Content uploaded by Sergey Chernomorets
Author content
All content in this area was uploaded by Sergey Chernomorets on Oct 16, 2018
Content may be subject to copyright.
61
Kriosfera Zemli, 2018, vol. XXII, No. 2, pp. 61–70 http://www.izdatgeo.ru
DOI: 10.21782/EC2541-9994-2018-2(61-70)
THE OUTBURST OF BASHKARA GLACIER LAKE (CENTRAL CAUCASUS, RUSSIA)
ON SEPTEMBER 1, 2017
S.S.Chernomorets1, D.A.Petrakov1, A.A.Aleynikov2, M.Y.Bekkiev3, K.S.Viskhadzhieva1,
M.D.Dokukin3, R.K.Kalov3, V.M.Kidyaeva1, V.V.Krylenko4, I.V.Krylenko1, I.N.Krylenko1,
E.P.Rets5, E.A.Savernyuk1, A.M.Smirnov1
1 Lomonosov Moscow State University, 1, Leninskie Gory, Moscow, 119991, Russia; devdorak@gmail.com
2 SCANEX R&D Centre, 1, Kiev highway, Business Park “Rumyantsevo”, 8 entrance, o ce 732, Moscow, 108811, Russia
3 High-Mountain Geophysical Institute, 2, Lenina prosp., Nalchik, Kabardino-Balkaria Republic, 360030, Russia
4 Shirshov Institute of Oceanology, RAS, 36,Nahimovskiy prosp., Moscow, 117997, Russia
5 Institute of Water Problems, RAS, 3, Gubkina str., Moscow, 119333, Russia
The outburst of Bashkara Lake on September1, 2017 refl ects the current stage of glacier downwasting
when the stable regime of lakes, debris complexes and glaciers is disturbed. We have estimated the parameters
of the lake outburst, as well as threshold and trigger conditions using ground-based and aerial observations,
satellite imagery analysis, and instrumental data obtained from a drone, an echo-sounder, an automatic weather
station, and a water level datalogger. An abnormal shower with 100mm of precipitation happened during a night
from August31 to September1 following the previous (August 30–31) 45mm rainfall, has been recognized as
the trigger of the lake outburst. The volume of the liquid phase of the debris fl ood was about 1.1mlnm3, includ-
ing water released during the lake outburst (0.8mlnm3). In the valley of the Adylsu River, the volume of
0.35–0.50mlnm3 of debris was entrained into the fl ow. Based on this research, we propose recommendations on
measures how to prevent emergencies which could take place in the area.
Glacial lake outburst fl ood (GLOF), debris fl ood, Caucasus
INTRODUCTION
At the night of September1, 2017, a disastrous
debris fl ood struck the valleys of the Adylsu River
and Baksan River in the Elbrus area (Kabardino-
Balkaria Republic, Russia). Three persons died. More
than 3.3km of the federal highway A-158 “Prokhlad-
ny–Baksan–Elbrus” and about 500m of the highway
in the valley of the Adylsu River were either fully de-
stroyed or signifi cantly damaged. Nearly 7750 local
residents and tourists turned out to be blocked in the
Elbrus area without transportation ties with the oth-
er part of the country, with the tourists’ evacuation
and delivery of essential goods exercised by way of a
helicopter. Gas supply was suspended for six settle-
ments – Terskol, Tegenekli, Elbrus, Neitrino, Baid-
ayevka, and Verkhny Baksan. There was no electrici-
ty and no telephone connection (neither regular tele-
phone nor mobile communication). Judging by
preliminary official data published in Rossiyskaya
Gazeta on September 8, the costs of the emergency
works amounted to 160million rubles, and the resto-
ration of the infrastructure required for ensuring sus-
tenance of the regions a ected by the disaster was
evaluated at 650million rubles [Myslivskaya, 2017].
The fl ood was caused by the outburst of glacial
Bashkara Lake (Fig.1). The coordinates of the out-
burst place are 43°12′35″N and 42°43′28″E. In the
late 1950s, the outbursts of Bashkara Lake resulted in
destructive debris fl ows in the valley of the Adylsu
River [Kovalev, 1961; Seynova and Zolotarev, 2001].
In those years, the body of the Bashkara Glacier was
the lake’s dam, and the outbursts moved along the
subglacial channels. After formation of a subglacial
(underground) drain channel with steady fl ow, the
lake’s outbursts stopped. The lake’s regime was stable
for about 40years.
Variations in the masses of the glaciers of the El-
brus region, including the Bashkara Glacier, in the
1980–1990s brought about changes in the moraine
landscapes and deformity of the existing and forma-
Copyright © 2018 S.S.Chernomorets, D.A.Petrakov, A.A.Aleynikov, M.Y.Bekkiev, K.S.Viskhadzhieva, M.D.Dokukin,
R.K.Kalov, V.M.Kidyaeva, V.V.Krylenko, I.V.Krylenko, I.N.Krylenko, E.P.Rets, E.A.Savernyuk, A.M.Smirnov,
All rights reserved.
