new on the onset of hf-induced airglow at...

AFRL-VS-HA-TR-2005-1061

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 109, A02305, doi:10.1029/2003JA010205, 2004

On the onset of HF-induced airglow at HAARP DISTRIBUTION STATEMENT A

E. V. Mishin Approved for Public ReleaseBoston College, Institute for Scientific Research, Chestnut Hill, Massachusetts, USA Distribution Unlimited

W. J. Burke and T. PedersenAir Force Research Laboratory, Hanscom Air Force Base, Massachusetts, USA

Received 21 August 2003; revised 2 December 2003; accepted 19 December 2003; published 13 February 2004.

[i] Observations of airglow at 630 nm (red line) and 557.7 nm (green line) during theFebruary 2002 campaign at the High Frequency Active Auroral Research Program(HAARP) heating facility are analyzed. We find that during injections toward magneticzenith (MZ) the green and red lines gain -5 R within ,-1 s and -20 R within ,-,10 s,respectively. We term this period the onset of the HF-induced airglow. A model of theonset at magnetic zenith is developed. It accounts for background photoelectrons anddissociative recombination of O. It is shown that heating and acceleration of backgroundelectrons dominate the airglow onset. We propose a scenario for the generation of strongLangmuir turbulence for injections outside the Spitze region, including magneticzenith. INDEX TERMS: 2403 Ionosphere: Active experiments; 2483 Ionosphere: Wave/particleinteractions; 2471 Ionosphere: Plasma waves and instabilities; 2481 Ionosphere: Topside ionosphere;KEYWORDS: HF modification experiments, artificial airglow onset

Citation: Mishin, E. V., W. J. Burke, and T. Pedersen (2004), On the onset of HF-induced airglow at HAARP, J. Geophys. Res., 109,A02305, doi: 10.1029/2003JA010205.

1. Introduction known as the parametric decay (PDI) or oscillating two-

[2] A distinctive feature of HF modification experiments stream (OTSI) instabilities [Fejer, 1979]. In the presence ofbackground suprathermal (e > Te) electrons, many more

is the excitation of airglow at 630.0 and 557.7 nm by high- energetic electrons are accelerated than would be in Max-power, high-frequency (HF) radio waves [e.g., Sipter et at., wellian plasmas [Mishin and Telegin, 1986].1974; Bernhardt et al., 1989; Pedersen and Carlson, 2001. [4] Recent observations show that the airglow maximizesGustavsson et al., 2001, 2002; Kosch et al., 2000, 2002]. during HF injections toward magnetic zenith [Kosch et al.,Enhancements up to -. 500 R (Rayleigbs) with the green-to- 2002; Pedersen et al., 2003]. The same is true for thered ratio (gr as high as >0.3 have been reported intensity of Langmuir waves [Isham .et al., 1999] and[Gustavsson et al., 2002; Kosch et al., 2002; Pedersen et electron heating [RietveMd et at., 2003] observed by theal., 2003]. The excitation energies EX of the O('D) and EISCAT UHF radar. Furthermore, the red line is excited atO(IS) states responsible for the red and green lines are er = magnetic zenith even at extremely low effective radiative

1.96 eV and ag = 4.17 eV, respectively. The population of powetic z 2 at eremet low 2003dt

energetic, C > e), electrons can increase significantly due to power P n - 2 MW [Pedersen et al., 2003].stochastic and resonant interactions with plasma turbulence the Spitze region, 0 > wt, reflectdt altitudes H0 below the

generated by a heating wave [e.g., Gurevich et al., 1985; standard reflection altitude Ho where the local plasma

Dimant et al., 1992; Mantas and Carlson, 1996; Gurevich frequency fp ,.o 104Vn-e Hz equals the driver frequency f.and Milikh, 1997; Istomin and Leyser, 2003]. The formerraises the electron temperature Te, the latter accelerates a Here Oc = arcsin( f4+'sinX), f, is the local electron gyro-group of electrons in the high-energy tail of the initial frequency, ne is the electron density in cm 3 , and X is thedistribution. Both effects are well documented [e.g., magnetic dip angle [e.g., Mjolhus, 1990]. At HAARP X -_Carlson et at., 1982; Gustavsson et al., 2001]. 14.50 andf. • 1.4 MHz at altitudes near 200 kin, so that for

[3] Gustavsson et al. [2002] emphasized that to interpret Jo = 7 MHz the Spitze angle is 0, ý_ 5.90., r> 0.1 in terms of electron heating requires unrealizable [6] Figure 1 shows a schematic of ray trajectories for

Te > 2 eV, pointing out the importance of electron acceler- ordinary HF waves injected vertically and toward magnetication. It is commonly believed that electrons are most zenith into a horizontally stratified ionosphere at HAARP.efficiently accelerated by Langmuir (1) turbulence. The Obliquely incident radiation does not form standing-wavegeneration of Langmuir waves is usually described in terms patterns and swelling is absent. Thus, given P0 = 150 MWof nonlinear instabilities of ordinary (o) mode pump waves, and distance R = 250 km, the wave amplitude is E0 L--

4.7.. RV'Po 0.2 Vim. With Jo = 7 MHz and Te = 0.1 eV,Copyright 2004 by the American Geophysical Union. the HF energy density at the reflection point is Wo = --

Copyight10-4 4T sufcen

0148-0227/04/2003JA010205$09.00 4eTe, sufficient to drive the PDI/OTSI at Ho [Fejer,

20050620 137 A02305 1 of 10

A02305 MISHIN ET AL.: HF-INDUCED AIRGLOW ONSET AT MAGNETIC ZENITH A02305

240 ____ _/ H0_ _ state from dissociative recombination grows significantly0_ - with Te and with the vibrational temperature of oxygen ions

B YThr[Guberman, 1997; Peverall et al., 2001]. Furthermore,Hmz- charge exchange 0+ + 02 -4 O + 0 is the dominant source

220 of 02 production in the nighttime F region. Its rate increases- _ significantly whenever 02 is vibrationally excited [Viggiano

"and Williams, 2001]. We expect a high degree of excitation. 200 - I of molecular species in the HF-illuminated region and thus

/[ enhanced yields of the green line emissions.[io] A consistent theory of the HF-induced airglowI / accounting for the ion/neutral chemistry, modification of

the heated spot, and self-focusing has yet to be worked out.

