Content uploaded by Wayne Scales
Author content
All content in this area was uploaded by Wayne Scales on Aug 12, 2015
Content may be subject to copyright.
Ann. Geophys., 33, 983–990, 2015
www.ann-geophys.net/33/983/2015/
doi:10.5194/angeo-33-983-2015
© Author(s) 2015. CC Attribution 3.0 License.
Stimulated Brillouin scattering during electron
gyro-harmonic heating at EISCAT
H. Y. Fu1, W. A. Scales2, P. A. Bernhardt3, S. J. Briczinski3, M. J. Kosch4,7, A. Senior4, M. T. Rietveld5, T. K. Yeoman6,
and J. M. Ruohoniemi2
1Key Laboratory for Information Science of Electromagnetic Waves (MoE) and School of Information Science and
Technology, Fudan University, Shanghai, China
2Bradley Department of Electrical and Computer Engineering, Virginia Tech, Blacksburg, USA
3Plasma Physics Division, Naval Research Laboratory, Washington, DC, USA
4Department of Physics, Lancaster University, Lancaster, UK
5EISCAT Scientific Association, Ramfjordmoen, Norway
6Department of Physics and Astronomy, University of Leicester, Leicester, UK
7South African National Space Agency, Hermanus, South Africa
Correspondence to: H. Y. Fu (fuhaiyang@gmail.com)
Received: 15 August 2014 – Revised: 16 June 2015 – Accepted: 24 July 2015 – Published: 11 August 2015
Abstract. Observations of secondary radiation, stimulated
electromagnetic emission (SEE), produced during iono-
spheric modification experiments using ground-based, high-
power, high-frequency (HF) radio waves are considered.
The High Frequency Active Auroral Research Program
(HAARP) facility is capable of generating narrowband SEE
in the form of stimulated Brillouin scatter (SBS) and stim-
ulated ion Bernstein scatter (SIBS) in the SEE spectrum.
Such narrowband SEE spectral lines have not been reported
using the European Incoherent Scatter (EISCAT) heater fa-
cility before. This work reports the first EISCAT results of
narrowband SEE spectra and compares them to SEE previ-
ously observed at HAARP during electron gyro-harmonic
heating. An analysis of experimental SEE data shows ob-
servations of emission lines within 100Hz of the pump fre-
quency, interpreted as SBS, during the 2012 July EISCAT
campaign. Experimental results indicate that SBS strength-
ens as the pump frequency approaches the third electron
gyro-harmonic. Also, for different heater antenna beam an-
gles, the CUTLASS radar backscatter induced by HF ra-
dio pumping is suppressed near electron gyro-harmonics,
whereas electron temperature enhancement weakens as mea-
sured by EISCAT/UHF radar. The main features of these new
narrowband EISCAT observations are generally consistent
with previous SBS measurements at HAARP.
Keywords. Ionosphere (active experiments; particle accel-
eration; plasma waves and instabilities)
1 Introduction
Ionospheric plasma turbulence can be created by injection
of powerful high-frequency (HF) radio waves from ground-
based transmitters. The interaction between high-power elec-
tromagnetic waves and plasmas in the ionosphere can pro-
duce stimulated electromagnetic emissions (SEEs), first re-
ported by Thidé et al. (1982) and reviewed by Leyser (2001).
SEE spectral lines in the scattered wave can be utilized to
remotely probe the properties of the ionosphere as well as
actively study radio pump-induced phenomena such as artifi-
cial airglow during modification of the ionosphere (e.g. Bern-
hardt et al., 2009, 2010; Pedersen et al., 2010; Mahmoudian
et al., 2013a).
Wideband SEE within 100kHz of the pump frequency
has been studied extensively for several decades. However,
due to updates of the HAARP facility in 2007, it has been
possible to investigate narrowband (within roughly 1 kHz of
the pump frequency) SEE near the resonance altitude in re-
cent years. Stimulated Brillouin scattering (SBS) has been
recently observed in the high-power, HF wave ionospheric
experiments, shifted by a few tens of hertz from the pump
Published by Copernicus Publications on behalf of the European Geosciences Union.
984 H. Y. Fu et al.: HF-pumped, stimulated Brillouin scatter at EISCAT
frequency (Norin et al., 2009). During the past few years,
SEE observations at HAARP have revealed a plethora of nar-
rowband SEE lines associated with SBS and stimulated ion
Bernstein scatter (SIBS) (e.g. Norin et al., 2009; Bernhardt
et al., 2009, 2010; Samimi et al., 2012, 2013, 2014). Fu et
al. (2013) further investigated SBS and SIBS and their in-
terrelationship in detail over a range of aspect beam angles
and frequency stepping near electron gyro-harmonics. Mah-
moudian et al. (2013b) investigated the threshold for SBS
at HAARP and experimentally showed that an effective ra-
diated power (ERP) ∼140 MW is required to excite SBS
associated with ion acoustic (IA) waves. These observations
motivated this campaign at EISCAT to produce SBS using
lower-power HF heating.
The physical process of SBS involves a nonlinear interac-
tion in which an incident (pump) electromagnetic wave de-
cays into an electrostatic IA wave and a scattered electro-
magnetic wave via the Brillouin instability. In such three-
wave interaction processes, the wave-matching conditions
are satisfied: ω0=ωS+ωLand k0=kS+kL, where ωis
the wave frequency; kis the wave propagation vector; and
the subscripts 0, Sand Ldenote the pump waves, the scat-
tered waves and low-frequency waves, respectively. Such
laser-induced parametric decay SBS processes have been
commonly detected and thoroughly studied in unmagnetized
plasmas as summarized by Kruer (1988). The first SBS de-
cay process has been detected only recently in high-power,
HF ionospheric modification experiments.
