) operating with a 15 nm

227 passband centered at 265 nm. Each camera has an overlapping FOV and a pixel size at the nadir

228 of ~ 2 km. The FOV of the camera system is 80º x 120º, centered at the sub-satellite point, with

229 the 120º axis along the orbit track (Fig. 5). Because of slant viewing, the spatial resolution

230 increases to ~6.4 km near the edge of the FOV of the forward and aft cameras. The combination

231 of images from the four cameras is referred to as a scene. CIPS records scenes of atmospheric

232 and cloud radiance in the summer hemisphere from the terminator to ~ 40o latitude along the

233 sunlit portion of the orbit. The near-polar orbit and cross-track FOV will cause the observation

10 234 swaths to overlap at latitudes higher than about 70o, so that nearly the entire polar cap will be

235 mapped daily by the 15-orbit per day coverage (Fig. 6). The CIPS images cover up to about 85

236 degrees latitude in each hemisphere.

237 Each camera has a focal ratio of 1.12, a focal length of 28 mm, a 25 mm lens diameter and

238 includes an interference filter and a Charge Coupled Device (CCD) detector system. The

239 throughput of the optical elements and their sizes are designed for a ±1% measurement precision

240 of the background sunlit Earth. The custom UV filters were manufactured by Barr associates and

241 centered at 265 nm. The CCD detectors are coupled with Hamamatsu V5181U-03 image

242 intensifiers (40 mm diameter active area) and have 2048 x 2048 useful pixels that are

243 electronically binned in 4 x 8 combinations for an effective 340 (cross track) x 170 (along track)

244 pixel images. The signal in each pixel is digitized to 12 bit resolution. On average, 26 images are

245 produced per orbit in the summer polar region with special ‘first light’ images just beyond the

246 terminator.

247 Imaging is achieved with this body-fixed camera assembly using an exposure time of 1 second,

248 which, when combined with the field-of-view, yields the nadir spatial resolution of ~2 km.

249 Between four and seven exposures of the same cloud volume are made during a satellite

250 overpass, at a rate of one scene every 46 sec. Each CCD is equipped with a Digital Signal

251 Processing (DSP) interface that incorporates a lossless Huffman compression algorithm,

252 reducing data volume by about a factor of two. Therefore each scene produces 523 Kbytes of

253 data yielding approximately 18 Mbytes per orbit. The CIPS requirements and measured

254 performance characteristics are listed in Table 3. As the AIM satellite moves in orbit, the CIPS

255 instrument provides seven images of the same region of the atmosphere at seven different

256 scattering angles from which the PMC scattering phase function is derived.

257

11 258 CIPS PMC High Resolution Structure

259 PMCs are identified as enhancements of brightness against the Rayleigh-scattered background

260 coming from the lower atmosphere. Thomas et al. [1991] proved the feasibility of this detection

261 method using 273.5 nm data from the Solar Backscattered Ultraviolet (SBUV) nadir-viewing

262 spectrometer on board the NIMBUS 7 spacecraft. They showed that the brightest PMCs could be

263 distinguished against the background despite the clouds under-filling the 175 x 175 km FOV.

264 One key to detecting relatively weak PMCs is to observe at wavelengths where the

265 Earth's albedo is at or near a minimum due to ozone absorption. The 265 nm central wavelength

266 of CIPS was chosen because it represents a peak in ozone absorption and therefore a minimum in

267 the background originating from scattering of sunlight from air molecules. This observed

268 background radiance is generally limited by ozone to scattering in the vicinity of 55-km and

269 above. A second key feature is the measurement of cloud brightness at small (< 35º) scattering

270 angles that greatly enhances the cloud contrast with respect to the Rayleigh background because

271 of the forward-scattering property of ice particles in the Mie regime.

272 CIPS observations at solar zenith angles (SZA) from 87 deg to about 94 deg (the 'shadow band')

273 experience a reduction in background signal by at least a factor of 10 from full daylight

274 brightness while the higher altitude PMCs remain at least 95% illuminated relative to an

275 overhead sun condition. This results in greatly enhanced PMC contrast because the background

276 is mostly in darkness. The shadow band is roughly 700 km wide and centered on the location of

277 SOFIE coincident observations. We refer to that location as the 'Common Volume' because both

278 CIPS and SOFIE observe the same volume of atmosphere within a period of six minutes.

279 The bright PMCs are confined to a summertime period about 70 days long, and mostly to a

280 region poleward of the Arctic and Antarctic circles [Thomas et al., 1989]. The CIPS observing

281 strategy allows for adequate collection of data from cloud free-regions, both before and after the

12 282 cloud 'season’, and at lower latitudes during the season (<55o). These regions are being used to

283 characterize signatures of small scale wave transport of energy from the lower atmosphere –

284 gravity waves, as well as planetary scale waves.

