Measurements of meteor smoke particles during the ECOMA-2006 campaign: 2. Results

ARTICLE IN PRESS Journal of Atmospheric and Solar-Terrestrial Physics 71 (2009) 486–496

Contents lists available at ScienceDirect

Journal of Atmospheric and Solar-Terrestrial Physics journal homepage: www.elsevier.com/locate/jastp

Measurements of meteor smoke particles during the ECOMA-2006 campaign: 2. Results Irina Strelnikova a, Markus Rapp a,, Boris Strelnikov a, Gerd Baumgarten a, Alvin Brattli b, Knut Svenes b, Ulf-Peter Hoppe b, Martin Friedrich c, Jo¨rg Gumbel d, Bifford P. Williams e a

¨ hlungsborn, Germany Leibniz-Institute of Atmospheric Physics, Ku Norwegian Defence Research Establishment (FFI), P.O. Box 25, 2027 Kjeller, Norway c Institute of Communication Networks and Satellite Communications, Graz University of Technology, Austria d Department of Meteorology, Stockholm University, 10691 Stockholm, Sweden e Northwest Research Associates, Colorado Research Associates Division, Boulder, CO, USA b

a r t i c l e in f o

a b s t r a c t

Article history: Accepted 22 July 2008 Available online 30 July 2008

The first sounding rocket of the European ECOMA-project (ECOMA, Existence and Charge state Of Meteoric smoke particles in the middle Atmosphere) was launched on 8 September 2006. Measurements with a new particle detector described in the companion paper by Rapp and Strelnikova [2008. Measurements of meteor smoke particles during the ECOMA-2006 campaign: 1. Particle detection by active photoionization. Journal of Atmospheric and Solar-Terrestrial Physics, this issue, doi:10.1016/j.jastp.2008.06.002] clearly showed meteor smoke particle (MSP) signatures in both data channels. The data channels measure particles directly impacting on the detector electrode and photoelectrons from the particles actively created using ionization by the UV-photons of a xenon-flashlamp. Measured photoelectron currents resemble model expectations of the shape of the MSP layer almost perfectly, whereas derived number densities in the altitude range 60–90 km are larger than model results by about a factor of 5. Given the large uncertainties inherent to both model and the analysis of our measurements (e.g., the composition of the particles is not known and must be assumed) we consider this a satisfactory agreement and proof that MSPs do extend throughout the entire mesosphere as predicted by models. The measurements of direct particle impacts revealed a confined layer of negative charge between 80 and 90 km. This limited altitude range, however, is quantitatively shown to be the consequence of the aerodynamics of the rocket flight and does not have any geophysical origin. Measured charge signatures are consistent with expectations of particle charging given our own measurements of the background ionization. Unfortunately, however, a contamination of these measurements from triboelectric charging cannot be excluded at this stage. & 2008 Elsevier Ltd. All rights reserved.

Keywords: Meteor smoke particles In situ measurements

1. Introduction Meteor smoke particles (MSPs) have long been suggested as recondensation products of meteoroid ablation in the altitude range between 70 and 110 km (Rosinski and Snow, 1961; Hunten et al., 1980). Today, it is thought that small subnanometer sized MSPs form as a result of complex metal- and silicon-oxide chemistry which may then grow further by Brownian coagulation (Plane, 2003). These particles have been suggested to be important for a variety of atmospheric processes such as the nucleation of noctilucent clouds and polar stratospheric clouds, and it has also been speculated that they are the form in which meteoric material is transported to the ground (e.g., Rapp and

 Corresponding author. Tel.: +49 38293 6820; fax: +49 38293 6850.

E-mail address: [email protected] (M. Rapp). 1364-6826/$ - see front matter & 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.jastp.2008.07.011

Thomas, 2006; Voigt et al., 2005; Gabrielli et al., 2004). While MSPs are obviously scientifically interesting, the experimental quantification of their properties has proven to be extremely challenging mainly due to their tiny dimensions. In consequence, only very few in situ measurements of charged MSPs have succeeded so far (see Rapp et al., 2007 for a recent review) and only recently first radar measurements were reported by Strelnikova et al. (2007). However, neither were these previous measurements able to provide estimates of the total number density of MSPs (i.e., charged and neutral particles) nor were these measurements successful to deliver data outside a limited altitude range between 78 and 90 km. In our companion paper (Rapp and Strelnikova, 2008) (further denoted RS08), we introduced a new particle detector, i.e., a combination of a classical Faraday cup similar to the type first utilized by Havnes et al. (1996) and a xenon-flashlamp for the active photoionization of MSPs and the subsequent detection of

ARTICLE IN PRESS I. Strelnikova et al. / Journal of Atmospheric and Solar-Terrestrial Physics 71 (2009) 486–496

corresponding photoelectrons. We presented raw data acquired with this detector during the first rocket flight of the ECOMAproject in September 2006 (ECOMA, Existence and Charge state Of Meteoric smoke particles in the middle Atmosphere). By comparison with laboratory measurements we demonstrated that we indeed measured photoelectrons created by the UV-photons of the xenon-flashlamp. Finally, we also demonstrated that measured currents could only be quantitatively explained assuming that the photoelectrons originated from MSPs. In contrast to this, consideration of photoionization of all other species with low enough ionization potential to become ionized by the photons of the xenon-flashlamp like NO, O2 ð1 Dg Þ, metal atoms and negative ions resulted in far too low currents below an altitude of 90 km. Above this altitude, however, we showed that it is to be expected that measured currents are dominated by photoionization of NO. In the current manuscript we present an overview of the measurements obtained during the first ECOMA rocket flight and corresponding ground-based measurements in September 2006. In Section 2 we present a summary of the different ground based and in situ techniques used to characterize MSPs and their atmospheric and ionospheric environment. Corresponding results are presented in Section 3. From our particle detector measurements we then derive altitude profiles of MSP-properties both from the photoelectron-data and from the conventional Faraday cup measurements (Section 4). These derived parameters will be critically discussed with respect to necessary assumptions regarding the composition of the particles, the influence of aerodynamics, and secondary effects like triboelectric charging and particle fragmentation. Finally, we conclude with a summary and future directions to be pursued within the ECOMA-project.

2. Methods of observations 2.1. Ground-based measurements The ECOMA 2006 sounding rocket campaign was supported by ground-based measurements with the ALOMAR RMR and Weber Na lidars and with the EISCAT VHF and UHF radars. The ALOMAR RMR-lidar of the Leibniz Institute of Atmospheric Physics (IAP) utilizes Rayleigh scattering from the molecular atmosphere to obtain a relative neutral air density profile. The latter is converted to a temperature profile assuming hydrostatic equilibrium. The ALOMAR RMR-lidar has been described in detail by von Zahn et al. (2000). The ALOMAR Weber Na-lidar, which is run by Colorado Research Associates and the Norwegian Defence Research Establishment uses resonant scattering from the D2-line of mesospheric sodium atoms to obtain sodium number densities (from the amplitude of the scattered signal) and temperature (from the shape of the D2-line). The ALOMAR Weber Na-lidar has been described in detail in She et al. (2002). Both lidars are located at the ALOMAR observatory which is part of the Andøya Rocket Range. The EISCAT UHF and VHF (930 and 224 MHz) incoherent scatter radars are located at Tromsø (69 N, 19 E) which is 130 km away from the launch site. During the launch window of the ECOMA 2006 campaign the radars were both run using an alternating code called ‘arcdlayer’ which is a further development of another low altitude modulation described in detail in Turunen et al. (2002). These measurements provided real time information of electron number densities in the altitude range from 60 to 140 km with a height resolution of 1 and 3 km below and above 101 km, respectively. Details about the EISCAT VHF and UHF radars can be found in Baron (1986) and Folkestad et al. (1983), respectively.