Observations of SBS at HAARP were considered to arise
from the plasma reflection resonance height where ω0≈ωp
and the upper hybrid UH resonance height ω0≈ωuh. How-
ever, for underdense plasmas, ω0ωp, it has been sug-
gested that SBS may be produced using the EISCAT inco-
herent radar facility as first discussed by Dysthe et al. (1977).
A modification in the double-humped spectra of incoherent
backscatter was predicted by Fejer (1977) at Jicamarca and
Arecibo. Experimental observation of SBS using the Jica-
marca 50 MHz incoherent scatter radar can cause asymmetry
as large as 25% in the incoherent ionic backscatter spectrum
(Fejer et al., 1978), resulting in errors of 10–15 m s−1in the
measured velocity.
The primary purpose of the experiment during the 2012
EISCAT campaign was to investigate the possibility of gen-
erating SBS using the EISCAT HF heating facility and
its modification effects using simultaneous incoherent EIS-
CAT/UHF radar and CUTLASS HF radar diagnostics. It
was also the intention to study the correlation between
SEE, field-aligned irregularities (FAIs) and electron tempera-
ture enhancement near the electron gyro-harmonic frequency
for different aspect angles during the ionospheric modifica-
tion experiment. This paper is organized as follows. In the
next section, experimental procedure and diagnostics are de-
scribed. Thereafter, the experimental observations and anal-
ysis are given. Finally, a summary and conclusions are pro-
vided.
2 Experiment setup
The EISCAT HF facility (69.59◦N, 19.23◦E) near Tromsø,
in northern Norway (Rietveld et al., 1993), was used to pro-
duce SEE during a campaign on 3–10 July 2012. The HF
transmitter was operated at O-mode polarization with full
power. The pump frequency was stepped upward and down-
ward through the third harmonic of the ionospheric elec-
tron gyro-frequency 3fce. The pump frequency steps every
20 kHz in a range of 3.9 MHz ≤f0≤4.2 MHz. The heater
duty cycle was typically 1min on and 1min off, unless oth-
erwise stated. All 12 transmitters on array 2 were used at
80 kW each, resulting in a gain of 22.4dBi and effective radi-
ated power (ERP) of approximately 148MW. The beam an-
gle was scanned in small 6◦steps in the magnetic meridian.
For each angle, the heating time period was 30min during
the frequency stepping cycle.
The SEE receiver was installed near Breivikeidet, Norway
(69.64◦N, 19.49◦E), about 13 km east-northeast of the EIS-
CAT site. The antenna was a broadband resistively loaded
folded dipole. The receiver was an Ettus Research USRP
N210 fitted with a GPS-disciplined oscillator to provide pre-
cise time and frequency references. The receiver was tuned
to 6 MHz and recorded at 6.25 MHz sample rate to cover all
heater frequencies. The sampled data are processed with the
fast Fourier transform (FFT) to yield low-frequency spectra.
The EISCAT 931MHz UHF radar was operated in Beata
mode, which enables measurement from 50 to 700 km with a
minimum of 5 s time resolution and 3.5 km range resolution.
The UHF radar data were integrated for 60s to reduce mea-
surement error, with approximately 14km resolution near
the reflection region. The UHF incoherent scatter radar can
provide ionospheric electron temperature, electron density,
plasma/ion-line frequency spectra and the reflection altitude.
A dynasonde, co-located with EISCAT, made a sounding ev-
ery 6 min, which can also provide the electron density profile
and the reflection altitude in the ionosphere.
The electron gyro-harmonic effects of SEE are associated
with HF, pump-induced FAIs, consisting of narrow filaments
of density depletions a few percent in amplitude and elon-
gated several tens of kilometres along the geomagnetic field
(Fialer, 1974). The heater produced FAIs were diagnosed by
CUTLASS HF coherent radars at Hankasalmi, Finland, and
Thykkvibær, Iceland. The CUTLASS pair of HF radars in
the Northern Hemisphere is part of the SuperDARN network
of HF coherent radars,which is a frequency-agile bistatic HF
radar system operating in the range 8–20 MHz (Robinson et
al, 1997). During this experiment, the CUTLASS radars op-
erated in “stereo” mode by utilizing some of the radar’s spare
duty cycle. The radar employed three frequency bands – 9–
10, 13–14 and 16–17 MHz – which are sensitive to FAIs with
spatial sizes of between 8 and 17 m. The dwell (integration)
time on each radar beam is 1 s for Hankasalmi.