0I i I I1 We emphasize that responsible processes develop within tens0 40 80 of seconds. However, considerable airglow appears within

first few seconds after the HF transmitter turns on. WeRange (km) designate this interval as the onset of HF-indsced airglow.[ii] This paper develops a model for the onset of

Figure 1. Schematic of ray propagation for ordinary airglow enhancements at magnetic zenith accounting forHF waves injected toward local vertical and magnetic the roles of O2 dissociative recombination and energeticzenith. The magnetic field B direction is indicated by photoelectrons. The following section describes the onsetarrows. The heights of reflection and upper hybrid characteristics from the HF heating experiments atresonance are shown by horizontal lines. A light dashed HAARP. Section 3 describes the model. In particular, weline shows the Spitze angle direction. The regions of propose a scenario (section 3.3) for the generation ofexcitation of Langmuir turbulence appear as bold lines near strong Langmuir turbulence outside the Spitze regionthe reflection points, including magnetic zenith with upper hybrid waves as

the primary source. The final section compares modeling1979]. However, mismatch of frequencies at <H0 suppress results with optical measurements.these instabilities for injections outside the Spitze region.

[7] Gurevich et al. [2002] suggested that decreased 2plasma densities within striations generated by heating Airglow Onset: HAARP, February 2002waves permit necessary phase matching [cf. Muldrew, [12] The HAARP facility is located outside of Gakona1978]. Changes in the refraction index (self-focusing) due Alaska (62.40 N, 145.15'W). During the course of HFto striations develop within tens of seconds after turn-on and heating experiments between 03:00 and 05:00 UT in Feb-explain some features of the spatial distribution of the HF- ruary 2002 several passes of Defense Meteorological Sat-induced airglow. However, the rise time of Langmuir waves ellite Program (DMSP) satellites flew to the east and west ofis ,-.10 ms [Isham et al., 1999]. Besides, striations are the HAARP location. Each of these spacecraft carries a pairgenerated in the upper hybrid layer [e.g., Vaskov et al., of upward looking particle spectrometers designed to mea-1981; Lee and Kuo, 1983]. Hence the reflection height of the sure fluxes of downcoming electrons and ions with energiesheating wave must be at or above this height. For magnetic- between 30 eV and 30 keV. Consistent with prevailing quietzenith injections at HAARP, this can be satisfied only if geomagnetic (Kp = 2) conditions, the equatorward boundaryfo < 5.4 MHz. However, the strongest airglow at HAARP of auroral precipitation was several degrees in latitudeoccurred with higherfo values [Pedersen et al., 2003]. poleward of Gakona. However, both before and after the

[8] Kuo et al. [ 1997] showed that Langmuir waves can be local time of sunset, the spectrometers detected fluxesexcited by upper hybrid (uh) waves with amplitudes exceed- of Rdowncoming electrons whose spectra monotonicallying the threshold value EuAh - 0.15 V/m. Such uh-waves can decreased with energies between 30 and 100 eV. Thesebe generated through the linear conversion of the o-mode on subauroral fluxes consist of photoelectrons that originatedpre-existing field-aligned irregularities [Wong etal., 1981] or in the still sunlit southern ionosphere.through the parametric decay o -- uh + lh [e.g., Istomin and [13] Intense green-line emissions were observed betweenLeyser, 1995]; lh stands for lower hybrid waves. Linear 03:48 and 05:00 UT (17:48-19:00 LT) on February 13,conversion within the upper hybrid layer proceeds with no 2002 events of the HAARP optics campaign. Readers areset threshold. However, the parametric decay develops ifE0 > referred to the report of Pedersen et al. [2003] for graphicEgh u 1.6 f2 mV/m, provided that the inequality 0.015 < examples of HAARP- induced airglow and the intensities ofIfo s • fJ < 0.5 MHz (integer s > 3) is valid. Here fo red/green-line emissions recorded during the period ofstands for the heating frequency in MHz. The rise time of interest. 0-mode waves were injected toward magneticthe uh-wave is Tuh - 1-3 ins. Consistent with observations zenith at fo = 7.8 MHz and at full power of 0.94 MW[Isham et al., 1999], this makes the generation of Langmuir (P0 "• 165 MW). The transmitter was programmed to turn onturbulence possible within ,-.10 ms. for 5 minutes exactly on the minute. Subsequent pulses

[9] Besides electron impact, dissociative recombination followed 5 minutes pauses. However, the fourth pulse in the(DR) of O2 is an efficient source of the 0('D) and 0('S) sequence started at 04:21:18 UT. A transmitter problemstates. Since the rate of dissociative recombination decreases produced a false start, 04:30:00-04:31:00 UT, prior to thewith Te, it is usually not considered in the theory of the HF- long pulse that began at 04:32:12 UT. In the course of thisinduced airglow. However, the quantum yield for the 0(1S) long pulse, the critical frequency of the F layer dropped

2 of 10


- volume and the background region. These data show that50• the difference between natural green-line emissions from

50 ....... ...... I ................. I "magnetic zenith and -70 km apart was less than 1 R. TheiI same is true for the red-line emission before the heater

40 . ....... .i........................ turned on. This suggests that the background airglow wascc .nearly uniform. When the heater turned on, green (red)-line

"...- .................. ... intensities increased by -,5 (-30) R within ,4.5 (-.16.5) s.II I It is relevant to note that Kosch et al. [2002] reported

L•O -,,500 R enhancements of the red- and green-line emissions<20 . ... i..................I produced by the EISCAT superheater (P0 "- 600 MW)

I ¶within the first 5-s frame.

1 0 ...4 . . . . .. .. . . . . I . . . . . . . . .. ..;

/ I \ •,. I 3. Modeling Airglow Onset

0 - ................... *-.@( .......... /.. [17] For modeling purposes we assume that the majorI species constituents of the neutral atmosphere and iono-

200 ........... i.... ...................... spheric plasma are well described by the MSIS-E [Hedin,o oq . I 1991] and IRI [Bilitza, 2001] models, respectively. Although

150\I I photodissociation of 02 accounts for ,-50% of the back-150 ........... . .................. . ... ground airglow, it is unaffected by heating and hence is not

SI . included. Basic processes and their rate coefficients areo I discussed at length by Solomon et al. [1988], Rees [1989],

100 .I .... ..................... Witasse et al. [1999], and Fox and Sung [2001]. Unless

5 0 .. .. . . . . . . . . . . 1 1 . . . . . . . . . . . . . . . . . i . 2 0 5 • . . . . . ... . . . . . . . . . .. .i : , . . . . . . . !... .. .

50 ...... 20.... .... .~2 0 .. . .. . . ... .. . . .. ,. .. : . .

0421 0424 0427 0430 0433 200................./UT - /

I.,: / *I,Figure 2. Heater-induced green (top panel) and red .195 /S•5.............. :........ •./(bottom panel) lines during February 13, 2002 injections. - 1 : .Vertical solid and dash-dotted lines indicate heater tum-ons ,and tum-offs, respectively. ,

1 9 0 g ... . . .. ............... ....below the transmitter frequency and the sequence was - -terminated. ...