285 Gravity wave effects on background CIPS signals result from the dependence of ozone

286 photochemistry on temperature. Temperature fluctuations associated with waves can alter the

287 ozone density. The resulting effect on the UV albedo in the highly-absorbing Hartley bands of

288 ozone is easily found from a single-scattering calculation to be linearly dependent upon the

289 ozone perturbation. Multiple scattering is negligible at 265 nm. Wave amplitudes as small as 1-2

290 K at 55 km will cause approximately a 3% perturbation in background signal, a change easily

291 detectable by CIPS above the otherwise smoothly-varying background due to Rayleigh

292 scattering. This capability has led to identification of gravity wave signatures in the CIPS data

293 [Chandran et al., 2008].

294 Clouds are identified by their large angular-dependent scattering signature, a well-established

295 property of PMC particles [Thomas and McKay, 1986; Gumbel and Witt, 1998]. Cloud ice

296 particles are strongly forward scattering. This effect is more pronounced for brighter clouds, or

297 more specifically clouds having larger particle size. The background Rayleigh scattered sunlight

298 follows a well known variation with scattering angle that is symmetric about 90° scattering

299 angle. Any brightness enhancement that shows forward scattering behavior can be identified as a

300 PMC, and not an underlying background irregularity. This leads to a methodology for separation

301 of the signature of the cloud albedo from that of the underlying background. The scene recorded

302 by CIPS includes a Rayleigh scattered sunlight background at altitudes near 50 km and Mie

303 scattered sunlight by PMC particles at ~ 83 km. The magnitude of the background is controlled

304 by ozone which absorbs radiation along the path. For solar zenith angles away from the

305 terminator, ozone limits contributions to the observed CIPS radiance to altitudes only above 50

13 306 km. The crux of CIPS measurement interpretation is determining the PMC signal by subtracting

307 the Rayleigh background. CIPS determines the scattering phase function by imaging the same

308 point in space at multiple angles. In one approach, the background signal can be inferred using

309 the difference between cloud and molecular scattering phase functions. Alternately, the

310 background can be calculated given profiles of temperature, pressure, and ozone taken from

311 SOFIE measurements within the common volume.

312 Given the water-ice composition [Hervig et al., 2001] and the shape factor of the irregular ice

313 particles [see Rapp et al., 2007 for further discussion of the shape issue], the CIPS angular

314 distributions of brightness at a single wavelength yield column mass, surface area, and mean

315 particle size. These results allow correlation of PMC properties with PMC extinction,

316 temperature, water vapor, and other atmospheric parameters. The particle size distribution can be

317 more constrained by combining the CIPS observations with SOFIE measurements of PMC

318 optical depths in the common volume rather than using the data from one instrument alone.

319 The Cosmic Dust Experiment (CDE)

320 CDE monitors the variability of cosmic dust influx into our atmosphere. It is an in situ dust

321 impact experiment, however the time for an incoming micrometeoroid from low Earth orbit to

322 reach the altitude region of 80 to 85 km is very short ( < 1 min). Hence CDE measurements

323 directly indicate the influx of cosmic matter to the mesosphere. CDE observations integrated

324 over several days are expected to show the temporal variability of the cosmic dust influx that

325 could influence the formation and evolution of PMCs, and results will be used to establish

326 possible correlations with remote sensing observations of PMC brightness and coverage, for

327 example. CDE is an impact dust detector based on thin (28 µ m) permanently polarized

328 polyvinylidine fluoride (PVDF) foil sensors [Simpson and Tuzzolino,1985]. PVDF dust

329 detectors were flown on the VEGA 1 and 2 missions to comet Halley [Simpson et al., 1986], on

14 330 the STARDUST mission to comet Wild 2 [Tuzzolino et al., 2004], on the Cassini mission to

