Document Actions

IJEGE-11_BS-Kogelnig-et-alii

background image
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
563
DOI: 10.4408/IJEGE.2011-03.B-062
A STUDY OF INFRASONIC SIGNALS OF DEBRIS FLOWS
a
Rnold
KOGELNIG
(*)
, J
oHannes
HÜBL
(*)
, e
mma
SURIÑACH
(**)
, i
Gnasi
VILAJOSANA
(**)
,
s
HuCHenG
ZHANG
(***)
, n
anyan
YUN
(****)
& b
Rian
w. m
C
ARDELL
(*****)
(*)
Institute of Mountain Risk Engineering, University of Natural Resources and Life Sciences, Vienna, Austria
(**)
Grup d’Allaus (RISKNAT), Dept. Geodinàmica i Geofisica, Fac. de Geologia, Universitat de Barcelona, Spain
(***)
Institute of Mountain Hazards and Environmnent, Chinese Academy of Sciences and Ministery of Water Ressources, Chengdu, China
(****)
Mechanics College, Southwest Jiao Tong University, Chengdu, China
(*****)
Swiss Federal Institute for Forest, Snow and Landscape Research WSL, Birmensdorf, Switzerland
(z
HanG
et alii, 2004; C
Hou
et alii, 2007; H
übl
et alii,
2008; k
oGelniG
et alii, in press). Infrasound signals (fre-
quency range 0.01-20 Hz) are longitudinal pressure waves
that travel through the air at a speed of 343 m/s, which is
the same as that of audible sound. Infrasound signals can
propagate over long distances in the atmosphere with little
attenuation. This is due to selective frequency absorption
of sound waves in the atmosphere - higher frequencies
(e.g. audible) are absorbed more readily than lower fre-
quencies (e.g. infrasound) (P
ilGeR
et alii, 2009). For de-
bris flow monitoring seismic waves as well as infrasound,
both have benefits and drawbacks. The latter is mostly
noise induced from wind or human activities that mask
the debris flow signal. The benefits include no structural
need for sustainability and monitoring from a remote lo-
cation not affected by the process activity. The quality of
monitoring results will depend on the relative positioning
between the mass movement and the sensors as well as the
specific characteristics of the site (e.g. topography).
The aim of this study is to present further results of
infrasound monitoring of debris flows at four interna-
tional sites and to illustrate the potential of infrasound
monitoring of alpine mass movements. The study sites
included the Lattenbach torrent (Tyrol, Austria), the Ill-
graben torrent (Valais, Switzerland), the MiDui Glacier
(Tibet, China) and the GuXiang Glacier (Tibet, China).
The specific equipment, setup and sensor placement
differed between sites. Where available, seismic signals
and flow depth data were used for comparison, correla-
tion and validation of the infrasound data.
ABSTRACT
Mass movements such as debris flows, rock fall
and snow avalanches are sources of sub-audible sounds
in the low frequency infrasonic and seismic spectrum.
Recent studies indicated that debris flow-generated sig-
nals are of significant amplitude and occupy a relatively
noise free band in the low frequency acoustic spectrum.
Infrasound signals have the ability to propagate kilome-
tres from the source, thereby allow monitoring of mass
movements from a remote location. This study presents
debris flow monitoring at four international sites - Lat-
tenbach, Tyrol (Austria), Illgraben, Valais (Switzerland),
and the MiDui and GuXiang Glacier, Tibet (China). The
infrasound sensors used were the Chinese sensor (DFW
I-III) or the German sensor (Gefell WME 960 H). The
results show that debris flows emit detectable low fre-
quency infrasonic signals (1-20 Hz) that are correlated to
seismic signals. The infrasound sensors detect the phe-
nomena before it reaches the sensors, depending on the
landscape, distances and the sensitivity of the equipment.
INTRODUCTION
Rapid mass movements (debris flows or snow ava-
lanches) are periodic or episodic phenomena that present
a hazard for people and property in inhabited alpine ar-
eas. Although efforts to develop debris flow monitoring
or warning devices have increased in the last decades
(a
Rattano
, 1999; i
takuRa
et alii, 2005; l
a
H
usen
, 2005;
b
adoux
et alii, 2009) further research is needed and only
few studies exist of infrasound monitoring of such events
background image
A. kOGELNIG, J. HÜBLE, E. SURIÑACH, I. VILAJOSANA, S. ZHANG, N. YUN & B. w. McARDELL
564
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
50%. Furthermore, an analysis with Power Spectra (PS)
were used, which show the frequency content of a sta-
tionary signal. Debris flows are generally described as
moving downhill in a series of waves or surges, whereby
the flowing body has a steep front with higher material
content and the flowing tail has a more gradual slope and
higher water content (i
veRson
, 1997). These particular
characteristics, which are common to all debris flows,
can also be seen in the flow height data from the ultra-
sonic gauges as well as the seismic and infrasonic data
from this event (Fig. 2 and 3, rectangles).