[14] To account for natural (background) variations, apolynomial fit was made to intensities measured along the - /magnetic meridian in each all-sky image, with the heated 340 ........... ....... : ./.- ..area blocked out. Throughout the studied period the differ- i..

ence between the background airglow b, and that of the 320 /heated volume Ix, during heater-off periods, was less than ' /:,-3-5 and ,-7-10 R for green- and red-line emissions,respectively. This small difference determines the accuracy .300............... / ...........of heater-induced airglow measurements A> = Ix - b>. / \

[15] Figure 2 presents a detailed view of the last three HF- 5280

induced enhancements of February 13, 2002 at fo =/ -7.8 MHz. Data points are from two red line exposures . ,compiled at 0 and 24 s after the beginning of each minute, 260 "k I . ....................and green-line exposure beginning at 12 s into the minute. * - , -After 04:00 UT, the exposure duration was 7.5 s. By 240 "examining cases when the heater turned on part way 0426 0427 0428 0429through an exposure, we see that the green-line emissions UTincreased by ,-'5 (,-10) R within -1.5 (-,7.5) s, while thered line gained ,-25 (,-30) R within 1-3.5 (,-,20.5) s. Figure 3. Heater-induced green (top) and red (bottom)

[16] Starting at 04:27:15 UT on February 9, 2002, a P 0 ý- lines during February 9, 2002 injections. Dots and asterisks124 MW wave at a frequency 6.8 MHz was injected toward stand for emissions from the magnetic zenith and back-magnetic zenith. Figure 3 shows red- and green-line emis- ground regions, respectively. Vertical dashed line indicatessions before and during the pulse turn on from the heated the heater turn-on time.

3 of 10


otherwise noted, we use the rate coefficients of Fox and 10 1 sSung [2001].

[18] As mentioned above, radiative emissions from aheated volume reflect the superposed results of complexwave, plasma, and chemical interactions. To render our Nmodeling of this complex chain of interactions intelligible, 0we divide this section into three parts. The first subsection • 10concerns the basic contributors to the natural and artificially 50excited radiation budgets. The subsection on the back- "T.ground ionosphere estimates the distributions of ion species ,and energetic photoelectrons present at heater altitudes prior 1

to turn on. The third subsection describes a flow chart 10encapsulating our concept of how heater-injected energy istransformed into the various wave modes that heat and/oraccelerate ambient electrons that interact with ambient 0 1neutrals to produce onset green- and red-line emissions. 10 " 10 102

Electron energy (eV)3.1. Basic Processes

[19] The red (r) and green (g) line-photons are emitted by Figure 4. Cross-sections of the 0(1D) and 0(1S) excita-atomic oxygen in the transition from the 0(1D) and 0(1S) tion and the Schumann-Runge dissociation by directstates to 0(3P) and O(1D) states, respectively. In photoelectron impact. Squares and dots are the measured cross-chemical equilibrium, the volume emission rate ,qx is sections from Doering and Gulcicek [1989]; solid lines arecalculated from interpolations by Majeed and Strickland [1997].

=A),. [C0,] = A), QX (1) [Guberman, 1988, 1997] appear to have reconciled thiseql = A "[ ]= A L ), + A:Cx Dcontroversy. They established that ot and >3 R depend not

Here [O] stands for the density o or in cm-3; only on T. but also on the vibrational population nu -s og pro(dut ad [02(j)]/[02] of O, where j = 0, 1,... designates 2the

Q, and LZ are the corresponding production and loss rates, vibrational level.respectively; Ar ý- 0.007, AE, •_ 0.009, Ag _- 1.2, and A g [22) For the ground vibrational state, 03(°) - 1.15, ct(°) - 21.3 are the Einstein transition probabilities in s-. The 10-7Te0,63 [cf. Fox and Sung, 2001], andcolumn emission rate (in R) is found by integrating (1)along the line of sight /(°) -0.063 exp 2 .23 (1 - 0'39 (4)

( \ 0re re9 1(41rIx = 10-6 Tl(z)dz (2)

for 1 < T < 10, where re = Te[K]/300. Importantly, theinequalities "9'(1,2) > f0(O) and 03(1,2) < 0(0) hold. For the

[20] The rates of electron impact excitation 0 + e - Maxwell-Boltzmann distribution of NU)1 with vibrational0(iD, S) + e and the Schumann-Runge dissociation 02 + temperatures Tv 0.25 eV, one obtainsD[3 •R 1 2 ande -- 0 + O('D) + e are calculated as follows PDR o

[23] Collisional quenching 0" + X -- 0 + X is importantQ' = [0]. ,= [01 .41 i a),(E:)(e)dE: (3) for the O(1D) state, while for 0('S) at altitudes >150 km it

can be ignored. Here X stands for 0, 02, N2, or eth. Thecorresponding loss rates are LY = 10-1 ,4•[X], where 0 -

Here '(Z) is the electron differential number flux, 7e,, ex, 0.65 Tro'iX, 2,-' 3.2 exp (0.225/Tr,), xN7 f-- 1.8 exp (0.36/T,),and a), are the rate coefficients, threshold energies, and and 4q - 2 8 .7 T-'91. Figure 5 shows the altitude profile ofcross-sections, respectively. We employ electron impact the total loss rate calculated with the MSIS-E and IRIcross-sections suggested by Majeed and Strickland [1997]. parameters at 04:30 UT, February 13, 2002.Figure 4 shows three of the major electron impact cross [24] Finally, the radiative cascade (RC) 0('S)'-- 0(1D) +sections. It is worth noting that % (Te) and %' (Te) calculated hv 5577 yields Q0 =with a Maxwellian distribution of thermal electrons areclose to the approximations used by Mantas and Carlson 3.2. Background Ionosphere[1996] and Gurevich and Milikh [1997], respectively. [25] Simultaneous observations from a digisonde located

[21] The dissociative recombination of 02 produces at the HAARP site were used to correct the IRI model and> x =%Rne[02]. The rate coefficients are usually approx- to determine that the reflection height at magnetic zenith

imated as otR o., where >R are the quantum yields. decreased from ,-250 to ,-235 km between 04:00 andEstimated yields for 0(iS) were subject to substantial 04:30 UT on February 13, 2002. At the same time, thediscrepancy in the literature for many years. Recent shadow height increased from - 150 to "-230 km indicatingexperiments [Kella et al., 1997; Peverall et al., 2001; the presence of photoelectrons [e.g., Doering et al., 1975].A. Petrignani et al., Vibrationally resolved rate coefficients Furthermore, the southern-hemisphere region magneticallyand branching in the dissociative recombination of O, conjugate to HAARP remained in sunlight, even after localsubmitted to Journal of Chemical Physics, 2003] and theory sunset. Thus a large fraction of the photoelectron spectrum

4 of 10


400 Figure 7. First, upper hybrid waves are generated due tothe parametric decay o -- uh + lh. Its threshold field isEoh ,- 0.08 V/m if the matching conditions are met at3. 2 e350 Xuh r 1 where r, is the thermal electron gyrora-dius and kuh is the uh-wave vector. Landau damping of

ES................ t-scale lower hybrid waves raises the threshold other-S300 ... .. . .."h r

S': wise [Mishin et al., 1997]. Forfo = 7.8 (6.8) MHz and 014.5', the matching conditions at the reflection height H0

_ 250 ............................... are easily satisfied for xuh ! 0.3 (•0.15). Finally, the freespace field of the incident wave Eo !-- 0.25 (0.2) V/mexceeds Eoh.

