Conference PaperPDF Available

Auroral Acoustics project – a progress report with a new hypothesis

  • Aalto University, School of EE, Espoo Finland

Abstract and Figures

Eerie sounds associated with bright and lively moving aurora borealis have surprised and even frightened observers for thousands of years. Since the 17th century their existence and cause have been at the centre of many speculations and hypotheses. These have ranged from a totally skeptical view of their existence to some possible physical or physiological explanation. No widely accepted physical model or reason for these sounds exists. There was at least one attempt to measure these sounds, performed at the Geophysical Institute in Alaska, during 1962-64. However, this did not yield a single recorded example or scientific publication. In February 2000 two researchers from the Sodankylä Geophysical Observatory (SGO) and the author agreed to start an informal co-operation on this research topic. The project was called Auroral Acoustics. It began by publishing some webpages to provide general information on the topic and to collect testimonies of auroral sound observations in Finland. These were also collected by SGO via telephone interviews. To date this dataset contains more than 300 observations. The design of the scientific setup for sound and EM-field measurements started in spring 2000. So far the Auroral Acoustics project has had two active phases; the first phase of the project occurred during solar cycle #23 (2000-05) and the second phase during cycle #24 (2011-ongoing). The project was basically inactive during the five years around the solar minimum between the activity peaks. This paper describes the two phases of the project and presents the main results of the sound and EM-field measurements recorded during more than one hundred auroral nights. The estimation of the locations of the sound sources has led to a new hypothesis for a possible physical mechanism to explain these sounds. The new hypothesis is based on the temperature inversion layer found in the lower atmosphere.
Content may be subject to copyright.
Auroral Acoustics project a progress report with a new hypothesis
Unto K. Laine
Aalto University, Department of Signal Processing and Acoustics, Otakaari 5A, 02150 ESPOO, Finland, unto.laine(at)
Eerie sounds associated with bright and lively moving aurora borealis have surprised and even frightened observers for
thousands of years. Since the 17th century their existence and cause have been at the centre of many speculations and
hypotheses. These have ranged from a totally skeptical view of their existence to some possible physical or
physiological explanation. No widely accepted physical model or reason for these sounds exists. There was at least one
attempt to measure these sounds, performed at the Geophysical Institute in Alaska, during 1962-64. However, this did
not yield a single recorded example or scientific publication.
In February 2000 two researchers from the Sodankylä Geophysical Observatory (SGO) and the author agreed to start an
informal co-operation on this research topic. The project was called Auroral Acoustics. It began by publishing some
webpages to provide general information on the topic and to collect testimonies of auroral sound observations in
Finland. These were also collected by SGO via telephone interviews. To date this dataset contains more than 300
observations. The design of the scientific setup for sound and EM-field measurements started in spring 2000. So far the
Auroral Acoustics project has had two active phases; the first phase of the project occurred during solar cycle #23
(2000-05) and the second phase during cycle #24 (2011-ongoing). The project was basically inactive during the five
years around the solar minimum between the activity peaks.
This paper describes the two phases of the project and presents the main results of the sound and EM-field
measurements recorded during more than one hundred auroral nights. The estimation of the locations of the sound
sources has led to a new hypothesis for a possible physical mechanism to explain these sounds. The new hypothesis is
based on the temperature inversion layer found in the lower atmosphere.
1 Introduction
Northern lights, aurora borealis, as it was called by Galileo Galilei in 1619 A.D., are caused by charged high-speed
particles (e.g., protons and electrons) carried by the solar wind and guided by the magnetic field of the earth towards its
magnetic poles. Different colours (wavelengths) of the northern lights are produced by atoms in the upper atmosphere
when they come into contact with these energetic solar wind particles. The process occurs in the ionosphere at an
altitude between 80-300 km in an oval shaped formation around both magnetic poles. Thus there are not only northern
but also southern lights. The phenomenon can be observed more often in the high latitudes (around 60°-70°N or S).
Only the most energetic events, that may occur once or twice in a decade, can be observed in the lower latitudes, e.g. in
central Europe. A strong coronal mass ejection in the vicinity of a large sunspot is the main source of the strongest solar
wind disturbances and geomagnetic storms that create bright auroras. Therefore, auroral activity is synchronized with
solar cycles that last about eleven years.
Strange sounds that sometimes accompany bright northern lights have been difficult to study and even more difficult to
explain since the light phenomenon occurs at such a high altitude. Due to the strong atmospheric attenuation of sound
waves, no audible components remain after their long travel time from the ionosphere to ground level. Only infra
sounds (< 20 Hz) are able to survive the long journey. Another strange aspect is connected to this phenomenon: the
sounds are observed sometimes simultaneously with movements or intensity modulations of the lights. This should not
happen if the sound source is located close to the light source (at the ionosphere level). These two facts have caused
many physicists to be deeply sceptical about the existence of these sounds entirely. According to them the sound
observations must be illusions or a kind of malfunction of our senses or our cognitive system. One hypothesis has been
that a strong geomagnetic storm with its electric and magnetic field disturbances will directly affect our brains and
cause them to create these sound illusions. Another hypothesis is that the bright and fast moving visual forms of the
aurora will create the illusion of sound in a synesthetic way. Thus the basic problem in relation to these observations has
been: i) are the observed sounds real and physical (objective) in nature, or ii) are they more like subjective illusions
caused by some other physical/physiological effect than acoustic sound waves, and further, if the sounds are real, where
is the sound source and what is its physical mechanism?