487

2.2. Rocketborne instruments A sketch of the ECOMA payload is shown in Fig. 1. The main instrument, the ECOMA particle detector, is mounted on the front deck of the payload, and is designed to measure MSP number densities. A detailed description of the ECOMA instrument is presented in the accompanying paper by RS08. In short, the ECOMA particle detector is a Faraday cup with two shielding grids biased at 6:2 V, similar to the one first used by Havnes et al. (1996). Heavy MSPs pass through the biased grids and—if they carry some charge—produce a current which can be measured by a sensitive electrometer. Measurements of this current are further referred to as DC-channel measurements. Note that neutral particles cannot be measured by this method. In addition, because of aerodynamical effects, this method only allows the measurement of particles with radii larger than 2 nm (Hora´nyi et al., 1999; Rapp et al., 2005; Hedin et al., 2007). To detect MSPs of a broad size range and independent of their charge state, we improved the Faraday cup design by adding a xenon-flashlamp for the active photoionization of MSPs and the subsequent detection of photoelectrons (see RS08 for more details). Corresponding measurements will further be designated flash-channel measurements. The other instrument directly dealing with the meteoric smoke onboard the ECOMA payload was the MAGIC particle sampler. This instrument is designed for the in-flight collection of MSPs and return to the ground for later analysis in the laboratory (Gumbel et al., 2005). The analysis of corresponding samples returned from the ECOMA-01 sounding rocket flight is not yet completed and will not be discussed in this paper. The radio wave propagation experiment yields a high precision electron density measurement. Basically, this is realized by transmitting a radio signal from the ground and receiving it by a pair of antennas on the rocket. The theoretical basis and practical application of this technique are described for example in Bennett et al. (1972) and Smith (1986). The transmitted linearly polarized electromagnetic wave may be considered to be the resultant of two circularly polarized waves with equal electric field vectors rotating in opposite directions. When propagating through the ionospheric plasma these two waves differ in absorption and phase velocity. The difference in absorption causes the resultant wave to become elliptically polarized (differential absorption). In addition, the difference in phase velocity causes the major axis of the polarization ellipse to rotate as the wave propagates (Faraday rotation). Since both differential absorption and Faraday rotation directly depend on the electron number density along the path of the radio signal through the ionosphere, the height resolved measurements of both quantities can be used to precisely derive the electron number density profile at a typical vertical resolution of 1 km (defined by the rocket velocity and the rocket spin rate). Importantly, these measurements are not influenced by undesirable effects due to the interaction of the payload with the ambient plasma such as payload charging. Two cold plasma probes (CPPs) are mounted on the rear deck of the payload and essentially utilize the retarding potential analyzer (RPA) technique. From analysis of the collected current it is possible to infer information about macroscopic plasma parameters like density and temperature as well as the payload potential. A schematic of the main features of this instrument is shown in Fig. 2, and it is described further in Svenes et al. (2005). In short, this approach calls for applying a sweeping potential to the outer grid of the sensor, and subsequently measuring the ensuing electron current. This is accomplished by varying the potential on the outer grid relative to the payload ground while biasing the internal collector to obtain electrons only.

ARTICLE IN PRESS 488

I. Strelnikova et al. / Journal of Atmospheric and Solar-Terrestrial Physics 71 (2009) 486–496

Fig. 1. Schematic of the ECOMA payload. The total length of the 14 in-diameter payload is 2.9 m.

Fig. 2. Details of the cold plasma probe (CPP) mounted on the rear deck of the ECOMA payload, see also Fig. 1.

As long as the grid potential is negative with respect to the plasma only electrons which are still able to penetrate the potential field are collected. By varying the applied potential systematically, an integral electron energy spectrum is obtained from these measurements. From the analysis of these spectra (by fitting the I2V curves to a Boltzmann distribution), the parameters characterizing the ambient electron population can be determined. As the sweeping potential enters into the attractive potential region (for electrons), the I2V curve will start to deviate from the Boltzmann distribution. Hence, a distinct change in the derivative of the I2V curve will occur here. The potential (relative to payload ground) where this happens is then the negative of the payload potential relative to the plasma. To minimize the plasma disturbance induced by the instrument itself, an auto-corrective potential sweep was used during the measurements. The algorithm to be used for this is based on the property of the second derivative of the I2V curve. The point where the second derivative drops to zero constitutes the maximum potential value. This value must be found on the fly, and a suitable retarding potential sweep is employed from this point. That is, the retarding voltage is increased by about 4 mV 1 until the measured current is 16 of the starting value (unless the noise level is reached first). Since the plasma population constituting the source of the current to be measured is fairly low in thermal energy, the spacecraft itself constitutes the main source of disturbance. The two dominant problems are that of any plasma sheets caused by charging of the rocket, as well as asymmetries in the particle populations caused by the plasma streaming around the moving vehicle. In order to minimize these problems, the spheres were

mounted on two 50 cm long booms close to the aft of the rocket. During flight the booms were released such that they were positioned perpendicular to the main axis in order to keep the probes out of the deep wake region. For stability reasons the booms were mounted symmetric to the spin axis. In addition to the CPP, there were two fixed biased probes on short booms mounted on the forward deck. They were intended to measure positive ions and electrons, respectively. Unfortunately, measurements with these two probes were impaired by payload charging effects and will not be discussed here any further. The payload was further equipped with the CONE (COmbined measurement of Neutrals and Electrons) instrument (see Giebeler et al., 1993 for more details). CONE is a combination of an ionization gauge for the measurement of neutral density and a fixed biased Langmuir probe to measure electron number density. CONE yields measurements with very high spatial resolution and allows the detection of small scale fluctuations in both species (neutrals and electrons) that arise due to processes like neutral air turbulence (Lu¨bken et al., 2002) or plasma instabilities (see e.g., Blix et al., 1994). In addition, the height profile of neutral number densities can be integrated assuming hydrostatic equilibrium to yield a temperature profile at 200 m altitude resolution and an accuracy of 3 K (Rapp et al., 2001, 2002). Finally, two Pirani gauges were mounted on the front deck. These gauges are designed for low-cost density and temperature measurements. Due to their off-axis position on the front deck, corresponding measurements require a complicated aerodynamic correction. This is subject to an ongoing investigation and will not be discussed here any further.