331 Saturn [Srama Et al., 2004], and the Earth observing ARGOS satellite [Tuzzolino Et al., 2005].

332 A copy of the CDE instrument is on the New Horizons mission to Pluto [Horanyi et al., 2007].

333 PVDF sensors have many advantages: they require no bias or high accelerating voltages, they are

334 electrically, mechanically and thermally stable, radiation resistant, do not respond to charged

335 particles, and can measure very high rates of dust impacts. The disadvantage of this sensor is

336 that the charge signal generated by an impact is a function of both the mass and the velocity of

337 the dust grain. Hence, in the analysis of the CDE data one has to assume a characteristic

338 meteorite speed and average flight angle with respect to the local zenith in order to infer

339 meteorite mass. A similar approach was used in the analysis of the impact craters on the Long

340 Duration Exposure Facility (LDEF) [Love and Brownlee, 1993]. The charge Q (in units of

341 electron charge) generated by a particle with mass m, (in units of grams), and relative velocity v

342 (measured in km/s) is: Qmv=×3. 8 1 017 1.3 3 343

344 The size of an individual PVDF detector ultimately determines the mass threshold as well as the

345 expected impact rate. The capacitance of the PVDF film is proportional to its surface area and, in

346 turn, the electronic noise level is set by the capacitance. The required mass (charge) threshold

347 ultimately sets the maximum size of an individual PVDF patch. The CDE science goal of

348 measuring an expected impact rate of ≥ 100 hits/week requires a mass threshold of ≤ 4x10-12 g

349 and a total sensitive surface area of ≥ 0.1 m2. To meet this requirement CDE (Figure 7) consists

350 of 12 active PVDF patches with surface areas of 85 cm2 each. In addition, there are two

351 additional sensors (identical to the front side patches), on the back side of CDE that cannot be hit

352 by dust. These reference detectors will be used to measure the noise background. The CDE

353 dynamic range provides mass resolutions within a factor of ≤ 3 in the mass range of 4x10-12g ≤

15 354 m ≤ 4x10-9g covering an approximate size range in particle radius of 0.8 µ m ≤ a ≤ 8

355 µ m.

356 Each of the 14 sensors has an adjustable threshold to optimize CDE operations. To follow the

357 possible degradation of its performance due to ageing, CDE has onboard calibration capabilities

358 for its electronics. Internal signals can be injected in each of the 14 channels with amplitudes

359 covering its entire dynamical range. Each dust hit generates a science event, where the time,

360 channel number and the impact charge are recorded, in addition to all relevant housekeeping

361 data. Using appropriate averages this can be turned into a time-dependent global map to show the

362 possible spatial and temporal variability of the amount of cosmic dust entering the atmosphere.

363 3. AIM Observation Scenario

364 Occultation and imaging measurements have inherent benefits, but also present unique

365 challenges to measurement interpretation. For example, CIPS offers excellent horizontal

366 resolution but little height information, while SOFIE provides excellent vertical resolution but

367 coarse horizontal resolution. Combining observations allows mitigation of certain measurement

368 limitations and offers opportunities to increase the information provided by each data set. Thus,

369 the AIM observation scenario provides space coincident CIPS and SOFIE observations on every

370 orbit. The SOFIE tangent point location and footprint defines the AIM Common Volume (CV).

371 CIPS will view the CV roughly six minutes before or after SOFIE, which is well under the 20-30

372 minute time scale of PMC variability [see e.g. Gadsden, 1982]. The SOFIE horizontal footprint

373 occupies 280 km along the limb by 5 km across-limb. Limb retrievals traditionally assume that

374 all atmospheric absorbers are distributed in a spherically symmetric fashion. While spherical

375 symmetry is an appropriate assumption when measuring gaseous species, PMCs can vary over

376 smaller scales than accommodated by the limb path geometry [Fritz, et al., 1993]. If ignored,

377 cloud inhomogeneity can result in fundamental errors, although the sign of the errors is at least

16 378 predictable. Variable cloud cover will always result in an underestimate of the instantaneous

379 cloud extinction because the measurement inversion assumes that the absorber occupies the

380 entire tangent path length. A cloud within the line of sight but not at the tangent point will

381 always be assigned an erroneously low altitude because its position is assumed to be at the

382 tangent point. SOFIE measurement interpretation will address these issues using CIPS images to

383 provide the detailed horizontal distribution of PMCs.