[29] When the amplitude of the primary uh-wave Euhexceeds -10 mV/m, it excites other, lower frequency,uh-wave with the growth rate of order yuh • T-

1 [Zhou150'u

10-1 et al., 1994]. The same is true for subsequently generateduh-waves. The frequency-step of this spectral transfer is

0(1 -1 quite small I zhJ/1Wh - O'lXuhbkuh/kuh - 0.2/vme/mi, whereD) loss rate, s- me/mi represents the electron/ion mass ratio. As many

spectral steps occur before the parametric instability satu-Figure 5. rates, the resulting uh-energy spectrum consists of a large

number A > 10 of spectral peaks, each of order E0/(y0hT~h)

generated in the conjugate ionosphere [e.g., Peterson et al., (well-known weak turbulence "cascading" [e.g., Sagdeev1977] had access to the ionosphere above HAARP. Electron et al., 1991]). Thus the total uh-energy density can bespectrometers on DMSP satellites observed the high-energy estimated as W./A - Wo " 7.5 (5) eV/cm 3 with the r.m.stail of the conjugate photoelectron spectrum. amplitude E. n 4'• , -W* ". --AE o.

[26] Figure 6 shows altitude profiles of the background [30] Upper hybrid waves with amplitudes Eh exceedingelectron temperature To and electron and 0+ densities the threshold value Etuh ! 0. /Ixu1/2 V/m generate short-calculated from the corrected IRI model. The twilight 02 scale, k, > 3-'/2/r,, Langmuir waves [Kuo et al., 1997].density above -200 km is defined by the charge exchange Due to a weak dependence of the growth rate y, on kuh,and dissociative recombination one can take Euh "- E*. As this value greatly exceeds the

threshold EuAh, the growth rate can be evaluated as -y, -[02 [0 [O÷]() eEuh/Euh [Kuo et al., 1997], which amounts to -y "[21=X (5) ' 2 - -

Ocne Ve.elO2XuhAE2. Here ve is the electron elastic (transport)collision frequency.

Here the rate of charge exchange %, after Viggiano and [31] The growth of Langmuir waves at T/i7 < 4 saturatesWilliams [2001] is yia induced scattering by ions that transfers energy toward

XC X(n)(n(0) + n(o1) [4.9 lg(", + 1) - 2.51 400--e- Ce 0"o _ 'l-. ', . '. -

) 2.4 (6)

' f + 6.05 exp 350 pnInj , s .. . .. ... .. .. . .. . ..... .. . . .. . . .

oL is the recombination rate in cm 3 /s, nJ = [02(j)]/[02],

[Y] stands for the density of species Yin cm- , and Tr in = E 300 ...............1.5Ti + T, - I with "i., = Ti,[K]I/300.

[27] The IRI model predicts Te = To < 0.2 eV (Figure 6). "TAt these temperatures and n. > 105 cm 3 , a Maxwellian • 250 ...

distribution FM is a good representation of the thermal "a T2electron distribution function (Feh) in the F region [Mishin .. •. •et al., 2000]. The differential number flux Ps(E) of photo- 200 . .......

electrons in the F region [e.g., Rees, 1989] at e > 6 = mv 2

5 eV can be approximated by a power-law function 150

500 1000 1500 2000 2500 3000s(e)= .- 0n5vs P,-- (7) electron temperature, K

3 0310 410 53 1where ns < 100 cm- 3 and p. = 3. number density, cm151

3.3. HF-Perturbed Ionosphere Figure 6. Altitude profiles of the electron temperature

[28] Our scenario for the excitation of plasma turbulence (solid lines), electron density (dashed line), and 0O densityand subsequent electron energization in the HF-illuminated (dotted line). To, T1, and T2 stand for 0, 1, and ý-1.5 s afterregion at magnetic zenith is represented schematically in turn-on.

5 of 10


--------- Bpattern at H0, where the PDI/OTSI instabilities developBacgr d Ino e during injections within the Spitze region [Fejer, 1979]. We

|-HF-wave --------- believe that this is the key factor in exciting strong airglowat magnetic zenith.

H r ul 3.3.1. Heatingturb ['o -- uh -e t> uh + lh ] _ [33] Collisional damping of high-frequency o-, uh-, and

io OS M" / +-waves is the major source of (stochastic) electron heating.OTSI ''.. Mishin et al. [2000] showed that at [N2] > 107f2 cm- 3 the

- process of N2-vibrational excitation dominates the forma-k,> 1/ 3 tion of the distribution function of ionospheric electrons

Swith energies 3.5 > E > Fvib - 1.8 eV, which can beInd. Scat~by Ions I represented as follows

Langmuir condensatekl---0F( > +ib) -Ft( b v._ K(E.b)

Mod.inst& collapse 1

Here r 2 (e) ViiKE)/Vee(E) + 0.5C/Te, where liil/vee is the ratio

between electron inelastic and electron-electron collisionfrequencies. A total HF wave energy density WHF <

heating acceleration 10-2noTe is assumed. At energies E < Cvib the distribution is. .close to FM, while at e > 3.5 eV it may slightly deviate from

..i..o..i.ti.. FM due to the 0(1D) and O(1S) excitation.S~~~~~isocization "..eeg aac

recombination excitation [34] The electron temperature can be evaluated from theionization energy balance

.......................... ............................... 1.5nei9T /Ot = r, - v.6 . neTe - V llq(9)

Ion/neutral chemistry where q,11 is the (parallel) electron heat flux, r.e" -- e WHF isnset airglow the volume heating rate, and 8 = (vii/ve) is the coefficient of