ARTICLE IN PRESS I. Strelnikova et al. / Journal of Atmospheric and Solar-Terrestrial Physics 71 (2009) 486–496

3. Atmospheric measurements From 6 to 18 September 2006, the European sounding rocket campaign ECOMA-2006 took place at the North-Norwegian Andøya Rocket Range (69 N, 16 E). The project is devoted to the study of MSPs and their atmospheric and ionospheric environment. The ECOMA-2006 campaign was the first sounding rocket campaign in the frame of the ECOMA project which is split into three subsequent campaigns and is planned to include a total of seven launches. During this campaign two sounding rockets were launched while ground-based measurements were performed with the ALOMAR RMR-lidar to measure temperatures and with the sodium lidar to measure sodium densities and temperatures. The EISCAT VHF and UHF radars located at Tromsø (69 N, 19 E) provided electron number density measurements. The radar observations were conducted during each launch-window (i.e., from 21 to 01 UT) throughout the entire campaign period while the Weber Na and ALOMAR RMR lidars were operating whenever weather conditions permitted. In this manuscript we discuss results from the first flight, labeled ECOMA-01, which was launched on 8 September at 22:17:00 UT and reached an apogee of 130.6 km. 3.1. Neutral and ionized background atmosphere In the night of the rocket launch, the EISCAT radars detected energetic particle precipitation from 21:58 UT which created enhanced ionization in the D-region. Just a few minutes later, i.e., at 22:04 UT a sporadic E layer (further designated Es layer) appeared with maximum electron number densities of 105 cm3 at 94.5 km. This layer was stable until 23:02 UT after which it slowly settled down (i.e., at 23:11 UT the Es layer maximum was at 93 km) and finally disappeared at 00: 00 UT. Electron number densities from both radars were integrated over 2 min around the rocket launch time and are shown in Fig. 3. In the same figure we also present the electron number density profile derived from the radio wave propagation measurements onboard the ECOMA-01 payload. All three measurements agree remarkably well indicating that the Es layer extended over more than 100 km in the horizontal direction (i.e., the horizontal distance from the Andøya Rocket Range to the EISCAT radars in Tromsø is about 130 km). Comparing the electron number density profiles shown in Fig. 3 to the large collection of D-region electron number density mea-

489

surements in the auroral zone with rockets and the EISCAT radars (Friedrich and Kirkwood, 2000) we see that the ionospheric conditions were moderately disturbed to quiet (this is also confirmed by measurements with the Andøya Imaging Riometer, http://alomar.rocketrange.no/iris-and.html which show absorption-values less than 0.1 dB). We will come back to this issue in Section 4.2.2 where we will discuss the observed charge sign of MSPs. Lidar measurements during the launch night were largely hampered by tropospheric weather conditions; however, it was possible to repeatedly run the ALOMAR RMR-lidar and the ALOMAR Weber Na-lidar for short time intervals in the period between 21:12 and 23:51 UT. In Fig. 4 we present temperature profiles inferred from measurements with the ALOMAR RMR-lidar, the ALOMAR Weber Na-lidar, and the CONEinstrument onboard the ECOMA-01 payload. Most remarkably, these measurements reveal minimum temperatures of about 160 and 170 K at 80 and 105 km, respectively. Between these altitudes, the measurements show a pronounced inversion layer with maximum temperatures of about 200 K at 90 km altitude. Differences between the CONE and lidar temperatures can likely be explained by the horizontal distance of 70 km between the two measurements and the different times when the profiles were obtained (i.e., the RMR-lidar measurements were taken 40 min before and the Na-lidar data were taken 55 min after the rocket launch). In any case, these temperatures are far above the frost point of water vapor (indicated by the dotted line in Fig. 4) and indicate that any particles recorded by the ECOMA-particle detector can certainly not consist of water ice. Weber Na-lidar measurements from approximately 1 h before and after the rocket launch are presented in Fig. 5. These profiles show that the Na layer extended from 80 to 105 km altitude with a sudden sodium layer superimposed at an altitude of 93 km (in the period 23:11–23:16 UT), coincident with the altitude of the Es layer recorded by the EISCAT radars. 3.2. In situ particle measurements An overview of the currents measured with the ECOMA particle detector for the ascent and descent of the rocket trajectory was already presented in RS08 and is reproduced here in Fig. 6. The DC-channel measurements (left panel in Fig. 6) reveal a prominent negative current layer in the altitude range between 80 and 90 km indicative of negatively charged particles.

Fig. 3. Left panel: electron number densities measured with the EISCAT UHF-radar on 8 September 2006. The red vertical line indicates the time of the ECOMA-01 start. Right panel: electron number density profile from the EISCAT UHF and VHF radars (black and blue lines) from the time of the ECOMA-01 launch compared to the in situ results from the Faraday rotation experiment (red line). Note that the difference in altitude of the sporadic layer seen in the radar and rocket measurements is likely due to the horizontal distance between the different observations ð130 kmÞ.

ARTICLE IN PRESS 490

I. Strelnikova et al. / Journal of Atmospheric and Solar-Terrestrial Physics 71 (2009) 486–496

Fig. 4. Temperature profiles derived from measurements with the ALOMAR RMRlidar (green line; corresponding error bars are indicated by the light green shading) integrated over the period 21:12–21:27 UT and the CONE instrument (black solid line). Note that the coincidence of lidar and CONE-temperatures around 75 km is by construction since the start temperature for the hydrostatic integration of the lidar density profile was taken from the CONE measurements. The orange line with yellow error bars indicates the temperature derived from measurements with the ALOMAR Weber Na-lidar where data were integrated over the period 23:11–23:16 UT. The black dashed line shows a profile from the MSIS90Eclimatology for the latitude of Andøya and the date and time of the rocket flight for comparison. The dotted blue line shows an estimate of the frost point temperature using the water vapor saturation formula of Marti and Mauersberger (1993) and assuming a constant water vapor mixing ratio of 5 ppm.

returns to zero. On downleg, the measured current shows a similar behavior with slightly enhanced values from apogee down to 80 km. The largest difference, however, is that on downleg the (negative) current further increases with decreasing altitude until it reaches a maximum value of 15 nA at 75 km and disappears at about 60 km, i.e., about 20 km lower than on upleg. As explained in detail in RS08, this difference between upleg and downleg measurements can be easily understood in terms of the mean free path of the photoelectrons in the environment of the sounding rocket. The mean free path is greatly reduced during the upleg part of the rocket flight when the instrument is facing the ram and the air in front of the detector is compressed such that the density increases (and the mean free path of the electrons decreases) by more than a factor of 10. In the next section, we discuss in detail the origin of these currents. We convert currents to MSP number densities and we critically discuss uncertainties owing to necessary assumptions, and effects like payload charging, triboelectric effects, and secondary charge production by particle fragmentation.

4. Discussion 4.1. Interpretation of ECOMA-01 flash measurements. In this section we will try to convert measured photoelectron currents to MSP number densities. Because of the reasons described above and discussed in detail in RS08, the downleg measurements cover a significantly larger altitude range than the upleg measurements, i.e., from 60 km to apogee. Therefore we only consider the downleg measurements for conversion to MSP number densities. In RS08 we showed that in the range from 60 to 90 km the signal is due to the photoelectrons, emitted from dust particles. Above 90 km, however, the signal is dominated by photoelectrons from nitric oxide. We start with the interpretation of the data from the lower part, where the measured current is due to photoionization/photodetachment of MSPs. It was shown in RS08 that peak photoelectron currents can be written as IMSP max ¼

Z

1 r min

Z

ve Dt 2:5 cm

Z

hc=W p 110 nm

!

P  dr p  dl  dl Fig. 5. Sodium number density profiles from measurements with the ALOMAR Weber Na-lidar from 21:19 to 21:28 UT (black line) and from 23:11 to 23:16 UT (red line). Note that measurements during the first period were done at one single frequency only, so that temperature estimates from this period are not available.