Ann. Geophys., 33, 983–990, 2015 www.ann-geophys.net/33/983/2015/
H. Y. Fu et al.: HF-pumped, stimulated Brillouin scatter at EISCAT 985
−100−80−60−40−20 0 20 40 60 80 100
−75
−50
−25
0
Frequency Shift (Hz)
Relative Power (dB)
−8 Hz 12 Hz
19: 31:00 f0=3.98 MHz
−100−80−60−40−20 0 20 40 60 80 100
−75
−50
−25
0
Frequency Shift (Hz)
Relative Power (dB)
19: 29: 00 f0=4.00 MHz
− 8Hz 12Hz
−100−80−60−40−20 0 20 40 60 80 100
−75
−50
−25
0
Frequency Shift (Hz)
Relative Power (dB)
−8 Hz
19: 27: 00 f0=4.02 MHz
12 Hz
−100−80−60−40−20 0 20 40 60 80 100
−75
−50
−25
0
Frequency Shift (Hz)
Relative Power (dB)
12 Hz
−8 Hz
19: 25: 00 f0=4.04 MHz
−100−80−60−40−20 0 20 40 60 80 100
−75
−50
−25
0
Frequency Shift (Hz)
Relative Power (dB)
12 Hz
−8 Hz
19: 23: 00 f0=4.06 MHz
−100−80−60−40−20 0 20 40 60 80 100
−75
−50
−25
0
Relative Power (dB)
Frequency Shift (Hz)
12 Hz
−8 Hz
19: 21: 00 f0=4.08MHz
Figure 1. Narrowband SEE frequency spectra of HF scattered sig-
nals showing strong emission lines at 8–12Hz using the EISCAT
HF transmitter operating at varying pump frequencies near 3fce
during 19:20–19:32 UT on 3 July 2012. The heating beam points
towards the magnetic zenith direction with 1min on/off duty cycle.
3 Experimental results
3.1 Narrowband SEE observations
Figure 1 shows narrowband frequency spectra of the scat-
tered HF pump wave for heating near the third electron gyro-
harmonic frequency, 3fce, during 19:20–19:32 UT on 3 July
2012. Electromagnetic backscattered waves were produced
by the EISCAT HF transmitter operating with an ERP of
148 MW for the magnetic zenith beam. Dynasonde data at
19:28 UT on 3 July 2012 indicate quiet ionospheric status.
The reflection altitude for the pump frequency 4.04MHz
is ∼215 km according to dynasonde data. Strong emissions
downshifted by ∼8 Hz and upshifted by ∼12Hz in the spec-
tra are clearly observed in Fig. 1 with power within 10dB
relative to the reflected pump wave. The power of the down-
shifted (or Stokes) emission line is larger than the upshifted
(anti-Stokes) emission.
These shifted spectral lines observed in Fig. 1 show a sim-
ilar frequency shift and relative amplitude of Stokes and anti-
Stokes lines when compared to experimental observations at
HAARP (e.g. Norin et al., 2009; Bernhardt et al., 2009). The
reflected pump waves and scattered electromagnetic waves
combine to produce upshifted SBS lines with lower intensity
and slightly higher 4–5 Hz frequency offset than the down-
0 500 1000
100
150
200
250
300
350
400
∆ Te (K)
0 2 4 6
100
150
200
250
300
350
400
Frequency (MHz)
Altitude (km)
20−21
22−23
24−25
26−27
28−29
30−31
Figure 2. The measured UHF radar plasma frequency and electron
temperature enhancement 1Teduring 19:24–19:30 UT on 3 July
2012. Note that the reflection altitude is approximately 215km for
the pump frequency 4.04MHz.
shifted SBS. Upshifted SBS lines have been previously ex-
plained as follows (Bernhardt et al., 2009). After the up-
ward pump reflects near-zero refractive index, yielding a
downward pump wave, it scatters with IA waves to pro-
duce another upward electromagnetic wave with wave vec-
tor k0=kS+kLand ω0=ωS+ωL. Bernhardt et al. (2010)
interpreted 1fIA = −(f −f0)=6∼12 Hz below/above the
pump frequency f0as SBS from the plasma resonance re-
gion. Theoretical and experimental works on SBS indicate
that the production of downshifted lines should be preferred.
The strength of observed sideband emissions in Fig. 1 de-
pends on the electromagnetic pump wave frequency as well.
During the frequency stepping, the SBS emissions were ob-
served at pump frequencies 4.04, 4.02 and 4.00MHz, rela-
tively close to 3fce. The frequency dependence of the emis-
sion may be attributed to the EISCAT HF transmitter power
being near the threshold for excitation. It may be postulated
that, when less anomalous absorption occurs near 3fce, more
heater power can be transmitted to a higher resonance al-
titude where SBS occurs. Anomalous absorption is due to
scattering of the electromagnetic waves on FAI with a wide
spatial spectrum. For pump frequency near nfce(n=3,4),
FAI intensity and anomalous absorption are minimum (see
Leyser, 2001, and references therein). This will be discussed
further in the next section.
Figure 2 shows the measured UHF radar plasma frequency
ωpand electron temperature enhancement 1Teprofile vs.
height during 19:20–19:32 UT on 3 July 2012. The integra-
tion time is 60 s. Incoherent radar data indicate that the re-
flection altitude is 215 km for the pump frequency 4.04MHz.