Fig.1.Bashkara Lake and Lapa Lake in Central
Caucasus after the outburst of September1, 2017.
The quadrotor photo was provided by E.A.Savernyuk,
08.10.2017.
62
S.S.CHERNOMORETS ET AL.
tion of new stadial moraines [Dokukin and Savernyuk,
2012]. In the 1990s, the overland fl ow from Bashkara
Lake turned out to be blocked by a new lateral mo-
raine up to 7–8m tall, through which water began to
fl ow down fi ltration channels [Dokukin and Shagin,
2014]. That in many ways determined the specifi cs of
the lake’s regime for the subsequent years. At the be-
ginning of the 21st century, new lakes began to be
formed at the place where the snout of the Bashkara
Glacier was receding; they were named Lapa and
Mizinchik [Chernomorets et al., 2003]. At that time,
the volume and the area of Bashkara Lake depended
on the water level fl uctuations, but the growth of the
lakes’ volume downstream from the glacier with epi-
sodic rises in the level of Bashkara Lake indicated the
increasing threat of the debris fl ow [Chernomorets et
al., 2007а]. The increased probability of Lake Basha-
ra’s outburst was repeatedly written about, for exam-
ple, in [Zalikhanov et al., 2009; Petrakov et al., 2012].
In 2008, overfl ow of water from the lake to the
moraine dam into the grotto of the Bashkara Glacier
was fi rst recorded [Kidyaeva et al., 2013], with the
probability of an outburst estimated to be high [Pet-
rakov, 2008]. No outburst occurred, and a pavement
was formed at the place of the overfl ow, which pre-
vented further erosion of the river channel. The Chief
Emergency Department of the Kabardino-Balkaria
Republic carried out certain preventive actions,
aimed at informing the population and at monitoring
the lake’s condition. In particular, at the place of the
overfl ow of Bashkara Lake, a cut-through 1.8m deep
was made in 2009, resulting in reduction of the maxi-
mum volume of the lake by approximately 100000m3.
Based on the calculated hydrograph obtained by the
specialists of “Sevkavgiprovodhoz” [Gnezdilov et al.,
2007], the probable outburst flood from Bashkara
Lake was simulated [Petrakov et al., 2012].
In the period of 2009–2014, the level of Bash-
kara Lake was not high, and there was no overfl ow of
water over the moraine dam, but in 2015–2017, the
water level rose high in the lake, the overfl ow was re-
corded every year, and its duration extended from
one week in 2015 to two months in 2017. The fi eld
observations and the aerial survey revealed water fi l-
tration through the moraine dam. Thus, after a rela-
tively stable period of 2009–2014, the probability of
the lake’s outburst rose again.
The goal of this work was to analyze the causes
of the outburst of Bashkara Lake and to make pre-
liminary estimation of the volume of the outburst
fl ood.
MONITORING OF BASHKARA LAKES BEFORE
THE OUTBURST
The methods of monitoring the group of Bash-
kara lakes have been described in detail in [Petrakov
et al., 2012]. The bottom of Bashkara Lake, dammed
up by the fl ank moraine and the glacier, was stable,
which allowed, as a result of bathymetric survey
(2001, 2002, 2003, 2004, 2005, 2008, 2009, 2012),
evaluation of the relations between the water volume
in the lake and the water level of the lake. The bathy-
metric curves were used to estimate the volume of
water in the lake before the outburst.
Since 2013, to monitor the water levels of Bash-
kara Lake, scientists have used Keller DCX data log-
gers (Switzerland), measuring water levels by the pi-
ezometric method. One of the loggers was placed un-
derwater and logged the total atmospheric pressure
and the pressure of the water column above the log-
ger in time, while the second logger was placed on the
shore and measured the atmospheric pressure for
making an adjustment by its value when calculating
the water level, as well as the air temperature. The
data reading was conducted simultaneously at the
end of the season. As a result of the data loggers’ op-
eration, detailed behaviors of the water levels and the
water and air temperatures were recorded for the
fl ood-hazardous period, with the data saving rate in
the device memory being once an hour. These water
level data recorded with the Keller DCX data loggers
are in good agreement with the foot-gauge measure-
ments and the results of the preceding measurements
conducted with the ADU water level gauge.
Thus, with the same level of detail, the behavior
of the water levels and of the water and air tempera-
tures was determined for the fl ood-hazardous periods
of 2013–2017. The error rate in measuring the water
level with the data logger was 1–3 centimeters.
The meteorological data for the period preceding
the lake’s outburst were obtained using the Davis
Vantage Pro2 automatic meteorological station. The
station was placed on a subhorizontal site at the alti-
tude of 2650m and at the distance of 500m from
Bashkara Lake, on the stationary Dzhankuat Glacier
gauge station. The temperature gauge was at the
height of 1.8m above the glacier surface. The tem-
perature measurement error was 0.5°С, and that of
precipitation was 0.5mm. All the measurements were
conducted once every 15minutes.