When our informal project started in February 2000 the name of the project was chosen to be broader than just focusing
directly on the ‘auroral sound’ aspect. The selected name Auroral Acoustics emphasizes different possibilities to solve
the problem based on the whole spectrum of methods and tools from physical acoustics to psychoacoustics. Today we
know that the usage of the whole spectrum of methods was also a necessity for the emergence of a new hypothesis
presented in this paper. Thus, the solution of this problem seems to be more strongly connected to acoustics and even
psychoacoustics than to classical geophysics, even though earlier attempts to solve the problem have purely risen from
that direction. However, as it will be seen, the new hypothesis is still strongly connected to the physics of the lower
atmosphere and therefore both fields of knowledge are needed to study and understand all the aspects of this apparently
complex process.
In the following sections, the two main phases of the project that are connected to solar cycles #23 and #24 are briefly
described, along with the preliminary results of the important event that occurred on March 17-18, 2013. This recording
finally provided strong scientific evidence for the physical existence of the sounds and opens the door for a new
hypothesis, of where and how the sounds may be created.
2 Development of methodology
There have been thousands of observations of the aurora related sounds around the auroral zones and hundreds of
speculative scientific papers of this topic published over hundreds of years. Surprisingly there has been only one known
attempt to really measure and study these sounds before the start of our project. This occurred at the Geophysical
Institute in Alaska, during 1962-64 in a project entitled: “Investigation of Auroral Sound Waves” (NSF Grant G-
22225). Here they mounted two condenser microphones about twelve miles apart to monitor vertical sound waves. The
signals were stored automatically by a photometer-triggered tape recorder. They met up with signal analysis problems
and reported: local noise sources caused interference which was impossible to eliminate. During the winter of 1962-
63 no clear-cut cases of aurorally associated noise were recorded”. During the next winter they abandoned audio signal
recording and rather concentrated on ground level electric field measurements. Their final report contains the following
summary: “This two year program for exploring auroral sound emissions has not given any firm, unambiguous
indication of the phenomenon. We believe that this is mainly due to the fact that the experiment was carried out too
close to sunspot minimum, when auroral events of sufficiently high intensity do not occur frequently enough over
College, Alaska, rather than non-existence of the phenomenon.” [1]
The fact that the study was started close to the solar min, just a few years after four intensive auroral events, and was
continued when the sun was quite inactive, probably supported the negative outcome. Even though the team clearly
indicated that this outcome should not be used as evidence for the non-existence of the sounds, this could not stop the
tendency to believe so [2]. Since all the information available of this experimental study, containing two annual reports,
fits on only one page, we could not find much help from their experience when planning our project. Naturally, the
audio technology has radically changed from those years where analog tape recorders were used. Today the best
microphones have a very low internal noise (around -3 dB) and the signals can be stored with 24-bit resolution at a 48
kHz sampling rate. Furthermore, by using a microphone array the direction of the sound source can be solved. Table 1
summarizes the audio equipment used during our project.
Our very first experiments with outdoor recordings were performed in February 2000 in Sodankylä together with Esa
Turunen and Jyrki Manninen from the SGO. Planning of the methodology had just started when a strong geomagnetic
storm occurred on April 6-7, 2000. A report of that event and an analysis of the partially corrupted signals were
presented at BNAM-04 [3]. In the summer of 2000 the project started to use a very low noise condenser microphone
(B&K 4179) with a preamplifier (B&K 2660) and power supply. The microphone was mounted at the focal point of a
parabolic reflector in order to increase its vertical directivity (see [4] for details). The system is still in use forming the
most valuable component of the setup.
The methodology of the first phase of the project (2000-05) was based on two-channel DAT recordings with one
microphone (B&K 4179) and a vertical VLF-antenna to measure the electric field fluctuations. At the beginning of the
second phase (2011-ongoing) a four-channel Zoom H4n recorder was included as a member of the microphone array.
Also, the vertical VLF antenna was changed to a more sensitive VLF loop antenna made of two hundred meters of wire
and having a symmetric output suitable for the microphone preamplifier. One of these nightly test recordings picked up
short clap sounds with all three microphones allowing, together with the simultaneous magnetic field measurement, for
the estimation of the direction and altitude of the sound source. The results were published in the summer of 2012 at
ICSV-19, Vilnius, Lithuania [5]. The most surprising result was that the sound source was found to be located in the
open sky just 70 meters above ground level. This was a preliminary result obtained from a single night of recording
with a few, relatively low intensity clap sounds. Therefore, we had to continue the effort in order to collect more data
with the hope that a more intense event would occur providing a better signal-to-noise ratio for the sounds. This finally
took place in March of 2013. However, before that event the system was developed further. The Zoom H4n had three
problems: first, it caused the signals on different channels to be slightly mixed due to channel leakage, and secondly, it
corrupted the signals with a certain amount of “internal infra sound” possible caused by the periodical update of its
internal memory. The third problem was related to the low sensitivity of its internal microphones. For these reasons the
recorder was changed to a Roland Edirol R-44 recorder and two AKG C480B+CK62 microphones replaced the internal
microphones of the H4n.