The electron temperature enhancement at ∼215 km min-
www.ann-geophys.net/33/983/2015/ Ann. Geophys., 33, 983–990, 2015
986 H. Y. Fu et al.: HF-pumped, stimulated Brillouin scatter at EISCAT
−10 −8 −6 −4 −2 0 2 4 6 8 10
−75
−50
−25
019:31:00 f0=3.98 MHz
Frequency Shift (kHz)
Relative Power (dB)
−10 −8 −6 −4 −2 0 2 4 6 8 10
−75
−50
−25
019:29:00 f0=4.00 MHz
Frequency Shift (kHz)
Relative Power (dB)
DP
−10 −8 −6 −4 −2 0 2 4 6 8 10
−75
−50
−25
019:27:00 f0=4.02 MHz
Frequency Shift (kHz)
Relative Power (dB)
DP
−10 −8 −6 −4 −2 0 2 4 6 8 10
−75
−50
−25
019:25:00 f0=4.04 MHz
Frequency Shift (kHz)
Relative Power (dB)
DP
−10 −8 −6 −4 −2 0 2 4 6 8 10
−75
−50
−25
019:23:00 f0=4.06 MHz
Frequency Shift (kHz)
Relative Power (dB)
DM UM
DP
−10 −8 −6 −4 −2 0 2 4 6 8 10
−75
−50
−25
019:21:00 f0=4.08 MHz
Frequency Shift (kHz)
Relative Power (dB)
DM
UM
DP
Figure 3. Wideband SEE frequency spectra of HF scattered sig-
nals from the EISCAT HF transmitter operating at varying pump
frequencies near 3fce. The heating direction is along the magnetic
zenith. According to DP measurement, f0≈3fceoccurs between
4.04 and 4.06 MHz.
imizes for pump frequency 4.04MHz during the interval
19:24–19:25 UT on 3 July 2012. The electron temperature
enhancement and anomalous absorption are correlated as ob-
served by Honary et al. (1995). Electron temperature and
anomalous absorption are minimized when pumping on a
gyro-harmonic frequency because the growth of small-scale
(1–10 m) field-aligned striations is suppressed, as will be dis-
cussed in further detail in the next section. Based on incoher-
ent scatter radar data, the pump-induced electron tempera-
ture enhancement reaches approximately 500–600 K for the
pump frequency 3.98MHz in the heated region in Fig. 2.
The wave-matching condition for SBS is kL∼
=2k0
(Bernhardt et al., 2010). An analytical expression for
the IA waves, propagating with an angle θto the
ambient magnetic field, can be expressed as ωIA =
q(k2
IAc2
IAcos2θ )/(1+k2
IAc2
IA/2
ci)when the ion sound
waves have wavelengths much larger than a Debye length
kIAλd1. Here, ciis the ion gyro-frequency and cIA =
√(γeTe+γiTi)/miis the IA velocity with γe=1 and γi=
3, Teand Tiare the electron and ion temperature, respec-
tively, and λdis the Debye length (Bernhardt et al., 2009).
The ion gyro-frequency representative of the conditions over
EISCAT at 215km is estimated to be fci≈46.0 Hz. Based
on incoherent scatter radar data, the electron temperature
is taken to be Te=2600Kand Te/Ti=2.5. The wave-
matching condition predicts that the strongest IA wave emis-
sions f1∼8 Hz are excited near the reflection resonance alti-
tude where the local plasma frequency becomes close to the
pump frequency.
3.2 Associated wideband SEE and irregularities
A classic feature of the steady-state SEE spectrum is the
downshifted peak (DP) when pumping near electron gyro-
harmonics (Leyser, 2001). The DP is located at 1fDP ≈1–
3 kHz below the pump frequency. An upshifted peak (UP)
feature occurs above the pump frequency at approximately
the mirror frequency of the DP. The DP, 2DP and UP can
be simultaneously observed when the pump frequency ap-
proaches the nth (n≥3) electron gyro-harmonic frequency
nfce. Stubbe and Kopka (1990) stated that the DP has been
found to be a strong feature for f0=3fceand weak sign for
f0=4fce,5fce. It is worth mentioning that Mahmoudian et
al. (2013a) recently noted a similar spectral feature for f0≈
2fcepumping with frequency offset 1f ≈500 −1000 Hz.
Figure 3 shows the dependence of wideband SEE fea-
tures on the pump frequency when pumping near 3fcefor
the same time period and experimental conditions as the
narrowband SEE in Fig. 1. For pump frequencies close to
4.04 MHz, the DP at approximately ∼2kHz below the pump
frequency develops. The DP frequency offset drops from ap-
proximately −2.5 to −1.6 kHz as the pump frequency ap-
proaches 3fce, consistent with previous experimental obser-
vations (Stubbe et al., 1994). The DP serves as a good in-
dicator for the pump frequency close to the third electron
gyro-harmonic frequency (Stubbe and Kopka, 1990; Stubbe
et al., 1994; Honary et al., 1995). The DP frequency off-
set for the pump frequency near 3fcecan be approximately
estimated based on existing theoretical models (Huang and
Kuo, 1995; Hussein and Scales, 1997; Mahmoudian et al.,
2013a). A detailed description of these DP models is beyond
the scope of the current paper and will be pursued in future
works. If the pump frequency increases further above elec-
tron gyro-harmonic, the downshifted maximum (DM) spec-
tral line (Leyser, 2001) at approximately 8–8.5kHz below
the pump frequency appears in the lower sideband spec-
trum. The DM involves electrostatic lower hybrid waves,
where the lower hybrid wave frequency is estimated to be
ωlh '7.5 kHz. The presence of a DM and upshifted maxi-
mum (UM) in the SEE spectrum, which is closely correlated
with FAIs, can also serve as a indicator of whether the pump
frequency is near to or far from a harmonic of the electron
gyro-frequency (Leyser et al., 1994). Another set of repeated
daytime experiments shows the dependence of the DP on the
pump frequency for different beam angles. The behaviour at
different angles is similar to the magnetic zenith case. If the
transmitter beam angle is tilted further off the magnetic field
line, the amplitude of the DP becomes weak and the second
downshifted peak (2DP) and UP may not appear in the spec-
tra.