This layer is qualitatively very similar to the previous measurements by Faraday cup instruments reported, e.g., by Lynch et al. (2005) and Rapp et al. (2005). Above this negative layer, the current turns positive and even shows a very prominent positive peak at 93 km which is the altitude where the EISCAT radars observed the sporadic E-layer. Since the same positive currents above 90 km are also seen on the downleg of the rocket flight when the instrument is not facing the ram (and no particles can enter the detector volume), we conclude that these currents must be due to contamination by positive ions. These measurements will be further discussed—for example with respect to the observed particle charge, payload charging and secondary effects like triboelectric charging and particle fragmentation—in Section 4. The right panel of Fig. 6 shows our measurements owing to the flash of the detector (see RS08 for details). On upleg, large negative currents of up to 10 nA were recorded above an altitude of 80 km. This current decays to about 2 nA at apogee but never



dNp dF MSP  s ðr p ; lÞ dr p dl photo

e

Dt

(1)

where e is the electron charge, r min ¼ 0:2 nm is the minimum assumed size of MSPs, ve is the velocity of a photoelectron, Dt ¼ 10 ms is the sampling interval in the fast data channel (and hence also the charge integration time of our integrator circuit), h is Planck’s constant, c is the speed of light, and W p is the threshold energy for photoionization/photodetachment of a particle, i.e., the workfunction or electron affinity of the corresponding material. dF=dl is the number of photons per wavelength interval emitted in one flash, and l is the distance from the particle detector. P ¼ 2 S=ð4pl Þ is the probability that the photoelectron is emitted towards the detector electrode with area S. dN p =dr p is the number density of MSPs per size interval dr p , and dl and dl are the length and wavelength elements over which the integrations above are carried out. Note that the integration over the wavelength l starts at 110 nm because of the transmission properties of the MgF2-window of the xenon-flashlamp. Finally, 2 sMSP photo ðr p ; lÞ ¼ pr p  Q abs ðr p ; l; nðlÞ; kðlÞÞ  Y

(2)

is the photoionization/photodetachment cross section of MSPs with radius r p at photon wavelength l. The cross sections are estimated using Mie-theory. In Eq. (2) Q abs is the Mie absorption

ARTICLE IN PRESS I. Strelnikova et al. / Journal of Atmospheric and Solar-Terrestrial Physics 71 (2009) 486–496

491

Fig. 6. Overview of current measurements from the ECOMA flight on September 8th, 2006. Left panel: current measurements due to direct particle impacts on the electrode recorded as a DC-current in the slow data channel. Right panel: corresponding peak photoelectron currents recorded in the flash data channel. Black lines are for upleg measurements, grey lines are for downleg. This figure is reproduced from Rapp and Strelnikova (2008).

efficiency which we calculated using the publicly available Mie-code from the text book by Bohren and Huffman (1983), n and k are the real and imaginary parts of the refractive index of the MSP material, and Y is the quantum yield for photoemision/ photodetachment. In order to convert measured currents to MSP number densities we proceed as follows. Assuming that all particles are identical, i.e., they have the same size rp ¼ r p0 and consist of the same material, the MSP number density is obtained from Eq. (1) as Np ¼

IMSP max

Z

hc=W p

d 110 nm

!1 dF MSP s ðr p0 ; lÞ  dl dl photo

(3)

where we introduced the new variable d¼

e Dt

Z

v e  Dt

P  dl

(4)

2:5 cm

which does not depend on the properties of the particles. Fig. 8 shows the results of the MSP number density estimation in comparison with number densities from the model by Megner et al. (2006) which is shown as the red line. The blue lines represent the MSP number density obtained from the flash-channel measurements using Eq. (3) under the assumption of a fixed particle radius of r p0 ¼ 0:4 nm, a quantum yield Y ¼ 1:0 and refractive indices for Fe2O3 for either photoionization (solid line) or photodetachment (dashed line, see RS08 for details). The disagreement between derived number densities and the model results is obvious. However, MSP models all suggest an altitudedependent MSP size distribution rather than the occurrence of altitude-independent monodisperse particles (Hunten et al., 1980; Megner et al., 2006). Hence, the anticipated altitude variation of an MSP particle size distribution should be taken into account in our retrieval of MSP number densities. Using the model results of e as Megner et al. (2006) we define an effective cross section s R 1 R hc=W p dN p dF MSP rmin 110 nm dr  dl  sphoto ðr p ; lÞ  dr p  dl p e¼ s Nptotal

(5)

where Nptotal ¼

Fig. 7. Upper panel: MSP size distributions from the model of Megner et al. (2006) e for different altitudes. Lower panel: altitude profile of the effective cross section s as defined in Eq. (5).

is a total number density of the dust particles. As can be seen from Fig. 7 this effective cross section increases significantly with decreasing altitude. e into Eq. (1), the MSP number density can be Substituting s estimated as Np ¼

Z

1 r min

dNp  dr p dr p

(6)

IMSP max e ds

(7)

Applying Eq. (7) to our measurements we obtain the number density profile shown in Fig. 8 as the black lines (shown again for

ARTICLE IN PRESS 492

I. Strelnikova et al. / Journal of Atmospheric and Solar-Terrestrial Physics 71 (2009) 486–496

Fig. 8. Altitude profiles of total MSP number densities derived from the flashchannel measurements assuming Fe2O3-particles with a fixed radius of 0.4 nm (blue lines) and assuming an altitude-dependent size distribution (black lines) taken from the model by Megner et al. (2006). Dashed and solid black and blue lines are for an electron affinity/workfunction of 2 or 5.5 eV, i.e., assuming photodetachment or photoionization, respectively. The red solid line shows the total MSP number density from the model of Megner et al. (2006).

the case of photoionization and photodetachment). The shape of the derived altitude profile of MSP number densities is in general agreement with model results. In terms of absolute values, however, we see that our analysis yields about a factor of 5 larger MSP number densities than expected from models below 80 km. This discrepancy increases further above 80 km, presumably because of the increasing abundance of NO. This difference must be compared to the combined uncertainty of both the model and the analysis of our measurement: In the case of the model there are unknowns like the total micrometeoroid mass influx and assumptions about the initial size of the recondensed MSPs, their microphysical properties (like particle shape, charge, magnetism, etc.), and the eddy diffusion profile used for calculating the vertical transport of particles. For an in depth discussion of these issues and their effect on the overall uncertainty of the model profile see Megner et al. (2006). As far as our analysis is concerned, we have made assumptions about the composition (and hence refractive indices) of the particles, the corresponding quantum yield for photoemission/photodetachment, and, of course, also the altitude variation of the MSP size distribution which is taken from a model! Hence, it is not surprising that our estimate and the model results do not fit one to one. Still, we find it promising that our results do resemble the general shape predicted by models and that our analysis reproduces predicted number densities within one order of magnitude. In particular, our results give very strong evidence for the existence of MSPs in the entire altitude range of 60–90 km, i.e., far below the altitude cut-off of 80 km usually seen by previous in situ measurements relying on classical Faraday cup measurements or comparable techniques alone. Finally, we turn to the photoelectron measurements above the altitude of 90 km: In RS08 it was shown that this current is most likely due to photoionization of NO. Using the formula ! Z ve Dt Z hc=W i dF e Imax ¼ (8)  si ðlÞ  P  ni  dl  dl  Dt 110 nm dl 0 the measured current above 90 km can be converted to NO number densities (values of the ionization potential ð9:25 eVÞ and of the photoionization cross section were taken from Watanabe et al., 1953). The resulting profile is shown in Fig. 9 where we also compare our estimate to results from the HALOE-climatology (Siskind et al., 1998). This shows that our derived NO-number densities are larger than the corresponding results from the

Fig. 9. Derived nitric oxide number densities. Black and red lines are the measurements from ascent and descent accordingly. The thick blue lines (solid and dashed) are the results from the HALOE-climatology and that multiplied by a factor of 6.