One more change was performed on the setup after a breakdown of one of the AKG microphones in cold weather. They
were replaced by Sennheiser MHK8020 microphones in 2014. This is the present form of the measuring system. The
microphones are typically mounted in a triangle at distances of 3-4 meters from each other. The B&K microphone (X)
is at the northern apex and the two Sennheiser microphones (L&R) at the east and west apices. The Sennheiser
microphones have a spherical directivity pattern and therefore can also monitor sounds coming from the vicinity of the
microphone array.
Table 1: Audio Equipment used during the Auroral Acoustics project during 2000-2016
Year Months
Phase I
2000 2
Sony DAT
2000 4
Sony DAT
2000 6-12
Sony DAT
Sony DAT
Vertical (E)
Phase II
2011 6-12
2012 1-11
B&K-2 + Zoom H4n
B&K-2 + Zoom H4n
Zoom H4n
Zoom H4n
Coil (M)
Coil (M)
2012 11
B&K-2 + 2*AKG-2
Coil (M)
2014 8- present
B&K-2 + 2*Sennheiser
Coil (M)
B&K-1 = Brüel & Kjær 4165
B&K-2 = B&K 4179 + preamplifier 2660
AKG-1 = C460B+CK61-ULS
AKG-2 = C480B+CK62
One important question in the methodology is when and where should the measurements be carried out? Based on the
short reports of the Alaska (1962-64) project, the measuring system started the recording automatically based on light
intensity, i.e., the system was turned on when the aurora illuminated the night sky. No one was monitoring the recording
process and the environment during the recording. This method is simple and practical but also has some weak points.
A sabotage caused by humans or animals could be conceivable even though unlikely. However, in today’s world,
instruments may be more readily stolen when left by themselves in more populated regions. For these reasons a decision
was made to monitor the whole recording session in a car close to the microphones. Naturally, this method is not very
convenient. This takes a lot of work and one should be ready to stay in the cold and dark night to monitor the
instruments and the environment. In the field of geophysics this type of work is often organized to take place in a form
of a campaign lasting 1-2 weeks. However, these weeks must be fixed months beforehand and there is no guarantee that
any auroral event will occur within the time frame. Therefore, the most efficient method is that someone is 24/7-ready
to go to a quiet place with the instruments and stay there overnight always when the local weather and space-weather
are preferable. This has been the main approach of our project.
One more question concerns what kind of environment is preferable for these measurements. This may be one of the
most difficult questions since there is not enough experience to make selections between “good” or “bad” measurement
landscapes. The most natural and important thing is that the location should be audibly as quiet as possible and located
at a reasonable distance from the main noise sources. Based on the collected observation testimonies we believed from
the very beginning of the study that the sound source exists in the open sky and not from trees or other objects.
Therefore, the focus of the most sensitive microphone was designed to be at the zenith and the recordings were
performed in open fields or on frozen lakes. The general shape of the landscape in the vicinity of the setup was selected
to be quite flat. Measurements in deep valleys on high hills have not been performed during this project. Hopefully, the
possible influences of the landscape can be understood better in the future.
3 Progress, crisis and turning points
The April 6-7, 2000 event [3] with a dramatic, purple aurora filling half of the sky and with a distinctly audible noise
was a very strong experience that will stay in the memory for the rest of the author’s life. The noise was so strong that it
masked the weaker sounds coming from a distance (e.g., birds or power-line hum). The recording method failed
partially, however, the comparison between the aurora night and reference night recordings at the same place and with
the same setup revealed the reason why the power line hum in the vicinity was not audible during the aurora night. It
was masked by auroral noise. The noise rose about 10 dB over the background noise level in the 200-500 Hz frequency
band (see Figure 1). By simple listening experiments the direction of the noise source was estimated to be overhead in
the open sky. This preliminary recording experiment and listening experience formed a strong motivation to continue
the work and to develop a better methodology to study the problem.
Figure 1: Left panel: Auroral noise above the background. Right panel: Power-line hum masked by auroral noise (blue)
and audible power-line hum in the reference recording (red).
The next strong geomagnetic storm under favourable weather conditions occurred on April 11-12, 2001. The
measurements were carried out at Koli in eastern Finland, about 500 km from Helsinki. The recording contains plenty
of different short sound events from crackling to loud reports. The material was analysed in parallel with two
researchers applying two different types of programs (Mathematica and Matlab). The main results are published in the
M.Sc. thesis of Janne Hautsalo [4]. The collected data still contain many details that need more analysis. Also, the
possible connections between the VLF and acoustic signals need much more work. However, the main result of the
analysis was that there is a relatively strong statistically significant correlation between geomagnetic activity and
measured sound power fluctuation. The same data even showed a strong correlation in the infra sound area with a delay
of about six minutes. The system was able to record sounds in the 1 Hz 24 kHz band. Some of the infra sound waves
were so strong that signal clipping occurred. After this experience a simple high-pass filter was designed and applied for
the B&K signal before its AD-conversion in order to avoid saturation.
Even though we were highly motivated to continue the study, we were not able to get any support for this work.
Therefore, the project was practically inactive for five years. We were also at the focal point of some quite hard
criticism from some scientists with statements such as: “Your measurements must be wrong” - “These sounds cannot
exist because there is no mechanism able to produce them”, etc. However, we received positive feedback as well.
People around Finland continued to discuss the topic and also helped us by sending reports of their personal sound
observations. These supports were very important during the “frozen” period of the project.