Ann. Geophys., 33, 983–990, 2015 www.ann-geophys.net/33/983/2015/
H. Y. Fu et al.: HF-pumped, stimulated Brillouin scatter at EISCAT 987
18:40 18:50 19:00 19:10 19:20 19:30
3.9
4
4.1
4.2
Frequency(MHz)
Universal Time
Altitude (km)
18:40 18:45 18:50 18:55 19:00 19:05 19:10 19:15 19:20 19:25 19:30 19:35
100
150
200
250
300
Electron Temperature (K)
0
500
1000
1500
2000
2500
3000
3500
4000
Figure 4. The results of HF pumping during 18:40–19:40UT on 3 July 2012. The upper panel shows the pump frequency, where the
black arrows indicate the pump frequency 4.04MHz (approximately 3fce) at 18:52–18:53 and 19:14–19:15UT. The middle panel shows
the electron temperature profile measured by the EISCAT UHF radar with the integration time 60s. The lower panel shows CUTLASS
backscatter power, Doppler velocity, and spectral width vs. slant range (line-of-sight distance) in the heating region over EISCAT.
Figure 4 depicts the HF pumping frequency scheme, the
electron temperature profile measured by the EISCAT UHF
radar and CUTLASS backscatter power, the Doppler ve-
locity, and the spectral width during 18:40–19:38UT on
3 July 2012. The upper panel shows the pump frequen-
cies between 3.92 and 4.2MHz. The arrows indicate the
pump frequency near 3fce(i.e. f0=4.04 MHz). The mid-
dle panel shows the electron temperature measured by the
EISCAT 931MHz UHF incoherent scatter radar. For the up-
ward frequency stepping, the electron temperature enhance-
ment minimizes during 18:54–18:55(f0=4.06 MHz) and
18:56–18:57 (f0=4.08MHz). For the downward frequency
www.ann-geophys.net/33/983/2015/ Ann. Geophys., 33, 983–990, 2015
988 H. Y. Fu et al.: HF-pumped, stimulated Brillouin scatter at EISCAT
stepping, the electron temperature enhancement reduces dur-
ing 19:22–19:23 (f0=4.06 MHz) and 19:24–19:25 (f0=
4.04 MHz), which correspond to strong DP as observed in
Fig. 3. The electron temperature enhancement minimizes
when the pump frequency approaches 3fce. This agrees with
previous experimental observations (Honary et al., 1995).
The electron temperature enhancement exhibits an asym-
metry for pump frequencies above and below 3fcein the mid-
dle panel of Fig. 4. It should be noted that the measurement
error increases during 19:08–19:15 for f0>3fcein the nar-
row altitude range close to the heater reflection height, since
the electron temperature retrieval algorithm is based on the
ion-line spectra which are modified by HF pump-induced ef-
fects. It is unclear whether there is more efficient electron
heating for f0>3fcefrom these observations. Further anal-
ysis will be required on the simultaneous ion-line spectra and
SEE measurements.
The lower panel shows the CUTLASS backscatter power,
Doppler velocity, and spectral width from beam 5 of the Han-
kasalmi radar. The backscattered signals are produced by
Bragg scattering of the sounding waves from pre-existing or
pump-induced field-aligned striations. The aspect-angle de-
pendence for scattering requires that the radio wave kvector
be close to orthogonal to the magnetic field B. The CUT-
LASS radar measures F-region irregularities with a plasma
E×Bdrift vd. The HF radar beam that is pointed to mag-
netic north measures an eastward (zonal) electric field or-
thogonal to B. The radar cycles through three frequencies:
9.9, 13.2 and 16.6MHz. Only 13.2 MHz corresponds to
strong backscatter from FAIs of wavelength ∼11 m. When
the pump frequency approaches 4.04MHz during the time
period 18:52–18:53, there is a clear reduction observed in
the backscatter power from a peak of 30dB to approxi-
mately 15dB. There exists a minimum of the backscatter
power during the time periods 18:54–18:55(f0=4.06 MHz)
and 18:56–18:57(f0=4.08 MHz). The minimum backscat-
ter power exactly corresponds to the minimum electron tem-
perature enhancement. Based on the temporal evolution of
the backscatter power between approximately 0 and 25dB,
the rise time of FAIs is estimated to be less than 10s and the
decay time is approximately 40 s.
The Doppler velocity during the HF pumping is due to the
E×Bdrift of HF-induced FAIs and reaches a maximum value
of approximately −50 m s−1corresponding to a frequency of
approximately 5 Hz. The spectral width is typically less than
5 m s−1.
Unfortunately, after 19:00UT in Fig. 4, the CUTLASS
radar was switched back to operate in its standard mode.
When pumping above 3fceafter 19:00, the spectral width
may vary but is not observed in these measurements. In sum-
mary, these observations indicate that when the electron tem-
perature is reduced and FAIs become weak, bothSBS and DP
are observed in the scattered signals.