STUDYING THE CONDITION
OF THE LAKES AND OF THE COURSE
OF THE OUTBURST FLOOD
To evaluate the situation before the outburst,
materials of the traverse studies conducted by
M.D. Dokukin and R.K. Kalov in June and by
V.V.Krylenko and A.A.Aleynikov in the second half
of August 2017 were used. Aerial survey and current
monitoring were carried out onboard a helicopter of
the Kabardino-Balkaria Department of the Russian
Ministry of Emergencies, with M.D. Dokukin,
R.K.Kalov, M.Y.Bekkiev and M.M.Khadzhiev par-
ticipating (September1 and 10, 2017). On Septem-
63
THE OUTBURST OF BASHKARA GLACIER LAKE (CENTRAL CAUCASUS, RUSSIA) ON SEPTEMBER 1, 2017
ber1–8, A.M.Smirnov, V.M.Kidyaeva, I.V.Krylenko
carried out the land studies of the depressions of
Bashkara Lake and of Lapa Lake, which included
tracing the lake contours, recording the shore line be-
fore and after the event and recording the runo pro-
cesses in the catchment of Bashkara Lake, and repeat-
ed photos of the locality from fi xed points. That al-
lowed the scientists to detect changes on the slopes
and in the river channel and to compare the situation
after the outburst with that of the previous years. In
addition, this team of researchers obtained informa-
tion about the discharge of water due to the GLOF
and the amount of the debris carried from the gauges
established in selected conditionally stable character-
istic points of the river channel. The maximum levels
of the fl ow and the value of the fall in the water level
of Bashkara Lake were evaluated by the high-water
benchmarks with Bushnell laser rangefinder. The
width and the depth of the gully were measured, and
the areas of the cross sections of the fl ow were deter-
mined. M.D.Dokukin and R.K.Kalov evaluated the
fl ood parameters in the downstream part of the fl ow
in the Adylsu Valley and the e ect of the shore-pro-
tecting structures on September15. Aerial survey of
the lakes and of the glacier-adjacent area was con-
ducted from the Phantom4 drone, and the bathymet-
ric survey of Bashkara Lake and of Lapa Lake, the
tracing of the external contours of the lakes and the
examination of the area along the fl ood course were
conducted by S.S.Chernomorets, E.A.Savernyuk,
M.D.Dokukin, A.V.Khatkutov on October7–9,
2017.
In addition, to analyze changes in the region of
the outburst fl ood, satellite and space station images
were used: from the Kanopus В1 satellite – on
22.08.2017, 19.09.2017; from the Sentinel 2A sa-
tellite– on 26.08.2017, 28.08.2017, 02.09.2017,
05.09.2017, 10.09.2017, 07.10.2017; and the images
taken by S.N.Ryazansky onboard the international
space station on 03.09.2017.
DISCUSSION OF THE CAUSES
OF THE LAKE’S OUTBURST
A high level of water at the end of the summer.
The mean long-term water level mark in the summer
period was about 2588m. As a rule, peak values of the
water level were recorded in the fi rst half of the sum-
mer [Kidyaeva et al., 2013], whereas by the end of
August, the water level decreased, according to the
readings of the water level gauge and to comparison
of the data of the aerial survey and of the fi eld obser-
vations (Fig.2). In total, before the outburst, the wa-
ter level in the lake had high marks not typical of the
end of summer (about 2591m); during 2.5months,
the water surface fl ow over the moraine dam was ob-
served; however, the water level in the lake, until the
middle of the day of August31, did not exceed the
values observed on the preceding days. Water over-
fl ow, which confi rmed the level gauge readings, was
recorded during the traverse measuring of June29,
July28 and August18.
Heavy and long-term precipitation. In the fi rst
half of the day of August31, 45mm of liquid precipi-
tation fell within 15hours (from 23:00 of August30
to 14:00 of August31). That caused a 33-cm rise on
the water level of the lake, and termination of pre-
cipitation in the second half of the day caused stabili-
zation and even certain decrease of the water level in
the lake. On August31 at 20.00, anomalously intense
rain, extremely rare in this region, started, according
to the observation data from the Dzhankuat Glacier
gauge station. In total, within 5hours (from 20:00
August31 to 01:00 September1), the liquid precipi-
tation according to the precipitation gauge was about
98mm, with three local peaks with the precipitation
intensity of about 40mm/h observed at 20:45, 22:00
and 24:00.
The anomalous shower resulted in the slope de-
bris flows into the valleys of the Dzhankuat and
Shkhelda Rivers, detected by comparing the Senti-
nel2A satellite images of August26 and September2
and recorded by the aerial survey of September1. In
the colluvial-proluvial deposits of a slope, a gully
about 800m long and 20–40m wide was formed. The
area of the deposits of the slope debris fl ooding the
pocket of the right flank moraine of the Shkhelda
Glacier amounted to more than 32thousandm2. The
debris fl ood deposits up to 100m wide were recorded
in the areas close to the glacier.