A more detailed explanation is given in k
oGelniG
et alii (in press). The acoustic sensors detect the debris
flow before it reaches the sensors - 50 s earlier in the case
of the geophone and 90 s earlier for the infrasound sen-
sor. Using an average flow velocity of 6 m/s (obtained
from the ultrasonic gauges), these time differences cor-
respond to 300 m and 540 m, respectively (Fig. 1, B and
C). The peak signal frequencies seen during this event
were approximately 6 Hz for the geophone and 17 Hz
for the infrasound sensor (Fig. 2 and 3). It must be noted
that the geophone device has a cut-off frequency of 10
Hz; however, it is expected that infrasonic waves have
a lower frequency content compared to seismic waves,
and there is generally little signal energy above 15 Hz in
infrasound. According to C
Hou
et alii (2007), peak fre-
quencies of infrasound debris flow signals are thought to
be correlated to the flow characteristics.
LATTENBACH (AUSTRIA)
A debris flow event was recorded on 01.09.2008 in
the Lattenbach torrent (catchment area 5.3 km
2
) (over-
view see Fig.1). The event had a duration of 867 s (de-
fined as time with flow depth >30 cm), a peak discharge
of 380 m
3
/s and a total volume of 14000 m3 within this
time. For further details of this event, the reader is re-
ferred to k
oGelniG
et alii (in press). Data was collected
using an infrasound microphone, a geophone and two
ultrasonic gauges (with an inter-distance of 47.2 m). The
infrasound sensor used at this site was the Gefell WME
960H, which has a frequency range from 0.5 to 20 Hz
and a sensitivity of 50 mV/Pa. The geophone sensor SM4
has a frequency range from 10 to 180 Hz and a sensitivity
of 28.8 V/m/s. The geophone was therefore not able to
register those seismic signals with a frequency less than
10 Hz, resulting in missing data. The infrasound sensor
was placed in the proximity of the upstream ultrasonic
gauge and the geophone for better data comparison. Fur-
thermore, this location has previously been shown to be
optimal for both infrasonic and seismic monitoring as
there is minimal background noise (k
oGelniG
et alii in
press). A Campbell Scientific CR1000 data-logger was
used with a sampling rate of 100 Hz. The signals were
analysed with Running Spectra (RS), which present the
temporal evolution of the frequency content of a signal,
using the Short Time Fourier transformation with a Han-
ning Window (length 128 samples) and an overlap of
Fig. 1 - Overview of Lattenbach torrent
- the catchment area and the af-
fected villages of Grins and Pians
are highlighted. The geophone
detected the debris flow 300 m
upstream (B) and the infrasound
sensor 540 m upstream (C) of
the actual sensor location (A)
(Source: Google Earth)
background image
A STUDY OF INFRASONIC SIGNALS OF DEBRIS FLOWS
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
565
have heights varying between 1 and 7 m and several are
either covered by sediment deposits or are destroyed.
Two infrasound capacity microphones, developed
by the Acoustics Institute at the Chinese Academy of
Science (CAS), were placed 38 m apart in the proxim-
ity of check dam 27. These devices have a frequency
range of 3 to 200 Hz and a sensitivity of 50 mV/Pa.
Unfortunately, this setup was not ideal as the distance
between sensors was inadequate to show a difference
in arrival time within the acoustic signals. Data will
therefore be presented for the upstream microphone
only. Additionally, a seismic velocimeter, model GS11,
The following sections provide a comparison of
peak frequencies during different debris flow events
in other countries.
ILLGRABEN (SWITZERLAND)
The Illgraben torrent is famous for its frequent sed-
iment transport and debris flow activity. This may be
accounted for by both its situation in an area of highly
fractured bedrock (b
adoux
et alii, 2009) and its size
(9.5 km
2
). In total there are 29 check dams located over
the course of the torrent (Fig. 4). Check dam 1 has the
greatest vertical height (48 m), whereas dams 2 to 29
Fig. 2 - RS (a), time series
(b), flow depth (c)
and PS (d) of the
infrasound signal
during a debris
flow on 01/09/08
in the Lattenbach
torrent. Different
debris flow surges
are marked by
the rectangle. The
initiation time cor-
responds to Fig.3.