HALOE-climatology by a factor of 6. However, a large variability of NO with time, particularly at high latitudes, was reported by different authors and is well known (Ackerman, 1979; Iwagami et al., 1998; Vitt et al., 2000). 4.2. Interpretation of ECOMA-01 DC measurements As already described in Section 3.2, the ECOMA-01 DC-channel measurements revealed a pronounced negative layer between 80 and 90 km with peak negative currents of 100 pA at an altitude of 85 km. Our downleg measurements, however, showed a positive ion leakage current. Hence, the measured current on the ascent of the rocket flight is the sum of the (positive) ion leakage current and the (negative) current due to MSPs. Thus, by subtracting the current measured on the downleg from that measured on the upleg we obtain the current which is only produced by MSPs. From this current MSP number densities can be estimated assuming singly negatively charged particles using the simple relation: I ¼ eSvNp

(9)

where e is the elementary charge, S is the area of the Faraday cup electrode, v is the rocket velocity, and N p is the MSP number density. Applying this formula to DC-channel measurements, we 3 obtain a maximum MSP number density of 200 particles/cm at 85 km which is in fair agreement with results from previous Faraday cup measurements of MSPs at this altitude (see Rapp et al., 2007, for a recent review). At this point, we also find it interesting to point out that the measured MSP layer sits underneath the maximum of the neutral sodium layer (compare Figs. 11 and 5). This location of the MSP layer relative to the sodium layer was also reported by Lynch et al. (2005). Plane (2004) showed that sodium species are permanently removed below about 85 km, both through the dimerization of NaHCO3 and the uptake of sodium species on MSPs. Thus, the decrease of the sodium density inside the layer of smoke particles is consistent with other experimental data (Lynch et al., 2005) and our current understanding of the sodium layer (Plane, 2004). While this agreement with independent previous measurements and our current understanding of the sodium layer is promising, we need to acknowledge, however, that several questions about this kind of measurements remain unanswered. These are: (1) Why do Faraday cup-type measurements always show distinct layers of MSPs in the altitude range 80–90 km?

ARTICLE IN PRESS I. Strelnikova et al. / Journal of Atmospheric and Solar-Terrestrial Physics 71 (2009) 486–496

(2) Are the measured currents truly due to charge residing on the MSPs which is deposited to the Faraday cup electrode upon impact or is the measured current due to secondary effects like particle fragmentation or triboelectric charging? In the following we address these two questions as they apply to our own measurements in detail. 4.2.1. Distinct MSP layers between 80 and 90 km All available model calculations of the MSP distribution in the middle atmosphere as well as our flash-channel measurements show a wide altitude extent of MSPs throughout the entire middle atmosphere. In this subsection we try to demonstrate quantitatively why Faraday cup and comparable measurements were so far only able to detect distinct MSP layers in the altitude range between 80 and 90 km. In the lower panel of Fig. 10 we compare the ECOMA-01 DCmeasurements to model results of the MSP distribution from Megner et al. (2006) where we have plotted total MSP number densities as well as profiles of number densities for MSPs larger than a certain radius. Obviously, the measured layer cannot be explained by any of these curves. However, as was already mentioned above, to understand our observations one has to take into account the aerodynamics of particle sampling by rocket borne detectors (Hora´nyi et al., 1999; Rapp et al., 2005; Hedin et al., 2007). In Rapp et al. (2005) we calculated the density fields inside and in the vicinity of our detector and simulated the trajectories of particles of given mass density, radius, and charge in this density field. From these calculations we determined the minimum particle radius r 

Fig. 10. Upper panel: minimum detectable radius r  of negatively charged particles with mass density of 3 (dashed blue line) and 2 (green line) g/cm3 as a function of altitude. The crosses are retrieved r from measurements using the model results of the density distribution of meteoric smoke particles. The yellow vertical line indicates the presence of some r min value for the entire altitude range. Lower panel: altitude profiles of the charge number densities of mesospheric nanoparticles measured (red) and calculated (black) for different size of particles. The green and dashed blue lines represent the amount of the dust which can be detected by a Faraday cup.

493

(for an assumed mass density) needed to allow the particles to reach the collecting electrode. The resulting altitude-profile of r  from Rapp et al. (2005) is reproduced in the upper panel of Fig. 10 by the blue line and symbols. With the green line and symbols we have further indicated r  assuming a slightly smaller mass density of 2 g=cm3 , whereas 3 g=cm3 was used for our original calculations. In order to estimate what kind of MSP number density profile we may expect assuming the model results of Megner et al. (2006) and taking into account the effect of aerodynamics, we have filtered the model output of Megner et al. (2006) with the r  profile. This means that at any given altitude we integrated the particle size distribution given by Megner et al. (2006) starting from r  to infinity. Resulting altitude profiles for the two considered mass densities of 2 and 3 g=cm3 are shown in the lower panel of Fig. 10 by the green and blue lines, respectively. While this estimate is certainly subject to many uncertainties due to unknown details about the mass density, shape, and charge state of MSPs, this figure clearly shows that the effect of aerodynamics can in fact provide a highly plausible explanation for the observed dust layer. We can actually go even one step further and estimate the values of r  from our measurements and the model by Megner et al. (2006). The obtained results are shown by red crosses in upper panel of Fig. 10. As can be seen, the derived r  -profile follows the ones from our calculations closely. While the aerodynamical explanation for the observed MSP layer described above is indeed plausible, we still have to mention that there might also be another factor that may have contributed to the observed layering. Since the DC-channel measures charged MSPs, neutral dust is invisible for this type of measurements. According to the present understanding, MSPs are charged by collisions with electrons, positive ions, and negative ions under nighttime conditions (e.g., Rapp and Lu¨bken, 2001; Rapp et al., 2005). Hence, if the ionization was not sufficiently high, it would be plausible to assume that the particles did not acquire any detectable charge below the lower boundary of observations. In order to investigate this possibility we compare the measured electron number densities from our rocket flight and the one described in Rapp et al. (2005). Note that these are the only flights reported so far where MSPs and absolute electron number densities (i.e., from a wave propagation experiment) were measured simultaneously onboard the rockets. Going back to Fig. 3 we see that the electron number density at the bottom of the MSP layer ð81 kmÞ is about 200=cm3 . This needs to be compared to the results of Rapp et al. (2005) where the MSP layer disappeared at an altitude of 82 km and an electron number density of less than 10=cm3 . This discrepancy between these two rocket flights is clearly not consistent with the idea that the detectable altitude range of MSPs is limited by the ambient ionization. In summary of this subsection, we conclude that there is strong quantitative evidence that the limited altitude range of MSP layers typically observed by Faraday cup-type instruments is due to aerodynamic effects.