Fast rise of the water level before the outburst.
As the subsequent reading of data from the logger
showed, already in the fi rst hour of the anomalous
shower (from 20 to 21o’clock) the level mark in the
lake grew by 12cm, with the rise continuing in the
Fig.2.Variations in the water level in in Bashkara
Lake in the summer of 2017.
The horizontal lines are drawn every 1m. After the outburst of
01.09.2017. The water level dropped by 16.5m, but the level
gauge recorded data only until it reached the surface.
64
S.S.CHERNOMORETS ET AL.
lake until the outburst. In the lake, the water level
was 55cm higher on September1, compared to Au-
gust31 at 20:00, and was 578cm above the level
gauge. In the period of direct observations, such fast
growth of the water level had not been observed. By
the following observation benchmark at 3o’clock in
the morning, the gauge level was already above the
water level, i.e. at least 6-m discharge of water had
already taken place (Fig.2).
Evaluation of the volume of the debris fl ood.
As a result of the outburst, the moraine dam was
washed out, and by the morning of September1, the
water level had dropped by 16.5m compared to the
time before the outburst (Fig.3), and the volume of
water in the lake decreased by 3/4 (from 1millionm3
to 200thousandm3). A catastrophic fl ood streamed
down the Bashkara Glacier and further along the
Adylsu Valley, in some areas turning into a debris
fl ow. The roughly estimated infl ow of water in the pe-
riod of the anomalous shower (the lake’s catchment
basin is about 3.8km2, the total precipitation is
100mm, and the runo coe cient is 0.6) could reach
200thousandm3. At least the volume of 50thou-
sandm3 of water was drawn into the fl ood during the
outflow of water from Lapa Lake located near the
lower end of the Bashkara Glacier (the mirror area of
which before the fl ood was about 60thousandm2) as
a result of the catastrophic fl ood. This is indicated by
the reduction in the water level of Lapa Lake by ap-
proximately 80–100cm below the long-term water
level observed before the outburst fl ood. In the valley
of the Adylsu River, the additional volume of water
involved into the outburst fl ood consisted of the river
channel supply and of the lateral infl ows. In the peri-
od of the rainfall fl ood, before the outburst fl ood, the
hydromorphological characteristics of the Adylsu
River channel were taken in the fi rst approximation
to be the following: the mean width was 8–10m, the
mean depth, 1m, and the mean velocity, 2m/s. Con-
sidering the flow velocity in the river and the ad-
vancement velocity of the frontal wave of the out-
burst flood, the volume of not less than 50thou-
sandm3 was involved into the outburst fl ood due to
the river channel water supply. Considering the
short-term passage of the outburst fl ood, the authors
supposed that the contribution of the lateral infl ows
was compensated by the losses of the flow due to
sрreading-out in the fl ood plain and run-ups. In the
preliminary estimate, the volume of the water compo-
nent of the debris fl ood amounted to about 1.1mlnm3.
The volume of the debris component was evalu-
ated on the basis of the profi les made during the ob-
Fig.3.Bashkara Lake before and after the outburst.
a – the photo was provided by V.V.Krylenko, 19.08.2017; b –
the photo was provided by A.M.Smirnov, 01.09.2017; c – the
space image Kanopus V1, 17.09.2017 (the solid contour – the
shoreline on 17.09.2017, the dashed line – on 22.08.2017).
65
THE OUTBURST OF BASHKARA GLACIER LAKE (CENTRAL CAUCASUS, RUSSIA) ON SEPTEMBER 1, 2017
servation, analysis of photographs taken onboard the
helicopter during the fl yover immediately after the
outburst and of space photographs taken onboard the
international space station. The indicated lithody-
namic zones are shown in Fig.4.
According to the preliminary estimates, the vol-
ume of the debris involved into the outburst fl ood
was 350–500thousandm3. The major mass of the de-
bris was delivered into the fl ow due to lateral erosion
in section 11 downstream from the bridge across the
Adylsu River (200–300thousandm3, Fig.4,j) and in
section9 downstream from the Dzhantugan sports
base (100–150thousandm3, Fig.4,g,i). Rather short
sections of involvement of the solid material with
prevalent bottom erosion section1 (a cut, 20thou-
sandm3, Fig.4,b,c), section4 (the frontal terrace of
the moraine downstream from Lapa Lake, 20thou-
sandm3, Fig.4,d) and section6 (the canyon covered
by moraines of the Little Ice Age, 20thousandm3,
Fig.4,e) are located in the upstream part of the val-
ley. The erosion sections are followed by the sections
of debris accumulation or its transport and subse-
quent accumulation (Fig.4,f,h,k,l).