Infrasound signal
in mV, sensor sen-
sitivity 50mV/Pa
Fig. 3 - RS (a), time series
(b), flow depth (c)
and PS (d) of the
seismic
signal
during a debris
flow on 01/09/08
in the Lattenbach
torrent. Different
debris flow surges
are marked by the
rectangle. The ini-
tiation time corre-
sponds to Fig. 2.
Geophone signal
in mV, sensor sen-
sitivity 28.8V/m/s
background image
A. kOGELNIG, J. HÜBLE, E. SURIÑACH, I. VILAJOSANA, S. ZHANG, N. YUN & B. w. McARDELL
566
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
Data of the infrasonic and seismic background noise at
the Illgraben torrent have been presented in k
oGelniG
et alii (in press); this site generates greater background
noise compared to the Lattenbach torrent, but the am-
plitudes are nevertheless low relative to the debris flow
signal. The torrential process discussed in this paper oc-
curred on 28.07.2009. Unfortunately no video data is
available of this event. Other measurements provided
by the WSL like bulk density (around 1600kg/m
3
) and
flow depth from laser sensors (flow front was small and
was placed near the upstream infrasound microphone.
This device has a frequency range of 4.5 to 100 Hz and
a sensitivity of 90 V/m/s. Data from all three acous-
tic sensors were collected with a Campbell Scientific
CR23 data-logger with a sampling rate of 50 Hz and
were stored on an Xplore iX104 C3 tablet computer.
Finally, ultrasonic gauges were placed at check dams 1,
10 and 27 to monitor flow depth (sampling rate 1 Hz).
These gauges were operated by the Swiss Federal Insti-
tute for Forest, Snow and Landscape Research (WSL).
Fig. 5 - RS (a), time series (b),
flow depth (c) and PS
(d) of the infrasound
signal during a debris
flood on 28/07/09 in Ill-
graben torrent. In order
to show only the debris
flood frequency content
a time window from 1.8-
2.2*104s was chosen for
the computation of the
PS. Infrasound signal
associated with a thun-
derstorm in the area are
marked by the rectangle.
The passing of the debris
flood at check dam 1 and
check dam 10 is marked
by the vertical lines in the
flow depth graph. Infra-
sound signal in mV, sen-
sor sensitivity 50mV/Pa
Fig. 4 - Overview of the Illgraben torrent - the catchment
area and the boarder between mountains and
Rhône valley are highlighted. The infrasound sen-
sor detects the debris flow 1500 m upstream (A) of
check dam 27 and the seismic sensor 2000m up-
stream (B) (Source: Google Earth)
background image
A STUDY OF INFRASONIC SIGNALS OF DEBRIS FLOWS
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
567
corresponds to a distance of 1500 m, which happens to
be the topographical transition between the mountains
and valley near the Bhutan Bridge (Fig. 4, A). Previous
work (k
oGelniG
et alii, in press) also reported that the
infrasound microphone, when placed at check dam 27,
detects the torrential processes at this location.
Infrasound signals generated from debris flows
are believed to be produced by the violent surge front
and the collisions (or abrasion) between the flow and
the channel loose boundary (C
Hou
et alii, 2007). Pre-
vious studies (z
HanG
et alii, 2004; H
übl
et alii, 2008)
reported that viscous debris flows recorded in the Ji-
nagjia Gully (China) have a frequency content of 6-10
Hz. In contrast, C
Hou
et alii, 2007 monitored stony
debris flows in Houyenchan (Taiwan) and reported
frequencies between 5-15 Hz and concluded that vis-
cous flows emit lower frequencies than stony flows.
The PS of the infrasound signal indicates that the
main frequency content from this debris flood was be-
tween 10 and 20 Hz. This differs from those results
seen at the Lattenbach torrent (peak frequency ca. 6
Hz) and those reported by k
oGelniG
et alii (in press)
for a previous event at the Illgraben (31.08.2008, peak
frequencies from 3 to 8 Hz). These results hint that
debris floods produce higher peak frequencies (10-20
Hz) than debris flows (< 10 Hz).
undular) point to a debris flood like event; the impulse
frequency of the geophone (operated by WSL, mounted
in the concrete of check dam 27) indicates only weak
activity at the flow front which could indicate that there
were not many boulders or just relatively small ones.
