Document Actions

IJEGE-11_BS-Seinova-et-alii

background image
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
1003
DOI: 10.4408/IJEGE.2011-03.B-109
REGIONAL SHORT–TERM FORECAST OF DEBRIS FLOW INITIATION
FOR GLACIATED HIGH MOUNTAIN ZONE OF THE CAUCASUS
i.b. seinova
(*)
, y.b. andReev
(**)
, i.n.kRylenko
(**)
& s.s.CHeRnomoRets
(*),(**)
(*)
University Centre for Engineering Geodynamics and Monitoring, Moscow
(**)
M.V. Lomonosov Moscow State University, Faculty of Geography, 119991, Moscow, Leninskie Gory, 1
E-mail: krylenko_i@mail.ru
also far from it. Through the channels of the big rivers
disastrous torrents can reach foothill plains with the de-
veloped agriculture and urbanization. The local experi-
ence of constructing debris flow protection structures
does not provide for sufficient safety of the population;
ABSTRACT
We present a methodology for generalised region-
al-scale forecast of debris flow activity for territories
with uniform conditions of debris flows formation. In
the high mountain region of the Central Caucasus the
trend of activation of debris flows processes is con-
nected with the degradation of glaciers. For the de-
velopment of debris flows forecast combined physi-
cal–statistical models based on long-term observation
series are shown to be optimal. The authors develop
an original and simple method of debris flows forecast
for high mountains region with glaciation. The meth-
od is based on predictors readily available in practice,
namely daily temperatures and precipitation. The
method of compilation of a short- term debris flows
forecast for high mountain zone of the Central Cauca-
sus is proposed and recommendations for its practical
usage are presented.
K
ey
words
: forecast, debris flows, physical-statistical model-
ling, Central Caucasus
INTRODUCTION
The high mountainous zone of the Caucasus
Mountains is well developed (Fig. 1) and is character-
ised by the presence of abundant infrastructure (towns,
resorts and communication routes). Significant part of
this infrastructure is subjected to risks, typical of moun-
tains regions. Among them debris flows cause appreci-
able damage not only in the zone of their formation, but
Fig. 1 - Map of debris flow channels for study area. Le-
gend: 1 – river channels with traces of debris
flows, 2 – rivers with no evidence of debris flows,
3 – boundary of Main Caucasus Range (and state
boundary), 4 – towns and settlements, 5 - glaciers,
6 – mountain summits, 7 – areas of temporary stay
for large numbers of people (ski resorts, mountai-
neering camps etc), 8 - area, shown on the figure
2 (see below) (Compiled by kapitanov A.N., Cher-
nomorets S.S. and Tutubalina O.V. on the base of
Landsat ETM+ satellite imagery of 2000-2001,
aerial imagery of 2004, topographic maps at sca-
les of 1:25000 to 1:200000 (various years), as
well as field data of Seinova IB., Dokukin M.D.
and Chernomorets S.S.)
background image
I.B. SEINOVA, Y.B. ANDREEV, I.N. kRYLENkO & S.S. CHERNOMORETS
1004
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
results we had to develop a more detailed physical-
statistical method.
The technique of forecasting uses information
about physico-geographical, geological, hydro-mete-
orological conditions of debris flows formation with
the purpose to find out the set of factors, both neces-
sary and sufficient for debris flows initiation.
The process of development of such techniques
consists of three stages: 1 - retrospection, 2 - diagnosis
and 3 - prospection (t
alanov
, 2007).
At the stage of retrospection the collection and
analysis of data about past debris flows are carried
out to reveal regional and local conditions of their
formation and to estimate their parameters quantita-
tively. As a result, catalogues, maps and databases
of debris flow basins are made and spatial-temporal
analysis of debris flow hazard for the area is carried
out (Fig. 2). At the stage of diagnosis deterministic
analysis of observation data is carried out. The first
step of diagnosis is determining the range of the val-
ues of meteorological parameters, which are neces-
sary and sufficient for initiation and development of
debris flow processes. The choice of the algorithm
for modelling depends on the forecast time inter-
val chosen (long-term and short-term) and research
area. The deterministic methods are used for local
areas from a specific debris flow initiation zone to a
river basin. For large regions, including several river
basins with homogeneous conditions of debris flow
formation, physical-statistical methods are more ap-
plicable. The final step at the stage of diagnosis is
the drawing up of recommendations, containing al-
gorithms and technology of forecast issues.
The stage of prospection includes experimental
forecasts and evaluation of the forecast success rate
by comparison of modeling results with data on ac-
tual debris flows.
In this work we adopt a physical-statistical ap-
proach for the short-term regional forecast, in which the
development of mathematical algorithms is preceded
by the analysis of the physical nature of debris flows.