384 The AIM spacecraft will accommodate CV measurements by transitioning from sun pointing

385 during the SOFIE occultation to an Earth inertial pointing mode for the CIPS imaging period.

386 This orientation also allows SOFIE and CIPS to view though the same 2-D plane, with CIPS

387 viewing multiple angles as it passes over the occultation tangent point. Earth rotation (~1°,

388 typically < 50 km) will prevent perfectly coincident planar viewing, but this is unimportant

389 considering the natural horizontal resolution of the SOFIE sample volume (~280 km).

390 4. Spacecraft Status

391 All AIM spacecraft systems have been functioning nominally since launch on April 25, 2007

392 except for the command receiver. The receiver has had periods of intermittent command signal

393 rejection, but the AIM Flight Operations team has been able to successfully work around these

394 difficulties. Full science operations began on May 22, 2007. Routine science data processing

395 started in February 2008. All data products are available to the public via the internet at the

396 main AIM webpage at aim.hamptonu.edu.

397 5. Early Science Results

398 Brief highlights of early science findings are presented in this section and more results as well as

399 additional details are contained in ten other AIM papers in this special issue. AIM is providing

400 the first global scale view of PMCs in the polar cap region with a spatial resolution that is

401 unprecedented. Results show that the clouds are highly variable from orbit to orbit and day to

17 402 day. Figures 8 and 9 show daily mean CIPS images on two successive days in late June and

403 early July 2007. As discussed above, these images have had the Rayleigh background signal

404 removed so that what remains is only PMC signatures. Further details of how the background

405 removal is accomplished are provided in papers by Rusch et al. [2008] and Bailey et al. [2008] in

406 this issue. Note the significant increase in the number of clouds from June 30 to July 1st. The

407 clouds on June 30th are not as bright and cover a much smaller area than on the next day. Also

408 note the very pronounced void in the cloud field in the first quadrant of Fig. 9. We refer to this

409 feature as an ice void. These features are currently unexplained but are common occurrences in

410 the CIPS images. They could be caused by energy deposition (heating) through gravity wave

411 breaking, but they may also occur due to turbulent mixing from vertically displaced air (which

412 can also be due to gravity wave breaking). Ice voids appear in circular or oval shapes with

413 diameters that vary from ten to several hundred kilometers or greater. Rusch et al., [2008]

414 discuss this feature in more detail. These structures bear a startling similarity to those seen in

415 tropospheric clouds and suggest that the mesosphere may share some of the same dynamical

416 processes responsible for weather near Earth’s surface. If this similarity holds up under further

417 analysis, it introduces an entirely different view of potential mechanisms responsible for PMC

418 formation and variability than existed before AIM was launched.

419 The multi-wavelength SOFIE observations are shedding new light on PMCs because of their

420 extremely high sensitivity and signal-to-noise. While previous instruments suggested that ice is

421 present about half the time near 70°N, SOFIE now shows that ice is almost always present, i.e.

422 100% of the time, at these latitudes during the PMC season. Hervig et al. [2008b] compare a time

423 series of SOFIE ice occurrence frequency to SME and HALOE results. They check their results

424 by downgrading SOFIE sensitivity to be consistent with HALOE and SME and find good

425 agreement with results from these experiments, thus confirming that the lower ice frequencies

18 426 recorded by the older were a result of reduced sensitivity. SOFIE is 15 - 200 times

427 more sensitive to PMC particles than previous instruments, including lidars. In addition, it

428 contains channels located at the peak in the ice absorption spectrum (~3 µm wavelength) where

429 PMC extinction is ~100 times greater than for surrounding NIR and IR wavelengths.

430 The time versus height cross section of SOFIE H2O in Figure 10 illustrates the development of

431 water vapor in the PMC environment during the 2007 northern cloud season. Water vapor, an

432 essential gas for PMC formation, is measured at a vertical resolution of 1.6 km and precision of

433 better than 21 ppbv. The seasonal evolution of water vapor is characterized by a rapid mixing

434 ratio rise in the beginning of the season and a steady increase after that time which is most likely

435 caused by vertical transport due to upwelling. SOFIE simultaneous O3 observations shown in

436 Fig. 10 indicate a mesospheric ozone minimum at about 81 km altitude at the beginning and end

437 of the season and near 79 km throughout most of the season, with a peak observed near 90 km.