There were not many choices: should we continue without any external support using our own means, or, stop the entire
effort. However, during this break new ideas and plans started to gradually emerge. Continuing with the old
methodology would probably not lead to any new results. Part of the criticism was based on the fact that we didn’t even
know the location of the sound source. Some people still thought that the project might locate the sound source, e.g., as
being inside the observer’s head, in the hair, on the ground, in the trees, etc. All this sounds quite natural since how
could the sound source be just above our heads in the clear, open sky where only the aurora is seen but at such a large
distance that no audible sound could reach us?
If the study was to continue locating the sound source was now understood to be the most important task. Knowing the
location could also help to understand the whole sound production mechanism. Finally, a decision was made to continue
the project by extending the setup from the two-channel system to a four-channel one consisting of three microphones
and a VLF loop antenna specially designed to match the microphone preamplifier of the digital recorder with a
symmetric input (XLR) (see Table 1).
One of the first experiments with the new four-channel setup occurred in September 2011 in the vicinity of the Fiskars
village in southern Finland [5]. A new Zoom H4n was tested as a component of the microphone array. A few days later
when the signals were analysed, twenty-one clap sounds were found. The strongest ones were present in all three
microphone-signals. This allowed for the estimation of the sound source direction for the first time in the project. When
the ground reflection and the VLF signal data were combined with the audio signals, the estimation of the altitude of the
sound source was possible. The obtained altitude estimates were about 60-70 m above ground level. This was a
surprising result and was even hard to believe. How could the sound source be located above our heads at that height
and eminating from the empty sky? This new result made the problem even stranger. A preliminary report was
published but it was clear that this question needed much more work and data collection from new aurora events.
3.1 March 17-18, 2013 aurora event
During this auroral event the weather in Fiskars village was cold (-20° C) and calm. The setup was reconstructed around
a Roland R-44 digital recorder with two new AKG microphones (see Table 1). The geomagnetic storm was strong with
kp values between 5-6 during 06-24 UTC and the dst-index min was -132 (Kyoto, provisional value [6]). Hundreds of
short sound events were detected in the recorded signals. The majority of them were normal environmental sounds.
However, over sixty loud sound events were detected just above the measurement location and they all had a
counterpart event in the magnetic field signal. The average time difference between the magnetic impulses to the
acoustic events indicated that the altitude of the sound sources were on average 75 meters above ground (see Figure 2).
Figure 2: Example of the magnetic field and audio events. Blue: coherent mean of 49 VLF pulses. Red: coherent mean
of 69 envelopes of the sound events. The average delay is used to estimate the sound source distance from the ground.
The energies of the magnetic pulses and the energies of the audio events were computed to solve the rank correlation
between them (Kendall’s Tau). The correlation is so high that we can say that these two sequences have a causal
relation with a 99.9% probability and with a 0.1% chance that this claim is wrong (confidence interval = 3.3 σ). This
new result confirms the one obtained in 2011. In science, the 95% confidence interval is commonly accepted as a
reliable confirmation of a claim under scrutiny. In this case the outcome is that these sixty short sound events had their
sources above the measurement location on average 75 meters above the ground and that the sound generating process
in that altitude radiated magnetic pulses as well. This result seems to confirm the hypothesis of brush discharging
mechanism in the atmosphere as a possible source of the aurora related sounds. The outcome forced to construct a new
hypothesis of the possible sound generating mechanism in the lower atmosphere.
4 Inversion hypothesis
At any season of the year, during a calm and clear evening, the temperature inversion layer builds up from the cooling
ground up to some hundreds of meters [7]. The altitude of the inversion layer in the studied cases when aurora related
sounds have been measured has varied between 60-400 meters.
According to Piper and Bennet [8], “Space charge builds up during the night and is trapped by the temperature
inversion, which acts as a lid to convective motion. This build-up of space charge under the temperature inversion
results in a negative PG (potential gradient at the ground)”. A negative PG means that the surface clear wetter electric
field may be reversed due to the concentration of negative charges under the inversion layer. From the point of view of
the space charge, the inversion layer can be called a convective boundary layer that hinders (like a lid) the convection of
negative space charges to the upper layers of the atmosphere (see also [9]).
Figure 3: The inversion hypothesis.
The inversion hypothesis is briefly described in Figure 3. A temperature inversion layer forms an isolating lid at an
altitude of 60-400 meters. Warm air with aerosols rises up and carries negative charges to the lower part of the inversion
layer. A geomagnetic storm increases the conductivity of the upper atmosphere. Positive charges are accumulated above
the inversion layer. The electric potential over the layer is increasing similarly to a capacitor during charging. Finally, a
sudden discharge occurs that produces a crackling or clapping sound together with magnetic pulses.
During the March 17-18, 2013 night, the temperature inversion layer was about 70-80 meters from the ground (see the
balloon measurement in Figure 4). The balloon measurement was performed in Jokioinen that is about 75 km to the
north of the Fiskars village. The altitude matches the estimated distance of the sound source from the ground. Thus the
sound source must be located near or in the inversion layer. Also a loud, controlled, short duration test sound was
created that night at ground level and its reflection from the inversion layer was measured. This result confirms the
inversion layer altitude.
Figure 4: Part of the temperature profile measured in Jokioinen at midnight, March 17-18, 2013. A strong temperature
inversion is observed at 70-80 meters from the ground (data from Finnish Meteorological Institute, archived at the
University of Wyoming). The red box is used to graphically estimate the distance to the inversion layer.