Ionosphere modification inelastic losses averaged over the total distribution Ft,.[35] Given v, - 500 s-', r, is >t40 (25) keV/cm 3/s. Note

that to match their optical observations, Mantas andSaturated airglow Carlson [1996] used the values of r", between ý--50 andF125 keV/cm 3/s at --260 km forf• A__ 5 MHz. Gustavsson

Figure 7. Block-diagram showing synergy of a model of t al. [2001] used r, ý 60 keV/cm Is forfo 4 MHz to fit

HF-induced airglow at magnetic zenith. Shaded boxes the electron temperature profile maximum 4000 K atindicate the steps to be worked out. _220 km observed by the EISCAT UHF radar. Neither of

them accounted for the decrease of b(Te) due to the deviationof the thermal electron distribution from Maxwellian. Taking

small k [e.g., Zakharov et al., 1976]. This process is that into account, the calculation results of Mantas andgoverned by the pump uh-wave power. In particular, if the Carlson [1996] and Gustavsson et al. [2001] can be scaledwidth Ak, of the parametrically unstable 1-wave spectrum to the T1.2-profiles in Figure 6, pertaining to our case.exceeds Ak, c!' k,'kv•ly,, the Langmuir wave energy W, 3.3.2. Accelerationaccrues in a state with k -+ 0 (Langmuir condensate) [36] Resonant lh- and 1-wave-particle interaction accel-[Zakharov et al., 1976; Zakharov, 1984]. The spectral width erates electrons. Musher et al. [1978, 1986] analyzed incan be evaluated as Ak ", 3(fo/f) 2 W*/(neTr,). Thus, for detail the dynamics of lh-waves excited by an externalE0 > 0.15(for/e)1 /V/A 7 0.2 Vim the "condensate" source. At 1 < TJTi < 4, high-frequency waves, fh >condition Ak, > klv••yl is fulfilled. The dynamics of the fc22me/mi, are dominated by induced scattering by ions.Langmuir condensate is defined by the modulational insta- The rate of spectral transfer toward smaller frequencies isbility and collapse leading to establishment of'strong "hh " (fZ/fh)WhfloTe, where Wh is the energy density of(cavitating) Langmuir turbulence [e.g., Gateev et al., lower hybrid waves. Equating the growth rate ýYuh to iYh1977; Zakharov, 1972, 1984]. yields the energy density at the saturated state WlhlnoTe,,

[32] We emphasize that this applies not only at the -Yuhfh/fc2 -, 10-5.5reflection layer H0 but also well below it, wherever the [37 The dynamics of h-waves in the low-frequency, fh <matching conditions for the parametric decay o -* uh + lh f, 2me/mi, region is dominated by the lower hybridare met and local values of Xuh < 1. Given the plasma collapse [e.g., Musher et al., 1978]. The threshold energydensity profile shown in Figure 6 andfo ,-" 6-8 MHz, the density is quite low W[hInTe Xlh(me/mi)(f/fo)2 < 106,altitude extent of this region is ,-10 km. One should and is surely exceeded in our case. In the course of collapse,compare this value with a few 100-m size of the Airy the longitudinal and transverse dimensions of a th-cavity,

6 of 10


II - l_Lj/mi/xlhme, diminish. Finally, the lh-wave energy in 10 8a collapsing cavity is absorbed by particles via Landaudamping [Musher et al., 1978, 1986; Sotnikov et al., 1978;Shapiro et al., 1993]. The density of accelerated electrons inthe magnetic field-aligned high-energy tail of the distribu- 010tion function, determined by assuming resonant particles adcarry away all of the energy pumped into collapsing cavities '9'[e.g., Shapiro et al., 1993], is of order <10-5ne. Thislh-contribution appears insufficient to account for the 10observed HF-induced airglow. It is worth noting thatlh-collapse also leads to transversely accelerated ions that EM 101 F(2) 0 2absorb the collapsing energy at approximately the same rate min m 0as electrons [Sotnikov et al., 1978]. energy, eV

[38] Strong Langmuir turbulence consists of an ensembleof collapsing cavitons that transfer the wave energy toward Figure 8. Differential number fluxes of photoelectronssmall scales. Small-scale, k > WkJVmin, waves are absorbed (solid lines) and accelerated electrons (dash-dotted lines) inby fast v > vmin > VT/m electrons via Landau damping, cm-s 1 ster-eV- for n= 1 (line 1) and 10 (line 2) cm 3 .thereby increasing the population of suprathermal electrons.The distribution of accelerated electrons Fa(E) = 21a(e) in and the wave energy flux transferred by collapsing cavitonsweakly magnetized f, < f, plasmas can be found from the and absorbed by accelerated particleskinetic equation [e.g., Galeev et al., 1977]

W Fý 1 ~ W D0a (m -E~, w, k (15)-dk 8F1 i 241rnoTe,9--5 7 5= - V -Tv W/]--d ( 1 0 )

Here Ek, is the wave amplitude in a cavity of the scalewhere w,/(27) -,'f, and Wk are, respectively, the frequency length lmin - k;' -_ Vmin/wp when collapse is arrested andand spectral energy density of Langmuir waves. The latter is 4o is the differential number flux of ionospheric electrons.determined from the requirement for energy balance [41] For a Maxwellian electron distribution, one has m

- 20T,. When photoelectrons are present, 4o(e) -+ 4,() atdWk d dk I >> Te, and mnin is defined bydt d k k ýdl] rk Wk (11)

Here esin - 30(nsTe/W')2/5 eV (16)

valid for the energy density of Langmuir turbulence W, >>f -o272 4 noTememi[Mishin and Telegin, 1986]. Given n, = 10 cm-3,

k - (2)"\n0 ) aF (-k'-) (12) ne =6 10 cm 3 , and W _ 1- lO 4 neTe, from (16) one getsEmin -- 15 eV. The number of accelerated electrons exceeds

is the rate of Landau damping; k(t) , (te - t- 2 is defined the photoelectron background in the energy range e > e'minby the collapse law in the absorption region [Pelleiter, by a factor of • ,• 10 For energies P < Emin, the1982], and t, is the collapse time. distribution is unaffected. Figure 8 shows differential

[39] The steady state solution of (10)-(12) at e > Emin = number fluxes of photoelectrons (7) and accelerated"ImeVmin is a power-law function [Pelleiter, 1982; Shapiro electrons (13) for n1= 1 and 10 cm-3 , n 6 10 cm 3,and Shevchenko, 1984] that yields the differential number and W- lO-4 neTe•flux

4. Discussion¢•(•) min X4 (Em) n naVmin\P) (13) [42] Figure 9 shows the components of the equilibrium4Itemin ~volume emission rates r,(1) for the period near 0430 UT.