EVALUATION OF THE VELOCITY
AND OF THE RATE OF DISCHARGE
OF THE OUTBURST FLOOD
According to the witnesses’ evidence, the frontal
wave of the flood reached the Shkhelda Gorge
(6.4km downstream from the source in Bashkara
Lake) at about 01:20a.m. and the settlement of El-
brus (10km downstream from the source), no later
than at 01:30a.m. by 02:30a.m. on September 2017,
the discharge of water in the Adylsu River signifi -
cantly decreased, as the loud sounds of crashing
stopped). According to those fi gures, the advance-
ment velocity of the frontal wave was 5–6m/s.
The fl ood reached the Baksan River and passed
several dozens of kilometers along it, causing destruc-
tion of roads, bridges, aqueducts, the gas pipeline and
the electric power lines at certain places. Due to the
passage of the outburst fl ood and due to the accompa-
nying Baksan River channel deformities in the town
of Tyrnyauz 40km downstream from the lake, the
crest of the debris fl ow dam on August 14–15, 2017
from the Gerkhozhan River was washed out by 2.0–
2.2m, and the level of the temporary dam pool, which
existed from mid-August 2017, significantly de-
creased.
Judging by the traces left on the Bashara Glacier
and in the upper part of the Adylsu Valley, the maxi-
mum outburst rate of water discharge could have
amounted to about 500m3/s, while the rates of water
discharge in the sections of local transformation into
the debris fl ow 1.5–4.0km downstream from the lake
could have been even greater. For example, in the
area of the bridge across the Adylsu River, the traces
of the fl ood passage were recorded at the height of
7–8m above the river channel. Considering the size
of the bridge and the fl ow velocity of about 5m/s, the
rate of water discharge could have reached 800m3/s.
Downstream in the Adylsu Valley, the maximum rate
of water discharge gradually decreased, and near the
exit to the Baksan Valley, it was approximately 200–
250m3/s.
Continuing reduction of the Bashara Glacier.
The surface of the Bashara Glacier, which propped up
the dam in 2008, signifi cantly decreased (Fig.5). Re-
treat of the right branch of the glacier adjacent to
Bashkara Lake on the front from 2009 to 2017 was
30–35m, and that of the main fl ow of Bashkara Gla-
cier adjacent to Lapa Lake from 2007 to 2017 was
100–105m.
Decrease of the surface of Bashkara Glacier con-
tributed to stability of the moraine dam. Whereas in
2008 it was backed by the glacier, this support has
recently disappeared.
The probable mechanism of the outburst. Due
to the infl ow of a large volume of water as a result of
anomalous precipitation, the fl ood with increased dis-
charge of water moved across the moraine dam. The
following factors contributed to that: 1)temporary
rise of the water level in the lake by 55cm due to cov-
er of the surface runo zone by the landslide masses
coming from the moraine crest and its fast washout
(the discharge of water could increase to several do-
zens of cubic meters per second – water was dis-
charged in the area wider than before); 2)increase of
the erosive ability of the fl ow as a result of its satura-
tion with loose landslide masses having a large
amount of fi ne soil.
Amid the abrupt rise of the water discharge over
the ridge to 6–10m3/s, seepage through the moraine
dam signifi cantly increased. The increase of the hyd-
rodynamic water head, together with reduced stabil-
ity of the dam due to its inundation, could have
caused violation of the dam pool’s integrity, its shift
and nearly instant crushing. Destruction of the mo-
raine dam, which kept the lake, presumably occurred
in two stages. At the fi rst, very short-term, stage, the
dam moved (crushed). At the second, longer-term,
stage, after reduction of the water level in the lake by
8–10m, a new narrow cut was formed due to bottom
and lateral erosion. At this stage, the lake level
dropped by 6–7more meters.
Traces of an instant splash of water onto the gla-
cier below (Fig.1, 4,d) and the absence of any traces
of terraces or of the stay of the interim water levels in
the lake, which is observed in erosional destruction of
the moraine dam (for example, [Chernomorets et al.,
2007b]) suggest the high degree of probability of such
a scenario. The trapeziform confi guration of the upper
part of the cut, where the main volume of the lake
water was discharged, also supports this hypothesis.
As a result, after fast filling of the funnel near the
66
S.S.CHERNOMORETS ET AL.
67
THE OUTBURST OF BASHKARA GLACIER LAKE (CENTRAL CAUCASUS, RUSSIA) ON SEPTEMBER 1, 2017
Fig.4.The disaster zone in the valley of the Adylsu River
before the infl ux into the Baksan River.