Based on the magnetic field pulses associated with the measured short duration sound pulses, the sound producing
mechanism must be somehow related to an impulsive current in the inversion layer. This speaks for a kind of electric
discharging mechanism, which has been proposed several times as a possible source for the aurora related sounds. For
example, Silverman and Tuan made a wide and comprehensive study of different physically possible and plausible
mechanisms that could produce these sounds [10]. Almost all the studied alternatives were not able to get much support
in their study. The outcome was that only a discharging mechanism in the lower atmosphere could form a physically
plausible reason for these sounds. Where and how these charges could then accumulate and what caused the rapid
discharged was left open. The inversion hypothesis discussed here may provide answers to these questions. However,
the hypothesis reveals also many new questions we have to be able to answer in the future.
4.1 Simultaneity of a sound with a visual stimulus
One branch of psychoacoustics deals with 3D-sounds and the perception of sound source direction. When sound has a
short duration, like a clap or report, a human observer can focus their attention very quickly to the direction of the sound
source. Typically, we are fast and accurate in estimating the location of such sound sources. This has probably helped
us to survive during human evolution. When the observers of the aurora-related sounds reported that the sound came
from the sky, there is a good reason to believe that these observations are valid. Therefore, the inversion hypotheses and
the measurements are in harmony with these observations.
Another branch in psychoacoustics deals with the problem of simultaneity: how much can a short duration visual
stimulus and sound deviate in time before the sensation of simultaneity is broken? If the sound precedes the visual
stimulus the sensation of simultaneity is broken much earlier than in the opposite case when the sound stimulus is
delayed in relation to the visual one. In front of a complex visual scene where many changes occur at the same time and
where the association of the sound to the visual stimulus is difficult, the allowed delay may be much larger than in the
case of a simple visual stimulus. Based on the present literature the sound stimulus may be delayed by more than 200
ms before the sensation of simultaneity is broken [11].
Under the active and complex visual scene of an aurora this delay may be even larger. Therefore, it is possible that
when the sound source is about 60-80 meters above the observer, it is very difficult to perceive the delay of the sound in
comparison to changes in the visual stimulus. The sounds emitted by this source are mostly perceived as simultaneous
effects with the visual source. Therefore, one more mystery in the field of aurora related sounds can be solved with the
presented inversion hypothesis.
5 Discussion and summary
Aurora-related sounds are not the only strange audible phenomenon outdoors during evenings and nights. The sounds
associated with meteors form a closely related topic that has also been discussed for a long time in history. Our project
has been contacted by researchers who are interested in these sounds. One team is currently working on this problem at
Armagh Observatory, Northern Ireland. The exchange of experiences and ideas with these researchers revealed that we
have worked on similar study lines independently of each other. One of their hypotheses is practically the same as
described in this paper. Not only aurora related sounds but also meteor related sounds can be solved within this same
The past sixteen years of active studies, measurements, and data analysis has finally led to a new hypothesis on the
origin and physical mechanism of aurora-related sounds. The journey has been quite long and demanding encountering
many surprises and turning points. The discovery of the location of the sound source opened the door for a new
hypothesis that combines the present understanding of the behavior of space charges in the temperature inversion layer
above ground to the measured evidence of a discharging mechanism found inside this layer. Additionally, the new
hypothesis is in harmony with the majority of documented sound observations. The sound source is localized in the sky
and the sound is often perceived simultaneously with active movements or other visual changes in the auroral forms. At
the moment, all details seem to be in harmony with this hypothesis. It also explains why the sounds cannot be audible
for every auroral event. However, it is still possible that this mechanism is just one in a plurality of different
mechanisms. For example, some new aurora-related sound sources might be found even much higher in the atmosphere.
To summarize, the ‘auroral theater’ is controlled by a geomagnetic storm and it has separate sound and light sources at
different distances from the observer. The inversion hypothesis at its present level of development does not explain in
detail how and why the discharging mechanism is triggered. Could it be caused by some high energetic particles, could
infra sounds have an effect, or are the rapid geomagnetic fluctuations strong enough to trigger the process? Future
studies may illuminate all these details.
6 Acknowledgements
The author is grateful to many friends and co-workers in SGO, FMI, and in the former Acoustics Laboratory at the
Helsinki University of Technology, TKK (up to 2010) and Aalto University (since 2010) for the practical help and
many fruitful discussions during the different phases of this project. TKK tukisäätiö has given valuable support to this
project during 2004-2005. The Finnish Cultural Foundation is also warmly acknowledged for their support (2015).
[1] Annual Report 1963-64, Geophysical Institute, University of Alaska, College, Alaska. 1964, 13.
[2] Can you hear an aurora?
[3] U. K. Laine, Denoising and Analysis of Audio Recordings made during the April 6-7, 2000 Geomagnetic Storm by
using a non-professional ad hoc setup, BNAM-04, Mariehamn, Finland, 2004, 8.
[4] J. Hautsalo, Study of Aurora Related Sound and Electric Field Effects. M.Sc. Thesis, TKK, 2005.
[5] U. K. Laine, Analysis of clap sounds recorded during the September 9-10, 2011 geomagnetic storm, Proc. ICSV-
19, Vilnius, Lithuania, 2012.