438 O3 enhancement is observed at altitudes from roughly 80 to 95 km during times that ice is

439 present. For example, at 92 km the mixing ratio increases from 0.8 ppmv on 22 May to ~1.1

440 ppmv by 11 July, returning to ~0.8 ppmv by 15 August. This change in ozone seems to be

441 consistent with the suggestion by Sisknd et al [2007] that ozone should be enhanced above the

442 ice layer due to dehydration resulting in lowered HOx. These observations of highly transient

443 features, combined with the other SOFIE products are giving a first-time detailed look at the

444 physics of PMCs.

445 An example of the quality of the underlying H2O profiles is demonstrated in Figure 11 which

446 shows a set of retrieved water vapor profiles for August 12, 2007. The various profiles indicate

447 where PMCs have formed as evidenced by reduced water vapor near 85 km, and where PMCs

448 have sublimated resulting in enhanced water vapor just above 80 km. The abrupt decrease in

449 H2O above PMC altitudes (83 km) is consistent with a loss of water vapor to ice growth, and also

19 450 with destruction of H2O by increasing photolysis at higher altitudes. These findings point to a

451 fundamental connection between ice particle characteristics and water vapor.

452 SOFIE has also demonstrated that mesospheric ice exists as one continuous layer extending from

453 ~80 km altitude to the mesopause (~87 km) and often higher, as shown in Figure 12. This view

454 is supported by model predictions [von Zahn and Berger, 2003; Rapp and Thomas, 2006], but is

455 in contrast with previous measurements that were not sensitive enough to detect the most tenuous

456 ice populations. The SOFIE results now confirm a previously suspected but never before

457 directly measured layer of small ice particles believed to be responsible for the high latitude

458 summer phenomenon called Polar Mesosphere Summer Echoes or PMSEs [see e.g. Rapp and

459 Lubken, 2004].

460 These results represent only a few of the rich findings about the properties of PMCs that have

461 come from the first northern hemisphere season of observations. AIM has also just completed

462 observations of the first southern hemisphere season and already conventional ideas about inter-

463 hemispheric differences in cloud properties are being challenged. The mission is providing high

464 quality data on particle size, size variation with altitude, particle shape and its altitude

465 dependence, characteristics that describe the onset and end of the PMC season, the interplay

466 between particles, water vapor and temperature, brightness variability over the entire polar cap

467 region and space and time variability. The unprecedented sensitivity and spatial resolution

468 provided by the AIM instruments is providing a rich data base to address all six AIM objectives

469 and to answer the question of why these clouds form and vary.

470 6. Summary

471 Much has been learned about PMCs from a number of satellite missions [see e.g. Deland et al.,

472 2006]. These studies were all serendipitous in the sense that none were conceived or designed to

20 473 address PMC phenomena. AIM is the first mission dedicated to the exploration of PMCs. All

474 systems designed to run the satellite are functioning as planned except for the command receiver

475 and work arounds have been established for this subsystem. All instruments are functioning

476 nominally. The appropriate mix of measurement capabilities and orbit parameters has been

477 brought together to answer the fundamental question of why these clouds form and vary. The

478 three science instruments are providing data needed to address a series of six objectives

479 formulated by the AIM science team. Initial science results from the mission have significantly

480 advanced our understanding of PMC properties, variability, and spatial distribution after one

481 northern hemisphere season. The presence of voids in the ice layer observed by CIPS appears to

482 be similar to a tropospheric weather phemenon and if true, it adds a potentially new mechanism

483 that must be understood in addressing the question of why these clouds form and vary.

484 Acknowledgements

485 The authors acknowledge the many contributions of expertise, time, talent and dedication given

486 to the mission by the team of engineers, scientists, and managers of organizations supporting

487 AIM at Hampton University, the University of Colorado Laboratory for Atmospheric and Space

488 Physics, the Space Dynamics Laboratory of Utah State University, Orbital Sciences Corporation,

489 GATS, Incorporated, the Goddard Space Flight Center and the Kennedy Space Center. Special

490 thanks is given to the NASA Headquarters AIM Program Executive, Victoria Elsbernd and the

491 AIM Program Scientist, Mary Mellott for their exceptional contributions to AIM through their

492 sincere and extraordinary dedication to the success of this mission right from its inception. We

493 also thank the AIM flight operations team at LASP, Orbital Sciences Corporation and GATS,

494 Inc. for their commitment to excellence and their duty beyond the reasonable at times in order to

495 make the mission succeed. It would not have been possible to present the exceptional data in this

496 and other AIM papers in this special issue without the support of these individuals and

21 497 organizations. Funding for the AIM mission was provided by NASA’s Small Explorers Program

498 under contract NAS5-03132.