Without any visual information and given the evidence
mentioned above it can be assumed that this event was
a debris flood or an event that had a front like a debris
flood and a body like a debris flow. Hence in this paper
we will refer to this event as debris flood (according to
the classification of H
unGR
et alii, 2001).
The infrasound signal is shown in Fig. 5 and the
seismic signal in Fig. 6. From the ultrasonic gauges it
is known that the main surge of the debris flood passed
check dam 1 at 11:18:00 pm (accuracy of +/- 1 min due
to installation issues), check dam 10 at 11:21:00 pm and
check dam 27 at 11:39:42 pm. This corresponds to a flow
duration of 1122 s between dams 10 and 27, and given
that this is a known distance of 2656 m, the average flow
velocity in this section can be calculated as 2.3 m/s.
The RS of the infrasound signal shows the arrival of
the first debris flood signal at 11:28:49 pm (Fig. 7). There
is also an observable increase in amplitude in the time
series in this section. This occurs approximately 653
s before arrival at check dam 27. Assuming the above
calculated average speed of 2.3 m/s, this time point
Fig. 6 - RS (a), time series (b), flow depth (c) and PS (d) of the seismic signal during a debris flood on 28/07/09 in the Illgraben
torrent. In order to show only the debris flood frequency content a time window from 1.8-2.2*104s was chosen for the
computation of the PS. Seismic signal associated with a thunderstorm in the area are marked by the rectangle. The passing
of the debris flood at check dam 1 and check dam 10 is marked by the vertical lines in the flow depth graph. Geophone
signal in mV, sensor sensitivity 90V/m/s
background image
A. kOGELNIG, J. HÜBLE, E. SURIÑACH, I. VILAJOSANA, S. ZHANG, N. YUN & B. w. McARDELL
568
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
The RS of seismic data shows the arrival of the
first debris flood signal at 11:25:13 pm (Fig. 8), which
is 869 s before the debris flood passes check dam 27
and 216 s before the infrasound sensor detects the
event. Applying the above distance calculation (i.e.
assuming a constant flow velocity of 2.3 m/s) this cor-
responds to a distance of 2000 m (Fig. 4, B).
The peak frequency content in the seismic PS was
20 to 30 Hz (Fig. 6), which, similar to the infrasound
frequency content, was higher than that of the Latten-
bach torrent (seismic range 10-20 Hz).
GUXIANG GLACIER (CHINA)
The GuXiang Glacier is well known for its frequent
debris flow occurrences. The first sizeable event was in
1953 - the event had a peak discharge of 12600 m
3
/s and
a total volume of thirty million cubic metres. The flow
structure was a mixture of fine sediment, stones and
boulders. This event blocked the Podou Zhangpu River
and formed the lake as it is now (Fig. 9). The catchment
area is 24 km
2
and debris flows can be classified as vis-
cous. The infrasound monitoring unit DFW-I III (which
includes a microphone and a data-logger) was installed
Fig. 7 - Magnified sec-
tion of Fig. 5; the
infrasound sensor
detects the debris
flood ca. 377s be-
fore it passes the
sensor
Fig. 8 - Magnified sec-
tion of Fig. 6;
the geophone de-
tects the debris
flood ca 593s
before it passes
the sensor
background image
A STUDY OF INFRASONIC SIGNALS OF DEBRIS FLOWS
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
569
resulted in lower data resolution due to both signal at-
tenuation (an effect of distance source sensor and build-
ing interference) and increased background noise. The
infrasound signal over 180 s is shown in Fig. 10. Other
measurements for comparison to this data were not
available. Local witnesses provided anecdotal evidence
of event time and date. The RS of the infrasound shows
a constant signal in the frequency range from 5 to 10
Hz (Fig. 10), which is assumed to be associated with
the debris flow. This frequency range is also observable
in the PS. These results correspond to the infrasonic
data reported by z
HanG
et alii. (2004) and H
übl
et alii.
(2008) for viscous debris flows in the Jinagjia Gully
at this site. The sampling rate of the unit is 100 Hz. The
data-logger was developed in 2004 by the Institute of
Mountain Hazards and Environment, the CAS and the
Southwest Jiao Tong University. The microphone was
created by the Acoustics Institute at the Chinese Acad-
emy of Science (CAS) and is a further development of
the original device described in z
HanG
et alii (2004).