[6] Kyoto Geomagnetic Data Service,
[7] Personal communications with meteorologist Petri Takala, Foreca Ltd., 2014.
[8] I. M. Piper and A. J. Bennett, Observations of the atmospheric electric field during two case studies of boundary
layer processes, Environ. Res. Lett. 7, 014017, 2012.
[9] S. V. Anisimov, S. V. Galichenko, and N. M. Shikhova, Formation of Electrically Active Layers in the Atmosphere
with Temperature Inversion, Izvestiya, Atmospheric and Oceanic Physics, Vol. 48, No. 4, 391-400, 2012.
[10] S. M. Silverman and T. F. Tuan, Auroral Audibility, Adv. Geophys., 16, 155-269, 1972.
[11] M. Zampini, S. Guest, D. I. Shore, and C. Spence, Audio-visual simultaneity judgments, Perception &
Psychophysics, 67(3), 531-544, 2005.
... When the Auroral Acoustics project [1] started and the author made one of his first AS observations, a TVreporter asked for a description of the sounds. The answer was: "they sounded like Niagara Falls from a distance of one kilometre". ...
... Tree tops are quiet. The present Inversion Layer Hypothesis (ILH) provides a physical explanation for the sound production [1]. It assumes that after a calm and sunny day the warm air at the ground rises carrying negative ions up forming a Temperature Inversion Layer (TIL) about 70-100 metres from the ground. ...
... The same measuring system as described in [1] was used in this recording, too. Its 'soul' is the high-quality measuring microphone of Brüel & Kjaer (4179 with preamp 2660) that was mounted at the focus of a parabolic dish antenna directed vertically to the sky. ...
Conference Paper
Full-text available
Sounds associated with aurora borealis, so called Auroral Sounds (AS), are explained by electrical discharges occurring in the Temperature Inversion Layer (TIL) approximately 75 meters above the ground [1]. It is assumed that under favourable weather conditions the number of AS events depends on the activity of the geomagnetic (GM) storm. Historical testimonies of AS indicate that the GM storm should be strong enough to rise bright and lively auroras to the zenith before these sounds can be observed. The goal of this study is to test the relevancy of this claim by studying the sensitivity of the AS producing mechanism under moderate GM activity. A four-hour sound measurement was performed around the local magnetic midnight on Jan 25–26, 2022 at Fiskars village in southern Finland. Sixty AS event candidates were manually selected and their temporal distributions and period histograms constructed for statistical comparisons with the GM data measured simultaneously by the Finnish Meteorological Institute (FMI) at Nurmijärvi Geophysical Observatory (NGO). The data was also used for regression models to study possible causal relationship between the GM activity and the AS events. A strong causality was found. The temporal distribution of the sounds was predicted by the GM activity with 90% accuracy after a delay of 21 min. The results show that in good weather conditions the AS events are much more common than previously thought.
... With the help of further improved instruments and methodology, locations of the sound sources in the clear sky were finally estimated [6]. A new hypothesis explaining the sounds was published in 2016 [7]. Based on the measurements, it appears that the sounds are created by corona discharges in the temperature inversion layer at 70-80 meters altitude. ...
... Both types of the measured auroral sounds, individual claps and longer sequences like cracklings, are connected to the space charges and the geomagnetism: both sound types have shown a clear causality to the preceding magnetic field pulses caused by the corona discharges (high RMS correlation) [7] and now the crackling sounds show a clear connection to the Schumann resonances. After these discoveries, there should not be any doubt that these sounds are of a special kind and tightly connected to space charges, geomagnetism, and the aurora borealis. ...
... ln 2001 when the Phase I of the Auroral Acoustics project was active, the setup used for sound and EM-field measurements was simple and easily portable. It consisted of Sony DAT recorder (TCD-D3), B&K microphone (4179) that was mounted at the focus of a satellite antenna (focused to the zenith) and a simple vertical VLF antenna connected directly to the R-channel of the DAT system [7]. The sampling frequency was 48 kHz (16-bit samples). ...
Conference Paper
Full-text available
Observations of unusual sounds like claps, snaps, and cracklings from a serene sky under actively moving, bright aurora borealis have been reported during a long history. Since year 2000 the auroral sounds have been studied in Finland. In 2011 a sound source associated with the aurora was localized for the first time in the sky approximately 70 meters from the ground. To accommodate for the surprisingly low altitude of the source, a hypothesis that the sounds are produced by corona discharges in a temperature inversion layer at the observed altitude was pub-lished in 2016. The focus of the present paper is in the acoustic analysis of the crackling sounds that are sequences with complex temporal structures. It is shown that the sounds share similar rhythmical patterns, which are connected to the electromagnetic resonances in the atmosphere, also known as the Schumann resonances.
... The first phase (2000)(2001)(2002)(2003)(2004)(2005) of the project called Auroral Acoustics succeeded to show statistically significant connections between the sounds measured on the ground level and the simultaneous geomagnetic activity measured by the Finnish Meteorological Institute (FMI) [1]. The ongoing second phase of the project (2011-) started by studying new methodology based on an array of three microphones (instead of one measuring microphone) and a loop antenna for magnetic field measurements (instead of a vertical antenna for electric field measurements) [2]. The development of the measuring system started by applying a Zoom H4n recorder. ...