To calculate the contributions of photoelectrons (7) and(pa 0 - 0.75). One-dimensional (magnetically field-aligned) accelerated electrons (13), n, 10 cm 3 and Wn - 10 4neT2numerical modeling of electron acceleration by strong were chosen, yielding Emin ý- 15 eV forfo = 7.8 MHz. AsLangmuir turbulence yields 4a(ill) - F-1, consistent with expected from (4), the dissociative recombination contribu-the 1D scaling law for Langmuir collapse [Galeev et al., tion to the red line is reduced much more than that for the1983; Wang et al., 1997]. A flat distribution of accelerated green line. Apparently, heated electrons dominate the red-electrons is consistent with the observations of Carlson et line emission, while accelerated electrons are the majoral. [1982]. contributor to the green line. Note a factor of about 2

[40] The minimum energy Emin and the density of par- difference between calculations with Fm(T1) and Flh(Tl)ticles in the high-energy tail na are determined by the consistent with Mishin et al. [2000] (it is !--3.5 for Te = T2),boundary condition [43] In the heated volume the densities of excited oxygen

atoms and of 02 ions grow. During the onset period [0('D)]a(Emnin) = •0(Emin) (14) and [02] remain well below the new equilibrium values that

7 of 10


from (2). The results are shown in Figure 10 together witha 630.0 n the observed intensities (section 2). There is good apparent

0@O0- agreement with the HAARP observations.350 0• [45] We emphasize that the onset values of the HF-

06 \\induced airglow amount to < 10% of the saturated values.00

E 0• As the reflection heights of 7.8 and 6.8 MHz waves at00 °magnetic zenith are below the upper hybrid resonance,

0 300 \I6 similar to Figure 1, striations are not generated. Thus self-._ -focusing due to striations seems unlikely. This indicates that

a the ionosphere's parameters in the heated volume are

250 significantly modified. Apparently, a complete explanationSof artificially induced airglow involves a complex chain of

plasma and chemical interactions within the heated volume200 !.• jand has yet to be worked out.

5. Conclusionb 557.7 nm [46] Observations of airglow at 630 nm (red line) and

350- 557.7 nm (green line) acquired during the February 2002

350, ,• 'campaign at the HAARP heating facility have been ana-lyzed. During injections toward magnetic zenith the inten-

" I' sities of green and red line emissions gain -5-10 R within6300o\ -,1 s and ,-.20 R within -10 s, respectively. We call this"o \\ \ period the onset of HF-induced airglow and develop an•_ \ explanatory model for the growth of emissions from mag-

250 \ \\ netic zenith. The model accounts for the roles played by250\ ambient photoelectrons and dissociative recombination of

\\ • 0' ions and shows that heating and acceleration of back-ground electrons, possibly by strong Langmuir turbulence,

200, \ dominate the airglow onset.

0 1 2 [47] We also propose a new scenario for the generation of10 10 10 strong Langmuir turbulence for injections outside the Spitze

photons cm-3 s 1 region, including magnetic zenith, with upper hybrid wavesas the primary source of Langmuir waves. Those waves

Figure 9. Equilibrium volume emission rate components accumulate in the region of small wavevectors due to

of the red (a) and green (b) emissions: Dissociative induced scattering by ions (Langmuir condensate). Waves

recombination with [O]02 from the background Te = To in the condensate are subject to the modulational instability

(thin solid line) and heated Te = T2 (bold solid line) regions; and collapse thus leading to strong turbulence. This scenario

ph6toelectrons (thin dashed line) and accelerated electrons(bold dashed line); 0(1S) cascading from the background(circles) and heated (dots) regions; thermal electrons with 30 0Te = To (crosses), T1 (filled triangles), and T2 (opentriangles); Maxwellian electrons with Te = T1 (asterisks). 250 .......

are of the order of QaILr and xce [0 2]/oc(T 2), respectively. 20.................Corresponding time-variations of the contributing compo-nents to the total volume emission rate during the onset can jr 15be evaluated assuming that A F < ( < min[1I/(neot2),lIL]-15 s and L,. '- const(t) at 200-300 km. Here C t - to and 1the heating time to -' 1 s. As a result, one gets

8[0+] = [O+]/[O-]o-1 f- [ n,[o(To) - o(T 2 )] - (17) 5

This yields 010 5 10 15 20

,%(() n,'q + i 1DR nfe[&(To) -- T(12 )] - time, s(18)

(qeqLr + 0.5TPR[Lr + no(T2 )]6[Oj])( Figure 10. Calculated intensities of the HF-induced red(solid line) and green (dashed line) emissions during the

[44] Using (18) and Figure 9, the variation of the red- and onset period. The observed intensities are shown by opengreen-line emissions during the onset period are calculated circles (red) and triangles (green), respectively.

8 of 10


applies not only at the reflection layer but also well below it, Kosch, M., M. Rietveld, T. Hagfors, and T. Leyser (2000), High-latitude

wherever the matching conditions for the parametric decay HF-induced airglow displaced equatorward of the pump beam, Geophys.Res. Lett., 27, 2817.o -+ uh + lh are met and local values of xuh < 1. For Kosch, M., B. Gustavsson, and M. Rietveld (2002), An overview of high-

injections toward magnetic zenith and heating frequencies latitude induced aurora from EISCAT, paper presented at World Space>5.4 MHz, this occurs below the altitude of the upper Congress, Am. Inst. of Aeron. and Astron., Houston, Tex.

Ta x hKuo, S., M. Lee, and P. Kossey (1997), Excitation of oscillating two streamhybrid resonance. The large altitude extent of the strong instability by upper hybrid pump in ionospheric heating experiments atturbulence region appears to be the key factor in exciting Tromso, Geophys. Res. Lett., 24, 2969.strong airglow at magnetic zenith. Lee, M., and S. Kuo (1983), Excitation of upper hybrid waves by a thermal

parametric instability, J. Plasma Phys., 30, 30,463.Majeed, T., and D. J. Strickland (1997), New survey of electron impact

[48] Acknowledgments. This research was supported in part by the cross sections for photoelectron and auroral electron energy loss calcula-HF Active Auroral Research Program (HAARP) under AFRL contract tions, J Phys. Chem. Ref. Data, 26, 335.F19628-02-C-0012 with Boston College and by Air Force Office of Mantas, G., and H. Carlson (1996), Reinterpretation of the 6300-A airglowScientific Research under tasks 2311 AS and 2311SD. enhancements observed in the ionosphere heating experiments based on