These DP line observations as well as electron temperature
and FAIs hold true for different heater beam angles. The FAIs
are suppressed when pumping very close to 3fce, resulting in
weak CUTLASS backscatter (Honary et al., 1999). While the
electron temperature is minimum in correlation with field-
aligned striation suppression, a prominent DP with a small
frequency shift is observed in the spectrum. When pumping
near electron gyro-harmonics, less absorption occurs near
the upper hybrid resonance level, giving rise to the simul-
taneous presence of a strong DP emission line (Huang and
Kuo, 1995). Huang and Kuo (1995) proposed a generation
mechanism for the DP and UP emissions through paramet-
ric decay of upper hybrid/electron Bernstein (UH/EB) wave
into another UH/EB sideband wave and a nearly perpendic-
ularly propagating IA decay mode wave in an altitude region
slightly above the double resonance layer. Such a DP gener-
ation mechanism process involves short-scale, field-aligned
density irregularities (k=k0) through a thermal oscillat-
ing two streaming instability (OTSI) process (Dysthe et al.,
1983). This differs from SBS, which does not involve field-
aligned density irregularities in its generation process. The
threshold for SBS is usually higher than DP emissions, and
DP appears in the spectra almost immediately after the heater
is turned on (Mahmoudian et al., 2013a).
4 Discussion and conclusions
Using an ERP of ∼148MW, the EISCAT HF facility may
generate SBS emissions. It is noted that this observed power
level is less than that required for SBS generation from the
plasma reflection altitude at HAARP (∼320 MW) (Mah-
moudian et al., 2013b). During this campaign, the IA-related
emission lines shifted by 6–12Hz from the pump are ob-
served for the pump frequency near the third electron gyro-
harmonic. Also, the amplitude of the downshifted ∼8 Hz IA
line is larger than the upshifted ∼12Hz IA line. These spec-
tral characteristics of IA emission lines reported in this paper
agree with SBS lines from the plasma resonance region pre-
viously observed at HAARP. As for the critical differences,
more carefully designed experiments are necessary in the fu-
ture to make substantive conclusions at this time.
To further investigate SBS generation near the third gyro-
harmonic, the DP lines are observed simultaneously with
electron temperature from EISCAT/UHF data and FAIs from
CUTLASS radar. During the frequency stepping, experimen-
tal results show that DP structures become prominent as
f0approaches 3fce, while FAIs are suppressed and elec-
tron temperature becomes minimum. It is postulated that the
enhancement of SBS near 3fcemay be explained by weak
FAIs, with the result that more power reaches the reflec-
tion altitude. The correlations between DP structures, plasma
line/ion-line spectrum and FAIsfor varying transmitter beam
angle have been observed during the experiments and show
qualitatively similar behaviour to pumping along the mag-
netic field but are not presented here. Further examination of
the data will be provided in the future.
Ann. Geophys., 33, 983–990, 2015 www.ann-geophys.net/33/983/2015/
H. Y. Fu et al.: HF-pumped, stimulated Brillouin scatter at EISCAT 989
Although fundamentally different physical processes, SBS
and DP SEE are both a result of parametric decay instabili-
ties with IA waves as the low-frequency decay modes. SBS
involves slow magnetosonic IA waves for frequencies below
the ion cyclotron frequency, while DP involves IA waves for
frequencies above the ion cyclotron frequency. These spec-
tral lines are therefore important consequences of IA waves
in the wideband and narrowband SEE spectrum leading to
additional diagnostic information of ionospheric conditions.
The characteristics of the two processes are compared from
the present experimental observations as follows:
1. For pump frequency stepping across electron gyro-
harmonics, the DP is strengthened as f0approaches
3fce. SBS is also observed to be enhanced as f0ap-
proaches 3fce. It is postulated that less absorption and
consequently more power near 3fceplays an important
role in exciting SBS emissions.
2. The frequency offset of SBS at 8–12 Hz appears roughly
independent of f0where the DP is highly sensitive
to f0with 1f varying with proximity of f0to 3fce.
When the pump frequency is increased towards 3fce,
the reduced frequency offset is explained by the wave-
matching condition of the upper hybrid/electron Bern-
stein (UH/EB) parametric decay process (Huang and
Kuo, 1995).
3. The SBS from the plasma resonance altitude is consid-
ered to be a process which does not depend on the pres-
ence of field-aligned striations. This is different from
models of the DP emission generation, which involves
the existence of FAIs (Huang and Kuo, 1995). The DP
may require a lower power threshold field than that for
SBS as observed in experiments. Previous experiments
have observed DP emissions at 4.04MHz with an ERP
of 86 MW (Stubbe et al., 1984) at EISCAT. According
to calculations by Huang and Kuo (1995), the heater
nominal power threshold is an ERP of ∼12MW by ig-
noring D-region absorption, a factor of 10 less than the
threshold for SBS as a rough estimation. The estimated
power level for the DP seems approximately on the or-
der of the power threshold for FAIs observed by Wright
et al. (2006).
4. The dependence of the SBS and DP lines on aspect an-
gle of the transmitter beam relative to the magnetic field
θ0is different. The SBS occurs where the parallel elec-
tric field undergoes swelling at the plasma resonance
altitude. The DP requires a large electric field compo-
nent perpendicular to the geomagnetic field in the UH
region. When increasing θ0for the pump wave, the par-
allel electric field component becomes smaller and the
electric field turns from parallel to the geomagnetic field
towards horizontal at a larger distance below the reflec-
tion height (Leyser, 1991). For tilting beam angles off
the magnetic field, the excitation of SBS from the reflec-
tion region becomes relatively less important compared
to SBS from the upper hybrid level (Fu et al., 2013).
For varying beam angle experiments, strong DP emis-
sions were essentially observed for the magnetic zenith
beam (Tereshchenko et al., 2006).