а – the prevailing types of the river channel processes in the valley of the
Adylsu River at passage of the debris fl ood on 01.09.2017; 1 – accumula-
tion of sediments; 2 – benthic and lateral erosion; 3 – sediment transport;
4 – the lake contour before the outburst on 22.08.2017; 5 – the lake
contour after the outburst as on 03.09.2017. Lithodynamic zones: 1 – cut;
2 – Bashkara Glacier; 3 – the debris fl ow fan; 4 – the massif of terminal
moraines with buried ice and of dead ice; 5 – sandur; 6 – a gully cut into
moraines of the Little Ice Age; 7, 8 – the gorge of the Adylsu River above
the Dzhantugan training base; 9 – the territory of the Dzhantugan
mountaineering camp and the cut slope; 10 – the river channel down-
stream of the Dzhantugan training base; 11 – a bridge across the Adylsu
River and the river channel downstream; 12 – the valley of the Adylsu
River before its outcoming into the gorge of the Baksan River and the
mouth of the Adylsu River.
The diagram was made by V.M.Kidyaeva. The base: a photo from the International Space Station (the photo was provided by
S.N.Ryazansky, 03.09.2017). b–l – photo illustrations for the diagram (the numbers of the lithodynamic zones are provided in red
in the photos). The photos were provided by: b – I.V.Krylenko, 03.09.2017; c – V.M.Kidyaeva, 03.09.2017; d –j, l – M.D.Dokukin,
01.09.2017; k – S.S.Chernomorets, 07.10.2017.
Fig.5.Part of a morainic dam, Bashkara Lake, in 2008 and in 2017.
a – 16.07.2008; b – 28.07.2017. Lowering of the surface of Bashkara Glacier by more than 10m. The red arrow illustrates the change
in the position of rocks in relation to the maximum water level (b) compared to the level of 2008 (a). The photo was provided by
M.D.Dokukin.
68
S.S.CHERNOMORETS ET AL.
grotto (about 17–20m deep in relation to the glacier
margin), the water masses splashed onto the surface
of the Bashkara Glacier. A large water wave passed,
probably fast, on its surface.
The condition of the Bashkara lake and the
glacier system after the outburst. The results of the
bathimetric survey of Bashkara Lake and of Lapa
Lake are shown in Тable1.
Bashkara Lake. Water flows from the lake
across the crest of the dam (the crest mark is about
2574m), the length of the dam is about 70m, below
the slope essentially increases, and the fl ow plunges
into the subglacial grotto, with the area of the en-
trance cross section being 10–15m2. Within the dam
water fl ows along a gully with the bottom 3–5m wide
and subvertical walls 15–20m tall formed by the mo-
raine (Fig.4,b,c). Falling o of large lumps of ground
from the walls of the cut (new material up to 30m3
was found) is hazardous, as it may cause rise of back-
water in the lake, followed by washout of the dam
pool. The possible rise of the water level may be deci-
meters–the fi rst meters, the outburst fl ood during the
washout would be insignifi cant and is likely to fully
pass through the subglacial/intraglacial fl ow channel,
performing the role of a regulator.
The area between the lakes. The width of the
splash-in was about 50m at the exist to the glacier
below the grotto; then the fl ow widened to more than
100m and divided into two branches (Fig.4,d,
zone6). The main left fl ow streamed to the depres-
sion in the central part of the glacier, while the right
fl ow passed as a relatively narrow stream 30–40m
wide closer to the right fl ank moraine and fell from a
steep (approximately 30m) cli over the grotto, into
which water fl ew from Bashkara Lake before the out-
burst. Due to a very high speed, the left fl ow “skipped”
the part of the glacier with opposite elevation, rising
to the height of more than 5–7m, and reached the
depression in the intensely fractured central part of
the glacier, where it made a deep (up to 15m) and
wide hole, in which (perhaps, already at the end of
Table 1. Parameters of Bashkara Lake
and Lapa Lake after the outburst of September 1, 2017
(as of October8, 2017)
Lake Area, thou-
sand km2Volume, thou-
sand km3Max
depth, m Average
depth, m
Bashkara 26 216 18 8
Lapa 43 218 12 5
Fig.6.Bathymetric maps of Bashkara Lake (right) and Lapa Lake (left) after the outburst.
The lake’s contour: 1 – according to the space image Kanopus V1 as of 22.08.2017; 2 – according to the quadrotor survey of
08.10.2017. The base – a space image from Sentinel-2 as of 07.10.2017, the orthophoto is made of quadrotor images of 08.10.2017.
The quadrotor survey and the bathymetric maps – E.A.Savernyuk, the orthophoto – A.I.Yarkova, E.A.Baldina.
69
THE OUTBURST OF BASHKARA GLACIER LAKE (CENTRAL CAUCASUS, RUSSIA) ON SEPTEMBER 1, 2017
the fl ood) large blocks of ice crashed down. On the
steep edge scarp, the left fl ow gushed into Lapa Lake
widely. It is likely that at the end of the fl ood the fl ow
of water made a kind of a “funnel” several dozens of
meters long and more than 20 m wide at the output in
the lower part of the scarp.