... The data collected with this system in September 9, 2011 gave the first estimates for the altitude of the sound sources [3]. Later, due to the quality problems, Zoom H4n was replaced by Roland R44 recorder first with AKG and then with Sennheiser microphones [2]. The statistical analysis of measured sounds and magnetic field pulses (MFPs) revealed the possible physical mechanism causing the MFPs and the so-called auroral sounds audible on the ground level. ...
... During a geomagnetic storm with bright and lively aurora a rapid change in the geomagnetic field may trigger a discharging process causing both MFPs and after a delay audible clapping or crackling noise on the ground. This is the basic idea of the AES 144th Convention, Milan, Italy, 2018 May 23-26 Page 2 of 7 present proposal for the sound producing mechanism that is called the Inversion Layer Hypothesis [2]. ...
Conference Paper
Full-text available
Audio- and magnetic field signals were recorded during a geomagnetic storm on March 7, 2012, on open fields at Karkkila, approximately 70 km north of Helsinki by using a Zoom H4n recorder. Almost 90 distinct sound events like short claps, loud reports, or even a crackling sound were recorded. The paper describes the methods used and the results obtained in the audio- and magnetic field signal analysis. Relationship between the instances of the sound events and the geomagnetic activity is described. It is shown that the spectral properties of the crackling sound and the reports are similar. The challenges in finding connections between individual sounds and the corresponding magnetic field fluctuations are demonstrated and discussed.
... This work is a continuum of the free running project Auroral Acoustics that started in the former Helsinki University of Technology (TKK) in the spring 2000 and continues in the Aalto University. The main goal was to study the problem of so-called auroral sounds [1,2]. Based on human observations made under bright and lively aurorae plenty of evidence of these sounds have been collected around the world for hundreds or even thousands of years. ...
... However, they may be large enough to trigger discharging in the temperature inversion layer, where the space charges from the ground and also from the upper atmosphere have accumulated during a calm and clear evening. This new hypothesis called the inversion layer hypothesis (ILH) was published in Stockholm in the summer 2016 [2]. ...
... The auroral event on March 17-18, 2013 and the sound measurements have been described in an earlier report [2]. In short, the set-up consisted of a four tract recorder (Roland R44), three microphones (AKG stereo pair CK62 called Land R-mic and Brüel & Kjaer 4165 called X-mic) and a self constructed VLF loop antenna with bifiliar winding and a symmetric output that fits directly to the microphone XLR connector of the R44-recorder. ...
Conference Paper
Full-text available
The sounds that occasionally accompany the bright and active northern lights – the aurora borealis, have been studied in a small project called Auroral Acoustics at the former Helsinki University of Technology (2000-2009) and the Aalto University (since 2010). In September 2011, the project succeeded in localizing sound sources in the open sky some 70 meters above the ground. On March 17–18, 2013 during an intense auroral display over southern Finland, hundreds of short duration sound events were recorded with a three-microphone array accompanied by a loop antenna for magnetic field measurements. In sixty cases the sound source was estimated to be directly above the microphone array. A pulse in the magnetic field preceded each of the strongest sounds, indicating an impulsive current in the vicinity of the sound source. Rank correlation between the RMS values of the magnetic pulses and the sounds showed a 99.9% probability that the sounds had a causal relation to discharging currents (possibly corona) about 75 meters above the ground. Concurrent atmospheric measurement data also show that a temperature inversion layer was located at the same altitude. This paper presents detailed analysis of one sound type in the March 2013 data. The inversion layer hypothesis (published in 2016) is described and tested by fitting a model to the data. The obtained results support the hypothesis and estimates that the altitude of the active inversion layer on that evening was at 78–80 meters. Moreover, the analysis revealed a new noise producing process that is associated with those low frequency sound events in both audio and magnetic field (VLF) components. This noise producing process likely takes place in the inversion layer, too.
... In more recent years, a group at Aalto University in Helsinki, Finland, led by Unto K. Laine has taken up the mantle of investigating the 'eerie sounds' of the aurora. 79 In the year 2000 they set up sensitive outdoor recording devices and a parabolic reflector to increase the vertical scope of the apparatus at Sodankylä, a small village in Northern Finland used by the Finnish research group during the First IPY. An estimate for the source of the sound held that it occurred, surprisingly, at 70 m above ground level. ...
Full-text available
This paper discusses heightened interest in the potential audibility of the aurora borealis during the First and Second International Polar Years (IPYs) of 1882–3 and 1932–3. Galvanized by a growing volume of local accounts expressing belief in the elusive noises, written by the inhabitants of the Shetland Islands, northern Canada, and Norway, auroral researchers of each era were determined to establish the objectivity of auroral sound. There was considerable speculation within the auroral research community as to whether the apparent noises were imagined or illusory, connected to discussions about the possibility of low-altitude aurorae. The anglophone auroral sound debate primarily played out within the official reports of IPY expeditions, the journal Nature , and a Shetland Island newspaper. I argue that the embodied senses were used exclusively to register the liminal sounds of the aurora across the two periods, despite developments in sound recording technologies, the primacy of mechanical objectivity, and instruments transported to the polar regions for the investigation of visual features of the phenomenon. This overlooked episode complicates narratives of polar science in the late nineteenth and early twentieth centuries by revealing a faith in the corporeal senses and the significant role of amateur observers.