It has a frequency range of 3 to 200 Hz and a sensitiv-
ity of 50 mV/Pa. For safety and convenience reasons,
the equipment had to be placed in the cultural room of-
fice in the GuXiang village, approximately 5 km east
of the debris flow channel (Fig. 9). This setup location
is less preferable compared to the European sites and
Fig. 9 - Overview of the GuXiang Glacier - catchment
area, debris flow channel, Podou Zhangpu River
and neighbouring town with sensor location indi-
cated. Clearly observable is the lake formed by the
event in 1953 (Source: Google Earth)
Fig. 10 - RS (a), time series
(b) and PS (c) of
the infrasound sig-
nal during a debris
flow on 12/09/07
flow at GuXiang
Glacier starting at
01:30:12am. Infra-
sound signal in mV,
sensor sensitivity
50mV/Pa
background image
A. kOGELNIG, J. HÜBLE, E. SURIÑACH, I. VILAJOSANA, S. ZHANG, N. YUN & B. w. McARDELL
570
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
As with the GuXiang Glacier, the infrasound moni-
toring unit DFW-I III was used (frequency range of 3
to 200 Hz, sensitivity of 50 mV/Pa and sampling rate
of 100 Hz). For safety purposes, the equipment had to
be placed in the local travel office which is close to the
debris flow channel. Fig. 12 and Fig. 13 provide the in-
frasonic data recorded with the DFW-I III unit. As this
device was developed for warning purposes, recording
is initiated only if the amplitudes reach over a threshold
value (3 mV). Fig. 12 illustrates a 100 s window with
recordings that are related to debris flow activity in the
channel. Fig. 13 provides a 17 s window that shows one
debris flood surge (according to local witnesses). An
increase in amplitude is observable in the time series
as well as a change in the frequencies in the RS. More
interestingly, in the PS the main frequency content has
shifted to 10 to 20 Hz (similar to the Illgraben, Fig. 5)
in comparison to the frequency shown in the larger time
window (Fig. 12, 5-10 Hz). No firm confirmation can be
given due to a lack of supplementary data; it can only
be assumed that the frequencies reflect a difference in
flow characteristics (i.e. debris flood) of the single surge.
CONCLUSIONS
Infrasound monitoring of debris flows at different
locations in Europe and China are presented in this
study. The infrasound data could be correlated with
seismic recordings and flow height measurements for
the Lattenbach (Austria) and Illgraben (Switzerland)
torrents. In all cases, the infrasound device was able
to detect the event before passing the sensor location.
At the Lattenbach torrent, the infrasound sensor de-
tected the debris flow before the geophone (cut-off
frequency 10 Hz), whereas the opposite was seen at
the Illgraben torrent (geophone cut-off frequency 4.5
Hz). Further studies are required to clarify the relative
detection capabilities of these sensors.
Data analysis for the two sites in China was more
(China). There is no observable increase in amplitude
in the time series nor an increase in the frequency in
the RS (Fig. 10), as was the case for the Lattenbach
and Illgraben torrents. An increase in amplitudes and
frequencies in the infrasonic signal is observed when
a debris flow is moving toward the sensor, and the
highest values are seen when the flow passes the sen-
sor (k
oGelniG
et alii, in press). The absence of these
increases may be due to the source-sensor distance. The
placement indoors or the rheology of the flow could be
further explanations for the constant signal amplitude.
There are no expected differences due to the infrasound
microphone, as this same device was used at the Ill-
graben torrent and only the data-logger differs.
MIDUI GLACIER (CHINA)
The MiDui Glacier is one of the most famous gla-
ciers in Tibet. It is situated east of the GuXiang Gla-
cier, approximately 131 km upstream in the Podou
Zhangpu River, and has a catchment area of 123.8
km
2
. The channel has a N-S orientation and flows into
the south bank of the Podou Zhangpu River (Fig. 11).
The debris flows occurring here originate at the glacier.
The first event occurred in 1988, resulting from a gla-
cial lake outburst. The peak discharge was 1270 m
3
/s.
The river was blocked, the highway was destroyed and
downstream villages and cities were flooded.
Since 1988 several smaller viscous debris flows
have occurred almost yearly, but they did not reach
the monitoring point (Fig. 11, A).
Fig. 11 - Overview of the MiDui Glacier - catchment area (black) de-
bris flow channel and the Podou Zhangpu River. The distance
between the infrasound sensor (A) and the area of debris flow
origin (B) is 7.5km (Source: Google Earth)
background image
A STUDY OF INFRASONIC SIGNALS OF DEBRIS FLOWS
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
571
tion of infrasound and seismic sensors and an analy-
sis of the frequency evolution of the signal (RS) are
the most promising for monitoring torrential hazards.