499 References

500 Bailey, S. M., G. E. Thomas, D. W. Rusch, A. W. Merkel, C. Jeppesen, J. N. Carstens, C. E.

501 Randall, W. E. McClintock, J. M. Russell III, Phase Functions of Polar Mesospheric Cloud

502 Ice as Observed by the CIPS Instrument on the AIM Satellite, J. Atmos. Solar-Terr. Phys., in

503 review, 2008.

504 Berger, U. and F.-J. Lübken, Weather in mesospheric layers, Geophys. Res. Lett., 33,

505 L04806,doi:10.129/2005GL02481, 2006.

506 DeLand, M.T., Shettle, E.P., Thomas, G.E. and J.J. Olivero, A quarter-century of satellite PMC

507 observations, J. Atmos. Solar-Terr. Phys., 68, 9-29, 2006.

508 Fritts, D.C., J.R. Isler, G.E. Thomas, and O. Andreassen, Wave breaking signatures in

509 noctilucent clouds, Geophys. Res. Lett., 20, 2039-2042, 1993.

510 Gadsden, M., Noctilucent clouds, Space Sci. Rev., 33, 279-334, 1982.

511 Gadsden, M. Observations of noctilucent clouds from North-West Europe, Annales Geophysicae,

512 3, 119-126, 1985.

513 Gordley, L. L., B. T. Marshall, and D. A. Chu, LINEPAK: Algorithms for modeling spectral

514 transmittance and radiance, JQSRT, 52, p563, 1994.

515 Gordley, Larry L., Mark Hervig, Chad Fish, James Russell III, James Cook, Scott Hanson,

516 Andrew Shumway, Scott Bailey, Greg Paxton, Earl Thompson, Lance Deaver, John Burton,

517 Brian Magill, Chris Brown, and John Kemp, The Solar Occultation For Ice Experiment, J.

518 Atmos. Solar-Terr. Phys., in review, 2008.

519 Gumbel, Jorg and Georg Witt, In situ measurements of the vertical structure of a noctilucent

520 cloud, Geophys. Res. Lett., 25, No. 4, 493-496, February 15, 1998.

22 521 Hervig, M., R. E. Thompson, M. McHugh, L. L. Gordley, J. M. Russell, M. E. Summers, First

522 confirmation that water ice is the primary component of polar mesospheric clouds, Geophys.

523 Res. Lett., 38, 971-974, 2001.

524 Hervig, Mark E., Larry L. Gordley, Michael H. Stevens, James M. Russell III, Scott M. Bailey

525 and Gerd Baumgarten, Interpretation of SOFIE PMC measurements: Cloud identification

526 and derivation of mass density, particle shape, and particle size, J. Atmos. Solar-Terr. Phys.,

527 in review, 2008a.

528 Hervig, Mark E., Larry Gordley, James Russell III, and Scott Bailey, SOFIE PMC observations

529 during the northern summer of 2007, J. Atmos. Solar-Terr. Phys., in review, 2008b.

530 Horanyi, V. Hoxie, D. James, A. Poppe, C. Bryant, B. Grogan, B. Lamprecht, J. Mack, F.

531 Bagenal, S. Batiste, N. Bunch, T. Chantanowich, F. Christensen , M. Colgan, T. Dunn , G.

532 Drake, A. Fernandez, T. Finley, G. Holland, A. Jenkins, C. Krauss, E. Krauss, O. Krauss,

533 M. Lankton, C. Mitchell, M. Neeland , T. Resse, K. Rash, G. Tate, C. Vaudrin, J. Westfall,

534 The Student Dust Counter on the New Horizons Mission, Space Science Rev.,

535 10.1007/s11214-007-9250- y, 2007.

536 Leslie, R. J., Sky Glows, Nature, 33, 245, 1885

537 Love, S.G., D. E. Brownlee, A Direct Measurement of the Terrestrial Mass Accretion Rate of

538 Cosmic Dust, Science, vol. 262, p550-553, (1993)

539 Rapp, M and F. –J. Lubken, Polar mesosphere summer echoes (PMSE): Review of observations

540 and current understanding, Atmos. Chem. Phys., 4, 2601 – 2633, 2004.