4.2.2. Observed particle charge: real or due to secondary effects? The apparent MSP charge measured in the ECOMA DC channel between 80 and 90 km was negative. This is consistent with predictions of present day charging models, according to which particles are expected to be negatively charged because of the much higher mobility of electrons as compared to (the much heavier) ions. Exceptions to this rule may be caused by significant photoemission or in the presence of negative ions being the dominant negative charge carrier instead of electrons (e.g., Rapp and Lu¨bken, 2001; Rapp et al., 2005). Since the ECOMA-01

ARTICLE IN PRESS 494

I. Strelnikova et al. / Journal of Atmospheric and Solar-Terrestrial Physics 71 (2009) 486–496

measurements were done under nighttime conditions, photoionization can be excluded as a charging mechanism, and electron number densities were also large enough to guarantee that the influence of negative ions was negligible (compare Fig. 3 to Figure 3 in Rapp et al., 2005). So in all, there is nothing controversial with the ECOMA-01 measurements. Nevertheless, we need to consider the possibility that the observed charge sign does not reflect the true MSP charge but that observed currents are rather caused by secondary effects such as charge production by particle fragmentation or triboelectric charging (Havnes and Naesheim, 2007; Amyx et al., 2008). (a) Charge production by particle fragmentation: Tomsic (2001) investigated the interaction of a neutral ice cluster beam with surfaces of different types. This investigation showed that a small cluster-size-dependent fraction of ice clusters carried away one negative charge when impacting with nearly grazing incidence. For impacting angles far away from grazing incidence, however, no effect at all could be observed. Reanalyzing measurements of charged ice particles performed in 1994 and initially described in Havnes et al. (1996) and Havnes and Naesheim (2007) proposed that this process occurred on the grids of the Faraday cup. During this flight the sounding rocket experienced severe coning such that the incidence of the particles on the shielding grids was indeed close to grazing. We now consider whether this process could have caused the negative charge layer observed between 80 and 90 km altitude during the ECOMA-01 flight. Let us assume that the dust particles are neutral and negative charge is produced due to fragmentation via impact on the grid’s edge. In the flight considered here the angle between ram velocity and payload axis was 5 on upleg, thus the particles’ impact was near normal to the grid area, i.e., far away from grazing incidence. This means that the effective area where this process could occur is a stripe at the edge of the grid wires where the particles can hit at grazing incidence. The width of this stripe is equal to the diameter of particles. One grid cell has an area Sc0 equal to 0:8  0:8 ¼ 0:64 mm2 and all cells are identical. If the diameter of the dust particles (or the width of the stripes) is equal to 4 nm, then the area of the stripes inside one cell, Scs , can be written as Scs ¼ Sc0  ð0:8 mm  4:0  106 mmÞ2 ¼ 6:4  106 mm2

(10)

This means that the effective area where efficient charge production could occur is reduced by factor of Scs =Sc0 105 as compared to the electrode area S. Thus, for conversion of measured currents to number densities, the electrode area (S) must be multiplied by this factor 105. Taking further into account that the probability of this charging process is at most 104 per particle impact for the particle sizes considered here (Tomsic, 2001), Eq. (9) becomes I ¼ 109 eSvN p

(11)

This implies that to explain the measured currents of 100 pA, the MSP number density needs to be at least 1011 cm3 which is a factor of 106 larger than any model estimate (e.g., Hunten et al., 1980; Megner et al., 2006). Hence, we may safely exclude this process as a potential cause of the charge observed on the electrode of the Faraday cup. (b) Charge production by triboelectric charging: The other process which may potentially produce a current from neutral particles is called triboelectric or contact charging and was first proposed by Amyx et al. (2008) as a potential cause for observed MSP charge signatures. It involves the transfer of electrons between materials of different work functions, i.e., electrons from the material with lower work function are transferred to the one with larger work function. The contact charging of various dust

particles in the 502200 mm size range was studied experimentally by Sternovsky et al. (2001, 2002). We consider this process here in detail, because it may potentially explain why during some of the previous MSPmeasurements positive currents and during others negative currents were observed. Rapp et al. (2005) reported the observation of positively charged particles between 82 and 90 km under nighttime conditions in October 2004 and attributed the positive charge to the predominant abundance of negative ions in the altitude range where MSPs were observed. As described above, during the ECOMA-01 flight negative charge signatures were observed. One major difference between the instruments employed by Rapp et al. (2005) and the one used in this paper is the surface material of the electrode. In the first case a copper surface was used and whereas the electrode surface was gold-plated in the ECOMA-01 flight. The work functions of copper and gold are 4.65 and 5.1 eV, respectively. The copper surface, however, was likely oxidized such that a work function of 5.5 eV is more appropriate (e.g., Sternovsky et al., 2002). As discussed in RS08 the dust composition is nowadays assumed to be oxides of metals like Fe2O3 and SiO with typical work functions on the order of 5.5 eV (Plane, 2003; Sternovsky et al., 2002). In the case of the gold electrode used during ECOMA-01 triboelectric charging would imply that electrons from the gold electrode (work function 5.1 eV) would be transferred to the MSPs (work function 5.5 eV) such that a positive current would be expected—i.e., opposite to what was observed. However, if the MSPs were not neutral but negatively charged from the beginning then the work function of gold would need to be compared to the electron affinity of the MSPs which is expected to be about 2 eV (Wang et al., 1996). In this case, electrons from the MSPs would indeed be transferred to the electrode such that a negative current should be observed. As far as the case with the oxidized copper electrode during the 2004 flight is concerned, we note that Sternovsky et al. (2001, 2002) showed that in contact of two oxidized surfaces the sign of the charge carried by the dust is unpredictable. In such a case the chemical composition, shape, or surface properties (e.g., impurities) play an important role. Hence, a prediction of the expected charge sign is impossible given the poor knowledge on the properties of MSPs that we currently have. More light on the issue of triboelectric charging as a potential source for observed MSP-signatures may be shed by the arguments presented by Barjatya and Swenson (2006). These authors argued that triboelectric charging should certainly also happen between MSPs and the payload itself and suggested that the observed payload potential of more than 3 V in their case was due to this process. For the ECOMA-01 flight we are actually in the favorable position that the payload potential was directly measured by means of the CPP. Hence we may test Barjatya and Swenson (2006)’s hypothesis with our own data. Fig. 11 shows the ECOMA DC-channel measurements and the payload potential obtained from the CPP measurements. At this point we only consider CPP downleg measurements since the CPP was in the payload wake during upleg which could have potentially hampered our measurements and because the CPP data quality was poor (i.e., very little current) on upleg in the 80–90 km altitude range. Under the reasonable assumption that the MSP layer did not change considerably between the upleg and downleg part of the rocket trajectory (see also our photoelectron measurements presented above), we consider a comparison of upleg MSP measurements and downleg payload potential measurements appropriate. Fig. 11 clearly shows that there is no detectable reaction of the payload potential at the altitudes of the MSP layer, i.e., between 80 and 90 km. Rather, the payload potential is near constant with typical values of about 1:4 V.