5. Both SBS and DP are associated with IA waves that
depend on electron temperature. As electron tempera-
ture, Te, is increased during heating, the frequency off-
sets of SBS and DP are predicted to increase. If f0is
sufficiently far from 3fce, the electron temperature may
be derived based on IA SBS emission from the upper
hybrid resonance level (Bernhardt et al., 2009). Elec-
tron temperature retrieval from SBS and DP lines is a
potentially powerful diagnostic capability; however im-
portant aspects of the theory are still lacking and more
work is required at this time.
Finally, it should be pointed out that both EISCAT (2≈
12◦) and HAARP (2≈14◦) HF heating facilities are located
at high latitudes with a comparable geomagnetic angle. The
EISCAT HF heater has approximately one-third the power
of the HARRP HF heater and only higher gyro-harmonic
(n≥3) heating capability is available at EISCAT. However,
with the unique advantage of the EISCAT/UHF radars and
CUTLASS radars, new SEE phenomena recently observed
at HAARP may be investigated in further detail at EISCAT
as well.
Acknowledgements. The authors would like to acknowledge Inge-
mar Häggström for access to EISCAT UHF radar data analysis and
the staff at the EISCAT facility for technical support. The work at
Virginia Tech was supported in part by the National Science Foun-
dation. The work at the Naval Research Laboratory was sponsored
by the NRL 6.1 Base programme. The work at Fudan University
was supported in part by National Science Foundation of China
(NSFC no. 41404122). The authors highly appreciate the construc-
tive comments from the referees.
The topical editor K. Hosokawa thanks S. Grach and T. Leyser
for help in evaluating this paper.
References
Bernhardt, P. A., Selcher, C. A., Lehmberg, R. H., Rodriguez,
S., Thomason, J., McCarrick, M., and Frazer, G.: Determina-
tion of the electron temperature in the modified ionosphere
over HAARP using the HF pumped Stimulated Brillouin Scat-
ter (SBS) emission lines, Ann. Geophys., 27, 4409–4427,
doi:10.5194/angeo-27-4409-2009, 2009.
Bernhardt, P. A., Selcher, C. A., Lehmberg, R. H. , Rodriguez, S. P.,
Thomason, J. F. , Groves, K. M. , McCarrick, M. J., and Frazer,
G. J.: Stimulated Brillouin Scatter in a magnetized ionospheric
plasma, Phys. Rev. Lett., 104, 165004, 2010.
Dysthe, K. B., Leer, E., Trulsen, J., and Stenflo, L.: Stimulated Bril-
louin scattering in the ionosphere, J. Geophys. Res., 82, 717–718,
1977.
www.ann-geophys.net/33/983/2015/ Ann. Geophys., 33, 983–990, 2015
990 H. Y. Fu et al.: HF-pumped, stimulated Brillouin scatter at EISCAT
Dysthe, K. B., Mjølhus, E., Pecsel, H. L., and Rypdal, K.: A ther-
mal oscillating two-stream instability, Phys. Fluids, 26, 146–157,
1983.
Fejer, J. A.: Stimulated Brillouin scattering and incoherent
backscatter, J. Geophys. Res., 4, 289–290, 1977.
Fejer, J. A., Rinnert K., and Woodman R.: Detection of stimulated
Brillouin scattering by the Jicamarca radar, J. Geophys. Res., 83,
2133–2136, 1978.
Fialer, P. A.: Field-aligned scattering from a heated region of the
ionosphere – Observations at HF and VHF, Radio Sci., 9, 923–
940, 1974.
Fu, H., Scales, W. A., Bernhardt, P. A., Samimi, A., Mahmoudian,
A., Briczinski, S. J., and McCarrick, M. J. : Stimulated Brillouin
scatter and stimulated ion Bernstein scatter during electron gyro-
harmonic heating experiments, Radio Sci., 48, 607–616, 2013.
Honary, F., Stocker, A. J., Robinson, T. R., Jones, T. B., and Stubbe,
P.: Ionospheric plasma response to HF radio waves operating at
frequencies close to the third harmonic of the electron gyrofre-
quency, J. Geophys. Res., 100, 21489–21501, 1995.
Honary, F., Robinson, T. R., Wright, D. M., Stocker, A. J., Ri-
etveld, M. T., and McCrea, I.: Letter to the Editor: First direct
observations of the reduced striations at pump frequencies close
to the electron gyroharmonics, Ann. Geophys., 17, 1235–1238,
doi:10.1007/s00585-999-1235-6, 1999.
Huang, J. and Kuo, S. P.: A generation mechanism for the down-
shifted peak in stimulated electromagnetic emission spectrum, J.
Geophys. Res., 100, 21433–21438, 1995.
Hussein, A. A. and Scales, W. A.: Simulation studies of paramet-
ric decay processes assosiated with ionospheric stimulated radi-
ation, Radio Sci., 32, 2099–2107, 1997.
Kosch, M. J., Rietveld, M. T., Kavanagh, A. J., Davis, C., Yeoman,
T. K., Honary, F., and Hagfors, T.: High-latitude pump-induced
optical emissions for frequencies close to the third electron gyro-
harmonic, Geophys. Res. Lett., 29, 27-1–27-4, 2002.
Kruer, W. L.: The Physics of Laser Plasma Interactions, Addison-
Wesley, New York, USA, 1988.
Leyser, T. B.: Parametric interaction between upper hybrid and
lower hybrid waves in heating experiments, Geophys. Res. Lett.,
18, 408–411, 1991.
Leyser, T. B., Thidé, B., Kopka, H., Waldenvik, M., Veszelei, E,
Frolov, V. L., Grach, A. M., and Komrakov, G. P.: Downshifted
maximum features in stimulated electromagnetic emission spec-
tra, J. Geophys. Res., 99, 19555–19568, 1994.