RETROSPECTIVE REVIEw OF CONDITIONS
AND FACTORS FOR DEBRIS FLOwS FORMA-
TION
The development and validation of a short-term
regional debris flows forecast was based on a 58 year -
long observation series (1951–2009) in the high moun-
therefore the problem of the forecast of debris flow situ-
ations is always important.
The forecast method, developed by the authors,
is based on the 6-day cumulative sums of daily tem-
peratures and precipitation and on forecasted values
of air temperature and precipitation for the next day.
The method has been successfully applied in Hydrom-
eteorological Center of the Republic of Kabardino-
Balkaria (Russian Federation).
THE FORECASTING METHODOLOGY
When complex glacial and rainfall-induced de-
bris flows are formed it is necessary to consider the
simultaneous influence of temperature and precipita-
tion (s
einova
& z
olotaRev
, 2001). The temperature
regulates the processes of ice and a frozen-ground
thawing, thus reducing internal cohesion and static
friction in glacial moraine deposits, where the debris
flows originate. Precipitation increases the weight of
moraine masses. It also reduces cohesive forces in
moraine due to washout of a material. Thus, the me-
teorological conditions define the state of debris flow
material and the time of debris flow initiation.
But the forecast success rate of our method,
based only on the critical values of meteorological
parameters for debris flow triggering, was 53% (a
n
-
dReev
& s
einova
, 1984). Such low forecast accuracy
testifies only to the presence of unstable potential de-
bris flow material on a slope. To improve forecasting
Fig. 2 - The map of debris flow basins for the Baksan Riv-
er tributaries. Legend: 1-7 –maximum volume of
debris flow deposits for one event (1000m
3
): 1. >
5000, 2. 5000 - 1000, 3. 1000 - 500, 4. 500 – 200,
5. 200 – 100, 6. 100 – 50, 7. 50 – 10; 8 – gla-
ciers, 9 – rivers, 10 –settlements, 11 – catalogue
numbers of debris flow basins, 12 –boundaries
of debris flow basins, 13 – boundary of the Main
Caucasus Range
background image
REGIONAL SHORT–TERM FORECAST OF DEBRIS FLOW INITIATION FOR GLACIATED HIGH MOUNTAIN ZONE OF THE CAUCASUS
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
1005
consolidated rock material (Figure 3). Tectonically ac-
tive high-mountainous relief with glaciation creates fa-
vorable preconditions for debris flow origination with a
high energy potential. The long-term geological factors
define the development of the abrupt bare slopes and
of unconsolidated rock material. The contribution of
glacial-nival factor leads to formation of a contrast re-
lief of the alpine type with the sharp peaks carlings and
grandiose walls. The intensive weathering of rocks of
high durability, such as granites, grandiorites, gneisses
and crystalline slates, is highly developed. This process
is considerably accelerated at the heights of 3100-3200
m above sea level, along the so-called “weariness line”
of the rocks, caused by seasonal thawing and freezing
of snow-ice cover (G
oRsHkov
, 1982). These areas are
characterised by constant rockfalls and transport of
denudated material to the bottoms of glacial cirques. At
the elevations from 2500 m (the regional level of valley
glacier terminuses) up to 3500 m (the regional border
of cirque glaciers) the potential debris flows source
area represents an interconnected system of moraine,
fluvioglacial and gravitational deposits, consolidated
by permafrost, with layers and lenses of buried stag-
nant ice, with marginal and thermokarst lakes and rock
glaciers. Here the potential debris flows source areas
are constantly replenished by solid, liquid and ice ma-
terials. Thermo-erosive processes of destruction of fri-
able unconsolidated rocks of the periglacial zone under
the influence of hydrothermal factors are the principal
cause of the formation of catastrophic debris flows in
the high mountains of the Central Caucasus.
THE CHARACTERISTICS OF THE DEBRIS
FLOw FOR THE PERIOD OF OBSERVATIONS
At the boundary of glaciosphere, where debris
flows originate, the nature system is unstable. In such
conditions insignificant deviations from stable trend of
weather and mass balance of glaciers could lead to sig-
nificant debris flow activity.
We have analysed the heterogeneity of debris flow
events in the time and space, the differences in their
scale and in the triggering meteorological conditions
during the period of observations. For the period from
1951 to 2009 there were 38 situations of high debris
flow activity in the Baksan River basin, during which
414 debris flow events occurred. The distribution of
their quantity from year to year was unstable - from
single cases to mass events on numerous tributaries
tain region of the Central Caucasus. Available data
were collected and systematized by authors (s
einova
&
t
atyan
, 1977; a
ndReev
& s
einova
, 1984; s
einova
&
z
olotaRev
, 2001). Our observations were carried out
in the upper stream of the Baksan River basin and in
its major tributaries Chegem and Cherek, with a total
area of 2100 km
2
, including 112 debris flow basins.