ARTICLE IN PRESS I. Strelnikova et al. / Journal of Atmospheric and Solar-Terrestrial Physics 71 (2009) 486–496

Fig. 11. Left panel: ECOMA-01 DC current from upleg (black line) and downleg (red line). Right panel: altitude profile of the payload potential of the ECOMA-01 payload as derived from CPP measurements on downleg.

Hence, we conclude that the payload charging observed during the ECOMA-01 flight was not affected by triboelectric charging between the payload and MSPs. Rather it was caused by other influences like the large unshielded positive potential on the ECOMA-front grid ðþ6:2 VÞ which drew a considerable current from the plasma to the payload and caused the large negative payload potential. In summary, we believe that during this flight we measured negatively charged dust particles. However, given the many unknowns, we cannot exclude that the measured current was caused by triboelectric charging effects. In order to prove this a dedicated experiment should be made using different electrode materials with known work functions. A corresponding instrument is currently being designed and will hopefully soon yield more conclusive results about the issue of MSP charge detection and triboelectric charging.

5. Summary In the current study we have described results from the first sounding rocket flight within the European ECOMA project which took place on 8 September 2006, from the North-Norwegian Andøya Rocket Range. The payload carried instruments to characterize properties of MSPs and their neutral and charged environment. The payload was launched under conditions of moderate ionization in the D-region as monitored by measurements with the EISCAT UHF and VHF radars and confirmed by in situ Faraday rotation measurements. Temperatures as derived from ground-based lidar and rocket borne neutral air density measurements revealed temperatures in large excess of the frost point of water indicating that ice particles could not form in the mesosphere at the time of our in situ measurements. The ECOMA particle detector registered particle signatures in both data channels, i.e., particles which directly impacted on the detector electrode and photoelectrons from the active photoionization/ photodetachment of the particles. The photoelectron measurements clearly show that particles existed in the entire altitude range between 60 and 90 km. We converted corresponding currents to MSP number densities which resulted in an altitude profile in qualitative agreement with model predictions in the sense that the relative shape of the altitude profiles from model and measurements is near identical. In terms of absolute values, however, model and measurements deviate by about a factor of 5 below 80 km and increases above, presumably due to the increasing abundance of NO. However, given the many unknowns concerning both model and measurements, we consider this a satisfactory agreement at this stage of our knowledge

495

of the properties of MSPs. Better data with smaller error bars will require independent information on the composition of the particles as well as their electrical and optical properties. These are currently not available. As an interesting side-result, our photoelectron measurements above 90 km altitude allowed us to estimate NO number densities up to altitudes of 130 km. The resulting values overshoot a climatology from the HALOE instrument by another factor of 5, which seems to be in order given the large known natural variability of NO. Our direct measurements of particle impacts on the detector electrode revealed a negatively charged layer in the altitude range between 80 and 90 km in fair agreement with independent previous observations with similar techniques. Importantly, however, the comparison of our two data channels as well as an aerodynamical analysis clearly shows that the direct Faraday cup measurements fail to detect MSPs outside an altitude and particle size range dictated by aerodynamics. I.e., the analysis of our photoelectron measurements yields many more MSPs as recorded by the conventional Faraday cup measurement and in a much larger altitude range. Given the observed absolute ionization level, we argued that the observed particle charge is plausible, however, we also discussed potential secondary effects which could have mimicked the observed charge characteristics. These effects are particle fragmentation of particles colliding with detector grids under grazing incidence and triboelectric charging. While we showed that the first process (charge generation by fragmentation) can safely be excluded in our case, the arguments concerning triboelectric charging are unfortunately not similarly conclusive: it is evident that triboelectric charging was not responsible for the payload charging itself, however, this does not prove that the currents observed in the 80–90 km layer were not contaminated by this effect. In order to clarify this issue a dedicated experiment will be made using different electrode materials of known work function. A corresponding instrument is currently being designed and will hopefully soon yield more conclusive results about this issue. Acknowledgments We are indebted to Zoltan Sternovsky for useful discussions on the issue of triboelectricity. The ECOMA project is sponsored by the German Space Center under DLR-grant 50OE0301. The Norwegian part of the project was funded by the Norwegian Space Centre as project ECOMA 2006 and the Research Council of Norway, as project 170848. The Austrian contribution was funded under Grant 18560 of the Austrian Science Foundation (FWF). EISCAT is an international association supported by research organizations in China (CRIRP), Finland (SA), France (CNRS, till end 2006), Germany (DFG), Japan (NIPR and STEL), Norway (NFR), Sweden (VR), and the United Kingdom (PPARC). The ALOMAR Weber Na-lidar was supported by US AFOSR contracts F49620-03-C-0045 and FA9550-06-C-0129, US NSF Grants ATM-0545262 and ATM-0436703, the Andøya Rocket Range, and Research Council of Norway projects 165573 and 170855. References Ackerman, M., 1979. In situ measurements of middle atmosphere composition. Journal of Atmospheric and Terrestrial Physics 41, 723–733. Amyx, K., Sternovsky, Z., Knappmiller, S., Robertson, S., Hora´nyi, M., Gumbel, J., 2008. In-situ measurement of smoke particles in the wintertime polar mesosphere between 80 and 85 km altitude. Journal of Atmospheric and Solar-Terrestrial Physics 70, 61–70.