Leyser, T. B.: Stimulated electromagnetic emissions by high-
frequency electromagnetic pumping of the ionospheric plasma,
Space Sci. Rev., 98, 223–328, 2001.
Mahmoudian, A., Scales, W. A., Bernhardt, P. A., Samimi, A.,
Kendall, E., Ruohoniemi, J. M., Isham, B., Vega-Cancel, O.,
and Bordikar, M.: Ion gyro-harmonic structuring in the stim-
ulated radiation spectrum and optical emissions during elec-
tron gyro-harmonic heating, J. Geophys. Res., 118, 1270–1287,
doi:10.1002/jgra.50167, 2013a.
Mahmoudian, A., Scales, W. A., Bernhardt, P. A., Fu, H., Briczin-
ski, S. J., and McCarrick, M. J. : Investigation of ionospheric
stimulated Brillouin scatter generated at pump frequencies near
electron gyroharmonics, Radio Sci., 48, 685–697, 2013b.
Norin, L., Leyser, T. B., Nordblad, E., Thidé, B. and Mc-
Carrick, M.: Unprecedentedly strong and narrow electro-
magnetic emissions stimulated by high-frequency radio
waves in the ionosphere, Phys. Rev. Lett., 102, 065003,
doi:10.1103/PhysRevLett.102.065003, 2009.
Pedersen, T., Gustavsson, B., Mishin, E., Kendall, E., Mills, T.,
Carlson, H. C., and Snyder, A. L.: Creation of artificial iono-
spheric layers using high-power HF waves, Geophys. Res. Lett.,
37, L02106, doi:10.1029/2009GL041895, 2010.
Rietveld, M. T., Kohl, H., Kopka, H., and Stubbe, P.: Introduction
to iono-spheric heating experiments at Tromsø, 1, Experimental
overview, J. Atmos. Terr. Phys., 55, 577–599, 1993.
Robinson, T. R., Stocker, A. J., Bond, G. E., Eglitis, P., Wright, D.
M., and Jones, T. B.: O- and X-mode heating effects observed
simultaneously with the CUTLASS and EISCAT radars and low
power HF diagnostics at Tromsø, Ann. Geophys., 15, 134–136,
1997,
http://www.ann-geophys.net/15/134/1997/.
Samimi, A., Scales, W. A., Bernhardt, P. A., Briczinski, S. J.,
Selcher, C. A., and McCarrick, M. J.: On ion gyro-harmonic
structuring in the stimulated electromagnetic emission spectrum
during second electron gyro-harmonic heating, Ann. Geophys.,
30, 1587–1594, doi:10.5194/angeo-30-1587-2012, 2012.
Samimi, A., Scales, W. A., Fu, H., Bernhardt, P. A. , Briczinski, S.
J., and McCarrick, M. J.: Ion gyroharmonic structures in stimu-
lated radiation during second electron gyroharmonic heating: 1.
Theory, J. Geophys. Res., 118, 502–514, 2013.
Samimi, A., Scales, W. A., Bernhardt, P. A., Briczinski,
S. J., and McCarrick, M. J.: Ion gyroharmonic struc-
tures in stimulated radiation during second electron gyrohar-
monic heating: 1. Simulations, J. Geophys. Res., 119, 502–
514, doi:10.1002/2013JA019341, 2014.
Stubbe, P. and Kopka, H.: Stimulated electromagnetic emission in
a magnetized plasma: A new symmetric spectral feature, Phys.
Rev. Lett., 65, 183–186, 1990.
Stubbe, P., Stocker, A. J., Honary, F., Robinson, T. R., and Jones,
T. B.: Stimulated electromagnetic emissions and anamalous HF
wave absorption near electron gyroharmonics, J. Geophys. Res.,
99, 6233–6246, 1994.
Stubbe, P., Kopka, H., Thidé, B., and Derblom, H.: Stimulated elec-
tromagnetic emission: A new technique to study the parametric
decay instability in the ionosphere, J. Geophys. Res.-Atmos., 89,
7523–7536, 1984.
Tereshchenko, E. D., Yurik, R. Yu., Khudukon, B. Z., Rietveld,
M. T., Isham, B., Belyey, V., Brekke, A., Hagfors, T., and
Grill, M.: Directional features of the downshifted peak observed
in HF-induced stimulated electromagnetic emission spectra ob-
tained using an interferometer, Ann. Geophys., 24, 1819–1827,
doi:10.5194/angeo-24-1819-2006, 2006.
Thidé, B., Kopka, H., and Stubbe, P.: Observations of stimulated
scattering of a strong high-frequency radio wave in the iono-
sphere, Phys. Rev. Lett., 49, 1561–1564, 1982.
Thidé, B., Hedberg, A., Fejer, J. A., and Sulzer, M. P. : First obser-
vations of stimulated electromagnetic emission at Arecibo, Geo-
phys. Res. Lett., 16, 369–372, 1989.
Wright, D. M., Davies, J. A., Yeoman, T. K., Robinson, T. R.,
and Shergill, H.: Saturation and hysteresis effects in ionospheric
modification experiments observed by the CUTLASS and EIS-
CAT radars, Ann. Geophys., 24, 543–553, doi:10.5194/angeo-
24-543-2006, 2006.
Ann. Geophys., 33, 983–990, 2015 www.ann-geophys.net/33/983/2015/