541 Rapp, M. and G. E. Thomas, Modeling the microphysics of mesospheric ice particles:

542 Assessment of current capabilities and basic sensitivities, J. Atmos.Sol.Terr.Phys., 68, 715-

543 744, 2006.

23 544 Rapp, M., G. E. Thomas and G. Baumgarten, Spectral properties of mesospheric ice clouds:

545 Evidence for nonspherical particles, J. Geophys. Res., 112, D03211,

546 doi:1029/2006JD007322,2007.

547 Rusch, D. W., G. E. Thomas1, W. McClintock, A. W. Merkel, S. M. Bailey, J. M. Russell III, C.

548 E. Randall, C. Jeppesen, and M. Callan, The Cloud Imaging and Particle Size Experiment

549 on the Aeronomy of Ice in the Mesosphere Mission: Cloud Morphology for the Northern

550 2007 season, J. Atmos. Solar-Terr. Phys., in review, 2008.

551 Simpson, J. A. et al., Dust counter and mass analyzer (DUCMA) measurements of comet

552 Halley's coma from VEGA spacecraft, Nature, vol. 321, p. 278-280 (1986)

553 Simpson, J. A.; Tuzzolino, A. J., Polarized Polymer Films as Electronic Pulse Detectors of

554 Cosmic Dust Particles, Nuclear Instruments and Methods, vol. A236, p. 187-202 (1985)

555 Siskind, D. E., M. H. Stevens, and C. R. Englert, A model study of global variability in

556 mesospheric cloudiness, J. Atmos. Sol. Terr. Phys., 67(5), 429-519, 2005.

557 Srama, R. et al., The Cassini Cosmic Dust Analyzer, Space Science Reviews, Volume 114, Issue

558 1-4, pp. 465-518 (2004)

559 Summers, M. E., et al., Discovery of a water vapor layer in the Arctic summer mesosphere:

560 Implications for polar mesospheric clouds, Geophys. Res. Lett., 28, 3601, 2001.

561 Taylor, M.J., Mesospheric cloud observerations at unusually low latitudes, J. Atmos Sol.-Terr.

562 Phys., 64, 991-999,2002.

563 Thomas, G. E. and C. P. McKay, On the mean particle size and water content of polar

564 mesospheric clouds, Planet. Space Sci., 33, 1209-1224, 1986.

565 Thomas, G. E., Olivero, J. J., Jensen, E. J., Schröder, W., Toon, O. B., 1989. Relation between

566 increasing methane and the presence of ice at the mesopause. Nature 338, 490-492.

24 567 Thomas, G. E., R. D McPeters, and E. J. Jensen, Satellite observations of polar mesospheric

568 clouds by the SBUV Radiometer: Evidence of a solar-cycle dependence, J. Geophys. Res..,

569 96, 927-939, 1989.

570 Tuzzolino, Anthony J. et al., Dust Measurements in the Coma of Comet 81P/Wild 2 by the Dust

571 Flux Monitor Instrument, Science, Volume 304, Issue 5678, pp. 1776-1780 (2004)

572 Tuzzolino, A. J. et al., Final results from the space dust (SPADUS) instrument flown aboard the

573 earth-orbiting ARGOS spacecraft, Planetary and Space Science, Volume 53, Issue 9, p. 903-

574 923 (2005)

575 vonZahn, U. and U. Berger, Persistent ice cloud in the midsummer upper mesosphere at high

576 latitudes: Three-dimensional modeling and cloud interactions with ambient water vapor, J.

577 Geophys. Res., 108(D*), 8451, ddo1:1029/2002JD002409(2003)

578

579

580 Tables

581 Table 1. SOFIE Channel Characteristics

582 Table 2. SOFIE Retrieval Characteristics.

583 Table 3. CIPS Requirements and Performance

25 584 Figures

585 Figure 1. photographed by Tom Eklund, July 28, 2001, Valkeakoski,

586 Finland.

587 Figure 2. SOFIE instrument

588 Figure 3. Calculated transmission spectra for key absorbers near the PMC height within the

589 SOFIE water vapor channel wavelength region.

590 Figure 4. CIPS Instrument

591

592 Figure 5. CIPS Field-of-View projected to 83 km. The satellite velocity vector direction is to

593 the right.