Lapa Lake. The dam pond of Lapa Lake was
formed by several terminal moraines containing bur-
ied ice. After the outburst, the lake water level re-
mained to be approximately 1m lower than the level
in the pre-fl ood period, was well recorded by abrasion
scarps on sandy beaches and by the thermal abrasion
niche in the vertical cli s of the ice banks. The over-
fl ow started as a wide fl at fl ow of water, 60m below
there are steep spillways 3–4m high. Below the spill-
ways, the branches cut into a mass of dead ice
60×100m in size, denudated by the outburst fl ood,
to the depth of 3–5m (Fig.4).
As at October8, 2017, the level of Lapa Lake de-
creased compared to the marks before the outburst by
approximately 1.8m. It is possible to forecast its re-
duction due to regressive erosion of the channel in
the ice, along which the water spills.
In any case, erosion of the dam will take place in
the shallow part of the lake (Fig.6), and the possible
discharge of water from the lake, without taking the
incoming water into account, will not exceed
200thousandm3.
THE DEBRIS FLOOD ON SEPTEMBER 7
IN THE ADYLSU VALLEY
Comparison of the aerial survey materials as of
September1–10, 2017 revealed the fact of a debris
fl ood that occurred on September7. There were re-
ports of the local residents of the partial erosion of the
restored road near the settlement of Neitrino on Sep-
tember7. Fig.7 demonstrates a part of the Adylsu
Valley in the area of the Elbrus training center after
the outburst of Bashkara Lake on September1 and
onSeptember10, after the debris flow. The debris
1.5–2.0m thick got deposited, and the flow went
above the protection wall onto the territory of the El-
brus mountaineering camp. The presence of numer-
ous erosions and deposits of loose material in the
channel of the Adylsu River, which emerged after the
outburst of Bashkara Lake on September1, 2017, as
well as intense liquid precipitation, a little over
70mm during 15hours, contributed to the debris
fl ow.
CONCLUSIONS AND RECOMMENDATIONS
The outburst resulted in fi vefold reduction of the
volume of water in Bashkara Lake. Despite this, there
exists the probability of a repeated outburst. The gul-
ly formed by the outburst is narrow and deep
(Fig.4,b,c). In the future, it is possible that the nar-
row cut in the moraine crest of Bashkara Lake may be
overlapped as a result of a landslide to form a tempo-
rary dam up to 5m high, resulting in the rise of the
water level in the lake and in the further outburst of
the formed dam.
In 2018, it is required that continuous control
should be exercised over the condition of the dam be-
tween Lapa Lake and Bashkara Lake, and observation
of their water level regimes should continue.
It is necessary to stop senseless and unprofes-
sional attempts to reduce the water level in the lakes
without examining all the related aspects. The global
experience [Liboutry et al., 1977] indicates that such
attempts may cause a disaster. A project of controlled
discharge of water from the lakes is required. In par-
ticular, for Bashkara Lake, it is recommended to
broaden the bottom of the evacuation channel to
10–15m and to reduce the steepness of the adjacent
slopes to 15–20°, to lay pipes, with subsequent reim-
bursement and making an emergency water dis-
charge, as well as to use syphon pumping of the lake
water annually in the period from the end of May to
Fig.7.The debris-fl ow-proof wall on the territory of the Elbrus training center before (a) and after (b) the
debris fl ood of September7 (the photo was provided by M.D.Dokukin).
70
S.S.CHERNOMORETS ET AL.
September. The measures to be taken in relation to
Lapa Lake require additional work (for example, it is
possible to apply hydromonitors for washing out and
broadening the ice dam).
The outburst of Bashkara Lake created new con-
ditions for forming debris fl ows along the channel of
the Adylsu River. There have appeared parts of un-
stable steep slopes composed of loose material. In the
case of new fl ows fl owing along the river channels or
fl oods, mobilization of this material will be much eas-
ier than in the situation of September1, 2017. There-
fore, even if the water impulse is not so large, addition
of solid material to the new fl ow may prove to be sig-
nifi cant and result in new destructions. Events simi-
lar to the debris fl ood of September7, 2017 will be
repeated on a larger scale.
Parts of the lake shores protected by concrete
walls withstood the destructive impact of the fl ood
and of the debris fl ood (Fig.7), implying the high ef-
fectiveness of these structures. In the future, rein-
forcement of the shores with concrete walls will con-
tribute to minimization of damage due to possible
recurrence of fl oods and debris fl ows. Destroyed and
cut part of roads are located at the turns of the chan-
nel of the Adylsu River (Fig.4,j), with the parts of
the roads on which large rocks appeared due to the
debris flood of September 1 appearing to be most
problematic.
The authors present their thanks to A.V.Khankutov,
R.G.Miskarova, V.V.Surkov for participation in the
collection of materials; to E.A.Baldina and A.I.Yarkova
for processing of the drone images; to L.V.Desinov,
U.D. Kurdanov, V.V. Borisov, B.H.Khokhanaev,
K.H.Zalikhanov, E.V.Zaporozhchenko, M.M.Khadzhiev
for valuable discussions, to the Chief Department of
Russia (Kabardino-Balkaria Republic), V.V.Popovnin,
A.H.Adzhiev for logistical support. The authors are
thankful to the specialists of Planeta Research Center
A.A.Nevsky, N.I.Abrosimov and V.V.Asmus for the
Kanopus-V1 satellite images.