Moreover, interfering noise in the signal arising from
a local thunderstorm are presented in the Illgraben
data. Variations in predominant infrasound and seis-
mic frequencies of over 15 Hz were seen between
study locations. It can be concluded that debris flows
emit infrasound signals with a lower frequency spec-
trum (<10 Hz) than debris floods (>10 Hz), and that
the frequency range is dependent on study site char-
acteristics, sensor location and process characteristics.
challenging and reference data were unavailable. Fur-
thermore, the DFW-I III is a warning device that initi-
ates recording only after the breach of a specific am-
plitude threshold and, as such, there is no knowledge
of signal patterns below this threshold. For further
studies at these two sites it is recommended to employ
a seismic sensor in addition to the infrasound sensor,
relocate the sensors to an outdoor location and imple-
ment a continuous recording scheme. These sites are
promising and the warning device is nevertheless a
powerful tool for debris flow alarming systems.
The preliminary results indicate that a combina-
Fig. 12 - RS (a), time se-
ries (b) and PS
(c) of the infra-
sound
signal
during a debris
flow on 10/08/09
flow at MiDui
Glacier starting
at 07:35:36am.
Infrasound signal
in mV, sensor sen-
sitivity 50mV/Pa
Fig. 13 - RS (a), time se-
ries (b) and PS
(c) of the infra-
sound signal of
a single surge
during a debris
flood on 05/09/08
at MiDui Gla-
cier starting at
09:07:38pm. In-
frasound signal in
mV, sensor sensi-
tivity 50mV/Pa
background image
A. kOGELNIG, J. HÜBLE, E. SURIÑACH, I. VILAJOSANA, S. ZHANG, N. YUN & B. w. McARDELL
572
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
REFERENCES
a
Rattano
M. (1999) - On the use of seismic detectors as monitoring and warning systems for debris flows. Natural Hazards and
Earth System Sciences ,20: 197-213.
b
adoux
a., G
Raf
C., R
HyneR
J., k
untneR
R. & m
C
a
Rdell
b.w. (2009) - A debris flow alarm system for the alpine Illgraben
catchment: design and performance. Natural Hazards and Earth System Sciences, 49: 517-539.
C
Hou
H.t., C
HeunG
y.l. & z
HanG
s. (2007) - Calibration of infrasound monitoring system and acoustic characteristics of
debris-flow movement by field studies. Debris-Flow Hazards Mitigation: Mechanics, Prediction, and Assessment, 571-580,
Amsterdam, Netherlands.
H
übl
J., z
HanG
s.C. & k
oGelniG
a. (2008)- Infrasound measurements of debris flow, In: d
e
w
RaCHien
, d., b
Rebbia
, C.a., l
enzi
,
m.a. (Eds.), wIT Transactions on Engineering Sciences, 60, Second International Conference on Monitoring, Simulation,
Prevention and Remediation of Dense and Debris Flows II, 17.-19. Juni 2008, The New Forest, 3-12; ISSN 1743-3533.
H
unGR
o., e
vans
s., b
ovis
m. & H
utCHinson
J.n. (2001) - A review of the classification of landslides of the flow type,
Environmental and Engineering Geoscience, 7(3):221-238, Alexandria, USA.
i
takuRa
y., i
naba
H. & s
awada
t. (2005) - A debris- flow monitoring devices and methods bibliography, Natural Hazards and
Earth System Sciences, 5: 971-977.
i
veRson
R. (1997) - The physics of debris flow. Reviews of Geophysics, 35: 245-296. Washington, USA.
k
oGelniG
a., H
übl
J., s
uRiñaCH
e., v
ilaJosana
i. & m
C
a
Rdell
b.w. (in press) - Infrasound produced by debris flow: Propagation
and frequency content evolution. Natural Hazards.
l
a
H
usen
R. (2005) - Debris-flow instrumentation. In: J
akob
m. & H
unGR
o. (2005 eds.) - Debris-flow Hazards and Related
Phenomena, 291-325, Springer Verlag Berlin, Germany in association with Praxis Publishing Ltd, Chichester, UK.
P
ilGeR
C. & b
ittneR
m. (2009) - Infrasound from tropospheric sources: Impact on mesopause temperature? Journal of
Atmospheric and Solar-Terrestrial Physics, 71: 816-822, Elsevier, Amsterdam, Netherlands.
z
HanG
s., H
onG
y. & y
u
b. (2004) - Detecting infrasound emissions of debris flow for warning purposes. Internationales
Symposium Interpraevent, VII: 359-364, Villach, Austria.
Statistics