Full-text available
We present measurements of potential gradient (PG) with associated meteorological variables and cloud profiles for two examples of convective boundary layer processes. Aerosol acts as a tracer layer to show lofting of the convective boundary layer; the rising aerosol layer results in a decrease in PG. In foggy conditions, the PG is seen to increase during the fog and then reduce as the fog lifts, as expected.
Full-text available
An unpredictably strong geomagnetic storm took place on April 6-7 2000. A scientific study on aurora related sounds and acoustical effects had just been started. There were two possible choices: to perform audio recordings with a non-professional ad hoc setup, or, to miss a promising possibility to study the phenomena. The first choice was selected and it produced over seventy-five minutes of data, corrupted with impulsive noise from a DC/AC-inverter used to power recording equipment in the field. The data consists of ten successive recordings. This paper describes a novel method to cancel the impulses found in data without affecting the microphone signal significantly. Next, the power of the cleaned signal is computed in 1/3-octave bands. The spectral comparisons of the different recordings showed a clear increase in the average power in recording #5 that was sampled somewhat after the peak of geomagnetic activity. Specifically, the frequency band around 100 Hz showed an increase in power and also in its fluctuation during the most active geomagnetic periods. This new result supports those obtained earlier [4]. Also, the role of related infrasounds is preliminary discussed.
Full-text available
The relative spatiotemporal correspondence between sensory events affects multisensory integration across a variety of species; integration is maximal when stimuli in different sensory modalities are presented from approximately the same position at about the same time. In the present study, we investigated the influence of spatial and temporal factors on audio-visual simultaneity perception in humans. Participants made unspeeded simultaneous versus successive discrimination responses to pairs of auditory and visual stimuli presented at varying stimulus onset asynchronies from either the same or different spatial positions using either the method of constant stimuli (Experiments 1 and 2) or psychophysical staircases (Experiment 3). The participants in all three experiments were more likely to report the stimuli as being simultaneous when they originated from the same spatial position than when they came from different positions, demonstrating that the apparent perception of multisensory simultaneity is dependent on the relative spatial position from which stimuli are presented.
The dynamics of the electric field in the planetary boundary layer (PBL) is thoroughly studied from in situ observations of the aeroelectric field and the height profiles of the wind-velocity components in the conditions of temperature inversion and incipient convection. It is established that the formation of a layer with temperature inversion is accompanied by a positive trend in the intensity of the aeroelectric field and by the generation of short-period aeroelectric pulsations. The transfer of a spatially nonuniform space charge and the formation of electrically active layers in PBL are studied by numerical modeling. The response of the electric field to the motion of the space charges simulating the coherent structures of electrogasdynamical turbulence is investigated for the vicinity of the observation point. The key parameters of the model distributions of the space charge are analyzed. The linear dimensions of the model structures range from 20 to 500 m, and the density of the transported charge varies from 0.1 to 1 nC/m3. The layer containing the model structures is located at a height of 60–300 m. It is shown that the spatial distribution and the transfer of the space charge form the dynamical component of the aeroelectric field in the surface layer. The short-period aeroelectric pulsations are induced by the transfer of the spatially heterogeneous space charge in PBL, while the positive trend is due to the accumulation of the space charge below the inversion layer. When the inversion was recorded by a sodar, the intensity of the field at the onset of the convection increased at a rate of 100 V/(m h) on average.
Sounds that accompany intense auroral displays have been reported throughout history. Because there has been no instrumental evidence of the sounds, the idea that these sounds are the result of auroral processes has been questioned. At the same time, the research done on the observational material of the sounds has suggested that these sounds are real physical effects. Infrasound produced by aurora has been registered by a number of research teams. However, the few attempts to record audible aurora related sounds have failed. The biggest reason for the failure is that a healthy human ear has been a more sensitive instrument for the detection of faint sounds than the recording technology of the past. This thesis gives a review of the previous studies of aurora related sounds, which almost entirely deals with the observational reports. The results of the survey of aurora related sound reports by the public in Finland are presented, arranged as a co-operation between the Sodankylä Geophysical Observatory and Helsinki University of Technology. A measurement system for aurora related sound and electric field effects is introduced. The data collected by the introduced measurement set up is analyzed with the aid of statistical analyses. The correlation results lend support to the previous findings on auroral infrasounds. In addition, in the geomagnetically most active part of the most interesting measurement, a correlation is found between the geomagnetic activity and acoustic power in audible frequencies when the geomagnetic data is delayed about 0-70 seconds with respect to the acoustic data.
Analysis of clap sounds recorded during the
  • U K Laine
U. K. Laine, Analysis of clap sounds recorded during the September 9-10, 2011 geomagnetic storm, Proc. ICSV- 19, Vilnius, Lithuania, 2012.
  • S M Silverman
  • T F Tuan
S. M. Silverman and T. F. Tuan, Auroral Audibility, Adv. Geophys., 16, 155-269, 1972.
7] Personal communications with meteorologist Petri Takala, Foreca Ltd
  • Kyoto Geomagnetic
  • Data Service
Kyoto Geomagnetic Data Service, [7] Personal communications with meteorologist Petri Takala, Foreca Ltd., 2014.
Denoising and Analysis of Audio Recordings made during the
  • U K Laine
U. K. Laine, Denoising and Analysis of Audio Recordings made during the April 6-7, 2000 Geomagnetic Storm by using a non-professional ad hoc setup, BNAM-04, Mariehamn, Finland, 2004, 8.