[49] Arthur Richmond thanks Bjorn Gustavsson and Michael analysis of Plateville, Colorado, data, J. Geophys. Res., 101, 195.T. Rietveld for their assistance in evaluating this paper. Mishin, E., and V. Telegin (1986), Spectrum of suprathermal electrons in

the auroral plasma, Soy. J. Plasma Phys., 12, 509.References Mishin, E., T. Hagfors, and W. Kofman (1997), On origin of outshifted

plasma lines during HF modification experiments, J. Geophys. Res., 102,Bernhardt, P., C. Tepley, and L. Duncan (1989), Airglow enhancements 27,265.

associated with plasma cavities formed during ionospheric heating Mishin, E., H. Carlson, and T. Hagfors (2000), On the electron distributionexperiments, J. Geophys. Res., 94, 9071. function in the F region and airglow enhancements during HF modifica-

Bilitza, D. (2001), International Reference Ionosphere 2000, Radio Sci., 36, tion experiments, Geophys. Res. Lett., 27, 2857.261. Mjolhus, E. (1990), On linear conversion in a magnetized plasma, Radio

Carlson, H., V. Wickwar, and G. Mantas (1982), Observations of fluxes of Sci., 25, 1321.suprathermal electrons accelerated by HF excited Langmuir instabilities, Muldrew, D. (1978), The role of field-aligned ionization irregularities in theJ. Atmos. Terr. Phys., 12, 1089. generation of the HF-induced plasma line at Arecibo, J. Geophys. Res.,

Dimant, Y., A. Gurevich, and K. Zybin (1992), Acceleration of electrons 83, 2552.under the action of intense radio-waves near electron cyclotron harmo- Musher, S., A. Rubenchik, and B. Sturman (1978), Collective effects asso-nics, J. Atmos. Terr. Phys., 54, 425. ciated with lower hybrid heating of plasma, Plasma Phys., 20, 1131.

Doering, J., and E. Gulcicek (1989), Absolute differential and integral Musher, S., A. Rubenchik, and I. Shapiro (1986), Nonlinear effects in theelectron excitation cross sections for atomic oxygen: The 3p -_

1D and propagation of an ion beam across a magnetic field, Sov. Phys. JETP, 63,3p _ 'S transitions from 4.0 to 30 eV, J. Geophys. Res., 94, 1541. 519.

Doering, J., W. Peterson, C. Bostrom, and J. Armstrong (1975), Measure- Pedersen, T., and H. Carlson (2001), First observations of HF heater-pro-ment of low-energy electrons in the day airglow and day side auroral duced airglow at the High Frequency Active Auroral Research Programzone from Atmosphere Explorer C, J. Geophys. Res., 80, 3934. facility: Thermal excitation and spatial structuring, Radio Sci., 36, 1013.

Fejer, J. (1979), Ionospheric modification and parametric instabilities, Rev. Pedersen, T., M. McCarrick, E, Gerken, C. Selcher, D. Sentman,Geophys., 17, 135. H. Carlson, and A. Gurevich (2003), Magnetic zenith enhancement of

Fox, J., and K. Sung (2001), Solar activity variations of the Venus thermo- HF radio-induced airglow production at HAARP, Geophys. Res. Lett.,sphere/ionosphere, J. Geophys. Res., 106, 21,305. 30(4), 1169, doi:10.1029/2002GL016096.

Galeev, A., R. Sagdeev, V. Shapiro, and V. Shevchenko (1977), Langmuir Pelleiter, G. (1982), Generation of a high-energy electron tail by strongturbulence and dissipation of high-frequency energy, Sov. Phys. JETP, 46, Langmuir turbulence in a plasma, Phys. Rev. Lett., 49, 782.711. Peterson, W., J. Doering, T. Potemra, R. McEntire and C. Bostrom (1977),

Galeev, A., R. Sagdeev, V. Shapiro, and V. Shevehenko (1983), Beam Conjugate photoelectron fluxes observed on Atmosphere Explorer C,plasma discharge and suprathermal electron tails, in Active Experiments Geophys. Res. Lett., 4, 109.in Space (Alpbach, Austria), Eur. Space Agency Spec. Publ., ESA-SP- Peverall, R., et al. (2001), Dissociative recombination and excitation of O:195, p. 151. Cross sections, product yields and implications for studies of ionospheric

Guberman, S. (1988), The production of O('D) from dissociative recombi- airglows, J. Chem.. Phys., 114, 6679.nation of 02', Planet. Space Sci., 36, 47. Rees, M. (1989), Physics and Chemistry of the Upper Atmosphere, Cam-

Guberman, S. (1997), Mechanism for the green glow of the upper atmo- bridge Univ. Press, New York.sphere, Science, 278, 1276. Rietveld, M., M. Kosch, N. Blagoveshchenskaya, V. Komienko, T. Leyser,

Gurevich, A., and G. Milikh (1997), Artificial airglow due to modifications and T. Yeoman (2003), Ionospheric electron heating, optical emissions andof the ionosphere by powerful radio waves, J Geophys. Res., 102, 1389. striations induced by powerful HF radio waves at high latitudes: Aspect

Gurevich, A., Y. Dimant, G. Milikh, and V. Vaskov (1985), Multiple accel- angle dependence, J. Geophys. Res., 108(A4), 1141, doi:10.1029/eration of electrons in the regions high-power radio-wave reflection in the 2002JA009543.ionosphere, J. Atmos. Terr. Phys., 47, 1057. Sagdeev, R., V. Shapiro, and V. Shevchenko (1991), Anomalous dissipation

Gurevich, A., K. Zybin, H. Carlson, and T. Pedersen (2002), Magnetic of strong electromagnetic waves in a laser plasma, in Physics of Laserzenith effect in ionospheric modifications, Phys. Lett. A, 305(5), 264. Plasma, vol. 3, edited by A. Rubenchik and S. Witkowski, p. 271, North-

Gustavsson, B., et al. (2001), First tomographic estimate of volume distri- Holland, New York.bution of HF-pump enhanced airglow emission, J. Geophys. Rex., 106, Shapiro, V., and V. Shevchenko (1984), Strong turbulence of plasma oscil-29,105. lations, in Basic Plasma Physics, edited by A. Galeev and R. Sudan,

Gustavsson, B., B. Brindstr6m, A. Steen, T. Sergienko, T. Leyser, vol. 2, p. 123, North-Holland, New York.M. Rietveld, T. Aso, and M. Erjii (20021, Nearly simultaneous images Shapiro, V., V. Shevchenko, A. Sharma, K. Papadopoulos, R. Sagdeev, andof HF-pump enhanced airglow at 6300 A and 5577 A, Geophys. Res. V. Lebedev (1993), Lower-hybrid turbulence at cometary bow wave andLett., 29(24), 2220, doi:10.1029/2002GL015350. acceleration of cometary protons, J. Geophys. Res.,.98, 1325.