594

595 Figure 6. CIPS daily mean image showing coverage for a full day of CIPS images on July 17,

596 2007. FOV overlap from one orbit to the next occurs at ~70N.

597 Figure 7. The 12 active PVDF sensors of CDE. Each patch has a surface area of about 85

598 cm2.The panel is mounted to point towards the local zenith direction at all times,

599 minimizing the impact rates from orbital debris.

600 Figure 8. CIPS daily mean image for June 30, 2007.

601

602 Figure 9. CIPS daily mean image for July 1, 2007.

603

604 Figure 10. Time – height cross section of SOFIE H2O and O3 observations for the 2007

605 northern polar summer. Mesopause altitude (solid line) and the altitude of peak ice

606 extinction (dashed line) are also indicated.

607 Figure 11. SOFIE mesospheric water profiles on August 12, 2007.

26 608

609 Figure 12. SOFIE altitude versus time cross section of ice mass density showing that the

610 mesospheric ice extends over a broad altitude range. The solid line is the

611 mesopause and the dashed line is the location of peak PMC extinction.

Tables

Table 1. SOFIE Channel Characteristics.

Chann Band Target* Center λ G∆V el (µm) 1 O s 0.292 1 3 30 2 O3 w 0.330 3 PMC s 0.867 2 300 4 PMC w 1.037 5 H O w 2.46 3 2 96 6 H2O s 2.618 7 CO s 2.785 4 2 110 8 CO2 w 2.939 9 PMC s 3.064 5 120 10 PMC w 3.186

11 CH4 s 3.384 6 202 12 CH4 w 3.479

13 CO2 s 4.324 7 110 14 CO2 w 4.646 15 NO w 5.006 8 300 16 NO s 5.316 *s indicates strongly absorbing band, w denotes weakly absorbing band

27

Table 2. SOFIE Retrieval Characteristics. Altitude Retrieval Measurement Range Bands 7, Channels 4 T 15 - 110 and 13

Sun sensor, 701 nm T 50 - 100 refraction angle O3 Band 1 (Channel 1) 15 - 105 H2O Band 6 (Channel 3) 15 – 105 CO2 Channels 4 and 7 15 - 95 CH4 Band 11 (Channel 6) 15 – 90 NO Channel 8 80 - 110

PMC Bands 2, 3, 4, 5, 8, 9, cloud extinction 10, 12, 14, and 15

Table 3. CIPS Requirements and Performance Parameter Reqmt. Design Measured Spatial Resolution (nadir) 3.0 km 2.1 km 2.1 km Absolute Accuracy 15% 10% 10% Precision (SNR for Earth 2.0 4.7 5.2 albedo=2x10-5 sr -1)1 1SNR=Signal-to-Noise-Ratio

28 Figures

Figure 1. Noctilucent cloud photographed by Tom Eklund, July 28, 2001, Valkeakoski, Finland.

Figure 3. Calculated transmission spectra for key absorbers near the PMC height within the SOFIE

Figure 2. SOFIE instrument water vapor channel wavelength region.

29

2000 1500

1000

500

0

km Crosstrack, -500

-1000

-1500

-2000 -2500 -2000 -1500 -1000 -500 0 500 1000 1500 2000 2500 Alongtrack, km

Figure 5. CIPS Field-of-View projected to 83 km. The satellite velocity vector direction is to the right. Figure 4 CIPS Instrument

Figure 7. The 12 active PVDF sensors of CDE. Each patch has a surface area of about 85 cm2.The panel is mounted

to point towards the local zenith direction at all times, minimizing the Figure 6. CIPS daily mean image showing impact rates from orbital debris. coverage for a full day of CIPS images on July 17, 2007. FOV overlap from one orbit to the next occurs at ~70N.

30

Figure 8. CIPS daily mean image for June 30, 2007

Figure 10. Time – height cross section of

SOFIE H2O and O3 observations for the 2007 northern polar summer. Mesopause altitude (solid line) and the altitude of peak ice extinction (dashed line) are also indicated.

Figure 9. CIPS daily mean image for July 1, 2007

Figure 11. SOFIE mesospheric water profiles on August 12, 2007

31

Figure 12. SOFIE altitude versus time cross section of ice mass density showing that the mesospheric ice extends over a broad altitude range. The solid line is the mesopause and the dashed line is the location of peak PMC extinction.

32