The study has been carried out with partial fi nan-
cial support of the Russian Science Foundation (proj-
ects 15-05-08694, 16-35-60042, 18-05-00520).
References
Chernomorets,S.S., Petrakov,D.A., Krylenko, I.V., Kry len-
ko,I.N., Tutubalina,O.V., Aleynikov,A.A., Tarbeeva,A.M.,
2007a. Changes of the Bashkara glacier-lake system and as-
sessment of debris fl oodhazard in the Adyl-Su river valley
(Caucasus). Kriosfera Zemli XI(1), 72–84.
Chernomorets,S.S., Petrakov,D.A., Tutubalina,O.V., et al.,
2007b. The outburst of a glacier lake on the north-eastern
slope of Mount Elbrus on August11, 2006: the forecast,
event and after-e ects. Materialy Glatsioloicheskikh Issle-
dovaniy, issue102, 211–215.
Chernomorets,S.S., Tutubalina,O.V., Aleinikov,A.A., 2003. New
hazardous lakes near the edge of Bashkara Glacier in Central
Caucasus. Materialy Glatsiologicheskikh Issledovaniy95,
153–160.
Dokukin,M.D., Savernyuk,E.A., 2012. The advance of glaciers
at the end of the 20th century as a factor of activation of
glacial debris fl ows (Central Caucasus), in: Debris fl ows:
disasters, risk, forecast, protection. Proceedings of the Inter-
national Conference dedicated to the centenary of Professor
S.M.Fleishman. Ed. by S.S.Chernomorets. MSU, Moscow,
pp.31–32. (in Russian)
Dokukin,M.D., Shagin,S.N., 2014. Features of dynamics of
glacial lakes with underground drain channels (analysis of
multi-temporal aerospace information). Earth’s Cryosphere
XVIII(2), 41–50.
Gnezdilov,Y.A., Ivashchenko,E.N., Krasnykh,N.Y., 2007. Esti-
mation of the hypothetical outburst of Bashkara Lake, in:
Proceedings of the North Caucasian Institute for Water-
works and Ameliorative Construction (Sevkavgiprovodhoz).
Pya tigorsk, issue17, pp.123–145. (in Russian)
Kidyaeva,V.M., Krylenko,I.N., Krrylenko,I.V., et al., 2013.
Fluctuations in the water levels of the glacier lakes of the
Mount Elbrus region. Georisk, No.3, 8–15.
Kovalev,N.V., 1961. On debris fl ows on the northern slope of
Central Caucasus. Proceedings of the Caucasian Expedition
(by the program of the International Geophysical Year).
Kharkov University Press, Kharkov, vol.III, pp.149–161.
Liboutry,L., Arnao,B.M., Pautre,A., Schneider,B., 1977. Gla-
ciological problems set by the control of dangerous lakes in
Cordillera Blanca, Peru. I.Historical failures of morainic
dams, their causes and prevention. Journal of Glaciology 18,
239–254.
Myslivskaya,G., 2017. A disaster may be controlled. Valentina
Matviyenko appreciates the work conducted by the autho-
rities of Kabardino-Balkaria Republic to overcome the e ects
of a debris fl ow. Rossiyskaya Gazeta, September8, 2017. –
URL: https://rg.ru/2017/09/08/reg-skfo/matvienko-vlasti-
kbr-horosho-spravilis-s-posledstviiami-selia.html.
Petrakov,D.A., 2008. The debris fl ood hazard of glacier lakes
and assessment of the probability of their outburst, in: Debris
fl ows: disasters, risk, forecast, protection. Proceedings of the
International Conference. Pyatigorsk, Russia, September
22–29, 2008. Chernomorets,S.S. (Ed.). Sevkavgiprovodhoz,
Pyatigorsk, pp. 309–312.
Petrakov,D.A., Tutubalina,O.V., Aleinikov,A.A., Cherno-
morets,S.S., Evans,S.G., Kidyaeva, V.M., Krylenko,I.N.,
Norin,S.V., Shakhmina,M.S., Seynova,I.B., 2012. Monitor-
ing of Bashkara glacier lakes (Central Caucasus, Russia) and
modelling of their potential outburst. Natural Hazards,
61(3), 1293–1316.
Seynova,I.B., Zolotarev,E.A., 2001. The Glaciers and the Debris
Flows of the Elbrus Region. Nauchny Mir, Moscow, 204pp.
(in Russian)
Zalikhanov,M.Ch., Anakhaev,K.N., Nedugov,A.N., 2009. On
mudfl ow hazardous Bashkara Lake. Meteorologia i Gidrolo-
gia, No.2, 89–92.
Received October 31, 2017