ARTICLE IN PRESS 496

I. Strelnikova et al. / Journal of Atmospheric and Solar-Terrestrial Physics 71 (2009) 486–496

Barjatya, A., Swenson, C.M., 2006. Observations of triboelectric charging effects on Langmuir-type probes in dusty plasma. Journal of Geophysical Research 111, A10302, doi:10.1029/2006JA011806. Baron, M., 1986. EISCAT progress 1983–1985. Journal of Atmospheric and Terrestrial Physics 48, 767–772. Bennett, F.D.G., Hall, J.E., Dickinson, P.H.G., 1972. D-region electron densities and collision frequencies from Faraday rotation and differential absorbtion measurements. Journal of Atmospheric and Terrestrial Physics 34, 1321–1335. Blix, T.A., Thrane, E.V., Kirkwood, S., Schlegel, K., 1994. Plasma instabilities in the lower E-region observed during the DYANA campaign. Journal of Atmospheric and Terrestrial Physics 56, 1853–1870. Bohren, C.F., Huffman, D.R., 1983. Absorption and Scattering of Light by Small Particles. Wiley, New York. Folkestad, K.T., Hagfors, T., Westerlund, S., 1983. EISCAT: an updated description of technical characteristics and operational capabilities. Radio Science 18, 867–879. Friedrich, M., Kirkwood, S., 2000. The D-region background at high latitudes. Advances in Space Research 25, 15–23. Gabrielli, P., et al., 2004. Meteoric smoke fallout over the holocene epoch revealed by iridium and platinum in Greenland ice. Nature 432, 1011–1014. Giebeler, J., Lu¨bken, F.-J., Na¨gele, M., 1993. CONE—a new sensor for in-situ observations of neutral and plasma density fluctuations. In: Proceedings of the 11th ESA Symposium on European Rocket and Balloon Programmes and Related Research, Montreux, Switzerland (ESA SP), ESA-SP-355, pp. 311–318. Gumbel, J., et al., 2005. The magic rocket campaign: an overview. In: Proceedings of the 17th ESA Symposium on European Rocket and Balloon Programmes and Related Research, Sandefjord, Norway (ESA SP-590), pp. 139–144. Havnes, O., Naesheim, L.I., 2007. On the secondary charging effects and structure of mesospheric dust particles impacting on rocket probes. Annales Geophysicae 25, 623–637. Havnes, O., Trøim, J., Blix, T., Mortensen, W., Næsheim, L.I., Thrane, E., Tønnesen, T., 1996. First detection of charged dust particles in the Earth’s mesosphere. Journal of Geophysical Research 101, 10839–10847. Hedin, J., Gumbel, J., Rapp, M., 2007. On the efficiency of rocket-borne particle detection in the mesosphere. Atmospheric Chemistry and Physics 7, 3701–3711. Hora´nyi, M., Gumbel, J., Witt, G., Robertson, S., 1999. Simulation of rocket-borne particle measurements in the mesosphere. Geophysical Research Letters 26, 1537–1540. Hunten, D.M., Turco, R.P., Toon, O.B., 1980. Smoke and dust particles of meteoric origin in the mesosphere and stratosphere. Journal of Atmospheric Science 37, 1342–1357. Iwagami, N., et al., 1998. Polar thermosphere–stratosphere photochemical coupling experiment: two rocket measurements in polar winter at 69 N. Earth, Planets and Space 50, 745–753. Lu¨bken, F.-J., Rapp, M., Hoffmann, P., 2002. Neutral air turbulence and temperatures in the vicinity of polar mesosphere summer echoes. Journal of Geophysical Research 107 (D15), 4273–4277. Lynch, K.A., 2005. Multiple sounding rocket observations of charged dust in the polar winter mesosphere. Journal of Geophysical Research 110, A03302, doi:10.1029/2004JA010502. Marti, J., Mauersberger, K., 1993. A survey and new measurements of ice vapor pressure at temperatures between 170 and 250 K. Geophysical Research Letters 20, 363–366. Megner, L., Rapp, M., Gumbel, J., 2006. Distribution of meteoric smoke—sensitivity to microphysical properties and atmospheric conditions. Atmospheric Chemistry and Physics 6, 4415–4426. Plane, J.M.C., 2003. Atmospheric chemistry of meteoric metals. Chemical Reviews 103, 4963–4984. Plane, J.M.C., 2004. A time-resolved model of the mesospheric Na layer: constraints on the meteor input function. Atmospheric Chemistry and Physics 4, 627–638.

Rapp, M., Lu¨bken, F.-J., 2001. Modelling of particle charging in the polar summer mesosphere: Part 1—general results. Journal of Atmospheric and SolarTerrestrial Physics 63, 759–770. Rapp, M., Strelnikova, I., 2008. Measurements of meteor smoke particles during the ECOMA-2006 campaign: 1. Particle detection by active photoionization. Journal of Atmospheric and Solar-Terrestrial Physics, this issue, doi:10.1016/ j.jastp.2008.06.002. Rapp, M., Thomas, G.E., 2006. Modeling the microphysics of mesospheric ice particles: assessment of current capabilities and basic sensitivities. Journal of Atmospheric and Solar-Terrestrial Physics 68, 715–744. Rapp, M., Gumbel, J., Lu¨bken, F.-J., 2001. Absolute density measurements in the middle atmosphere. Annales Geophysicae 19, 571–580. Rapp, M., Lu¨bken, F.-J., Mu¨llemann, A., Thomas, G.E., Jensen, E.J., 2002. Small scale temperature variations in the vicinity of NLC: experimental and model results. Journal of Geophysical Research 107 (D19), doi10.1029/2001JD001241. Rapp, M., Hedin, J., Strelnikova, I., Friedrich, M., Gumbel, J., Lu¨bken, F.-J., 2005. Observations of positively charged nanoparticles in the nighttime polar mesosphere. Geophysical Research Letters 32, L23821. Rapp, M., Strelnikova, I., Gumbel, J., 2007. Meteoric smoke particles: evidence from rocket and radar techniques. Advances in Space Research 40, 809–817. Rosinski, J., Snow, R.H., 1961. Secondary particulate matter from meteor vapors. Journal of Meteorology 18, 736–745. She, C.Y., Vance, J.D., Williams, B.P., Krueger, D.A., Moosmuller, H., Gibson-Wilde, D., Fritts, D.C., 2002. Lidar studies of atmospheric dynamics near polar mesopause. Eos Transactions AGU 83, 289–293. Siskind, D.E., Barth, C.A., Russell III, J.M., 1998. A climatology of nitric oxide in the mesosphere and thermosphere. Advances in Space Research 21, 1353–1362. Smith, L.G., 1986. Electron density measurements in the middle atmosphere by radio propagation techniques. In: Handbook for MAP, vol. 19, pp. 17–46. Sternovsky, Z., Hora´nyi, M., Robertson, S., 2001. Charging of dust particles on surfaces. Journal of Vacuum Science & Technology A 19, 2533–2541. Sternovsky, Z., Robertson, S., Sickafoose, A., Colwell, J., Hora´nyi, M., 2002. Contact charging of Lunar and Martian dust simulants. Journal of Geophysical Research 107 (E11), 5105, doi:10.1029/2002JE001897. Strelnikova, I., Rapp, M., Raizada, S., Sulzer, M., 2007. Meteor smoke particle properties derived from Arecibo incoherent scatter radar observations. Geophysical Research Letters 34, L15815, doi:10.1029/2007GL030635. Svenes, K.R., Blix, T.A., Hoppe, U.-P., Gumbel, J., Strelnikov, B., 2005. In-situ measurements of neutral temperature in the middle atmosphere by using electrons as proxy. In: Proceedings of the 17th ESA Symposium on European Rocket and Balloon Programmes and Related Research, ESA SP-590, pp. 191–196. Tomsic, A., 2001. Collisions between water clusters and surfaces. Ph.D. Thesis, Gothenburg University. Turunen, T., Westman, A., Ha¨ggstro¨m, I., Wannberg, G., 2002. High resolution general purpose D-layer experiment for EISCAT incoherent scatter radars using selected set of random codes. Annales Geophysicae 20, 1469–1477. Vitt, F.M., Cravens, T.E., Jackman, C.H., 2000. A two-dimensional model of thermospheric nitric oxide sources and their contributions to the middle atmospheric chemical balance. Journal of Atmospheric and Terrestrial Physics 62, 653–667. Voigt, C., et al., 2005. Nitric acid trihydrate (NAT) formation at low NAT supersaturation in polar stratospheric clouds (PSCs). Atmospheric Chemistry and Physics 5, 1371–1380. von Zahn, U., von Cossart, G., Fiedler, J., Fricke, K.H., Nelke, G., Baumgarten, G., Rees, D., Hauchecorne, A., Adolfsen, K., 2000. The ALOMAR Rayleigh/Mie/Raman lidar: objectives, configuration, and performance. Annales Geophysicae 18, 815–833. Wang, L.S., Wu, H., Desai, S.R., 1996. Sequential oxygen atom chemisorption on surfaces of small iron clusters. Physical Review Letters 76, 4853–4856. Watanabe, K., Marmo, F.F., Inn, E.C., 1953. Photoionization cross section of nitric oxide. Physical Review 91, 1155–1158.