Hedin, A. (1991), Extension of the MSIS thermospheric model into the Sipler, D., E. Enemark, and M. Biondi (1974), 6300-A intensity variationsmiddle and lower atmosphere, J. Geophys. Res., 96, 1159. produced by the Arecibo ionospheric modification experiment, J Geo-

Isham, B., M. Rietveld, T. Hagfors, C. LaHoz, E. Mishin, W. Kofinan, phys. Res., 79, 4276.T. Leyser, and A. van Eyken (1999), Aspect angle dependence of HF Solomon, S., P. Hays, and V. Abreu (1988), The auroral 6300 A emission:enhanced incoherent backscatter, Adv. Space Res., 24(8), 1003. Observation and modeling, J. Geophys. Res., 93, 9867.

Istomin, Y., and T. Leyser (1995), Parametric decay of an electromagnetic Sotnikov, V., V. Shapiro, and V. Shevchenko (1978), Macroscopic effects ofwave near electron cyclotron harmonics, Phys. Plasmas, 2, 2084. collapse at lower hybrid resonanse, Sov. J. Plasma Phys., 4, 250.

Istomin, Y., and T. Leyser (2003), Electron acceleration by cylindrical Vaskov, V. V., A. Gurevich, and A. N. Karashtin (1981), Thermal self-upper hybrid oscillations trapped in density irregularities in the iono- focusing instability of plasma waves near resonance, Geomagn. Aeron.,sphere, Phys. Plasmas, 10, 2962. Engl. Transl., 21, 724.

Kella, D., L. Vejby-Christenson, P. Jonson, H. Pedersen, and L. Andersen Viggiano, A., and S. Williams (2001), Ion-molecule kinetics at high tem-(1997), The source of green line emission determined from a heavy-ion peratures (300- 1800 K): Derivation of internal energy dependencies,storage ring experiment, Science, 276, 1530. Adv. Gas-Phase Ion Chem., 4, 85.

9 of 10

A02305 MISHIN ET AL.: HF-[NDUCED AIRGLOW ONSET AT MAGNETIC ZENITH A02305

Wang, J., D. Newman, and M. Goldman (1997), Vlasov simulations of Zakharov, V., S. Musher, and A. Rubenchik (1976), Weak Langmuir turbu-electron heating by Langmuir turbulence near the critical altitude in the lence of an isothermal plasma, Sov. Phys. JETP, 42, 80.radiation-modified ionosphere, J. Atmos. Sol. Terr Phys., 59, 2461. Zhou, H., J. Huang, and S. Kuo (1994), Cascading of the upper hybrid/

Witasse, 0., J. Lilensten, and C. Lathuil!ire (1999), Modeling of the 01 electron Bernstein wave in ionospheric heating experiments, Phys.630.0 and 557.7 nm thermospheric dayglow during EISCAT-WINDII Plasmas, 1, 3044.coordinated measurements, J. Geophys. Res., 104, 24,639.

Wong, A., G. J. Morales, D. Eggleston, J. Santoru, and R. Behnke (1981),Rapid conversion of electromagnetic waves to electrostatic waves in theionosphere, Phys. Rev. Lett., 47, 1340. W. J. Burke and T. Pedersen, Air Force Research Laboratory, Space

Zakharov, V. (1972), Collapse of Langmuir waves, Sov. Phys. JETP, 35, Vehicles Directorate, Hanscom Air Force Base, MA 01731, USA.908. ([email protected]; [email protected])

Zakharov, V. (1984), Collapse and self-focusing of Langmuir waves, in E. V. Mishin, Boston College, Institute for Scientific Research, 402 St.Basic Plasma Physics, edited by A. Galeev and R. Sudan, vol. 2, p. 81, Clements Hall, 140 Commonwealth Avenue, Chestnut Hill, MA 02467-North-Holland, New York. 3862, USA. ([email protected])

10 of 10

Form Approved

REPORT DOCUMENTATION PAGE OMB No. 0704-0188Public reporting burden for this collection of information is estimated to average 1 hour per response, Including the time for reviewing instructions, searching existing data sources, gathering and maintaining thedata needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducingthis burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of Information if it does not display a currentlyvalid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.

1. REPORT DATE (DD-MM-YYYY)13-06-2005 REPRINT4. TITLE AND SUBTITLE Sa. CONTRACT NUMBEROn the Onset of HF-induced Airglow at HAARP

Sb. GRANT NUMBER

Sc. PROGRAM ELEMENT NUMBER61102F

6. AUTHOR(S) 5d. PROJECT NUMBERE.V. Mishin*, W.J. Burke and T. Pedersen 2311

Se. TASK NUMBER

SD5f. WORK UNIT NUMBERA3

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORTAir Force Research Laboratory/VSBXP NUMBER29 Randolph RoadHanscom AFB MA 01731-3010

AFRL-VS-HA-TR-2005-1061

9. SPONSORING I MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORIMONITOR'S ACRONYM(S)

11. SPONSOR/MONITOR'S REPORT

NUMBER(S)

12, DISTRIBUTION I AVAILABILITY STATEMENTApproved for Public Release; Distribution Unlimited.

*Boston College, institute for Scientific Research, Chestnut Hill, MA

13. SUPPLEMENTARY NOTESREPRINTED FROM: JOURNAL OF GEOPHYSICAL RESEARCH, Vol 109, A02305,doi: I0.1029/2003JA010295, 2004.14. ABSTRACT

[I] Observations of airglow at 630 nm (red line) and 557.7 nm (green line) during theFebruary 2002 campaign at the High Frequency Active Auroral Research Program(HAARP) heating facility are analyzed. We find that durring injections toward magneticzenith (MZ) the green and red lines gain ,-,5 R within I-l s and -20 R within -,10 s,respectively. We term this period the onset of the HF-induced airglow. A model of theonset at magnetic zenith is developed. It accounts for background photoelectrons anddissociative recombination of O2. It is shown that heating and acceleration of backgroundelectrons dominate the airglow onset. We propose a scenario for the generation of strongLangmuir turbulence for injections outside the Spitze region, including magneticzenith.

1S. SUBJECT TERMSWave/particle interactions Topside ionosphereHF modification experiments Artificial airglow onset

16. SECURITY CLASSIFICATION OF: 17. LIMITATION 18. NUMBER 19a. NAME OF RESPONSIBLE PERSONOF ABSTRACT OF PAGES W. 1Burke

a. REPORT c. THIS PAGE SAIR 19b. TELEPHONE NUMBER (include areaUNCLAS UNCLAS UNCLAS code)

781-377-3980Standard Form 298 (Rev. 8-98)Prescribed by ANSI Std. 239.18

new on the onset of hf-induced airglow at...

Documents