The tributaries of the main river collect their waters
on the north slope of the Main Caucasus Range with
maximum elevations over 5000 m. The glaciers cover
an area of 550 km
2
. The 75 tributaries with the highest
debris flow hazard originate from glaciers. The Baksan
River originates from glaciers of the highest mountain
of Caucasus, the Elbrus volcano (5,642 m) and flows
into the Terek River, which runs into the Caspian Sea.
At the first stage of the debris flows forecast the
assessment of deposits, available for entrainment in de-
bris flows is carried out.
GEOLOGICAL AND GLACIAL PRECONDI-
TIONS OF POTENTIAL DEBRIS FLOw MASSES
FORMATION IN THE DEBRIS FLOwS ORIGI-
NATON SITES
In the conditions of glacier degradation the margin-
al moraine complexes of retreating glaciers are sources
of material for potential debris flows. The debris flow
source area is a morphological formation with sufficient
slope, capable to concentrate runoff and containing un-
Fig 3 - Debris flow initiation zone at the marginal mo-
raine of kyaarty glacier, 2000 (helicopter photo
by Valeriy Perekrest)
background image
I.B. SEINOVA, Y.B. ANDREEV, I.N. kRYLENkO & S.S. CHERNOMORETS
1006
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
when the level of snow line was above 4000 m, at 4 am
on 19 July a rainfall of 33.7 mm occurred and led to
mass debris flow releases.
The local debris flow releases (in 10-20 basins si-
multaneously) were observed more frequently and oc-
curred after showers with amount from 20 to 52 mm.
Such situations were observed in 1970, 1972, 1979,
1980, 1984, 1985, 1986, 1994, 1995, 2001, 2003 and
2005.
For the 400 cases of precipitation from 20 mm
(annually repeated) to 86 mm (1% probability) only
38 debris flow events were registered, so their prob-
ability is less than 10%. The observation data do not
show a clear dependence between the magnitude of
debris flow events and quantity of precipitation. Part
of the reason is in different phase states of precipita-
tion (liquid, solid). In high mountains precipitation is
accompanied by sharp decrease of air temperature. As
a result of summer snowfall the glaciers are covered
by snow and thaw runoff decreases. The role of rain-
fall is 14% from the runoff of glacial-nival zone of the
Caucasus (w
ateR
R
esouRCes
, 1973). The possibility
of debris flow initiation during rainfall also depends
on air temperature. There is not enough time for thaw-
ing of frozen masses in a cold summer season and
after extremely cold winters, as a result debris flows
processes do not developed. Such situation was ob-
served at the Baksan River basin in 1985-1995. Dur-
ing this decade in high mountains debris flows were
not observed, while at lower heights in 1986–1987
catastrophic debris flows did occur.
In all observed situations with various combina-
tions of water –temperature regime, the genesis of de-
bris flows formation was complex (glacial and rainfall-
induced), where the rainfall had the triggering role.
During periods of extremely high temperatures one
of the catastrophic mechanisms of debris flows forma-
tion is the lake outburst. Such debris flows are glacial-
induced and take place usually without rainfall trigger-
ing. In the Baksan River basin debris flows of glacial
genesis had especially individual character unlike mass
events of debris flows of glacial - rainfall genesis. In
1958-1959 year the outburst of Lake Bashkara in the
upper reach of the Adyl-Su River occurred twice. In
1968, 1975 and 1983 the outbursts of the lake Izmyaltsi
at the Cherek River basin occurred. Debris flow due to
lake outburst also took place in 1978 and 1979 in the
upper sources of the Azay and Adyr-Su valleys respec-
of different hydrological orders. Mass occurrences of
debris flow events, from 30 to 100 of simultaneous de-
bris flow events, took place in 1953, 1966, 1967, 1975,
1977, 1983, 1996, 1999 and 2002.
The release of millions of cubic meters of mud and
debris material from the tributaries in the main chan-
nels formed powerful debris flows, which reached the
foothills of Caucasus mountains. The trigger of debris
flows in all known cases were extreme rains with quan-
tity of precipitation from 23 to 86 mm. The values of
precipitation were recorded in the zone of formation of
catastrophic debris flows in 1953, 1967, 1975, 1977,
1983 and 1996 and also by data of the representative
Terskol meteorological station.
The most catastrophic debris flows in high moun-
tain zone of the Central Caucasus as a whole were
caused by an extreme rain of 1% probability. A 85.7
mm rain occurred on 5-6 August 1967. The maximum
intensity was 4 mm/min (data from Terskol meteoro-
logical station). The average day air temperature at the
height 2100 m on this day was 9°C, the level of the
0°C isotherm was 3500 m above sea level. It provided
liquid precipitation at heights 3000-3200 m, in perigla-
cial zone of debris flow origination. Catastrophic debris
flow passed along the channel of the Terek River, the
main river of the Central Caucasus. In its valley the
highway Vladikavkaz–Tbilisi, a gas pipeline and other
communications were washed off. In the valley of the
Baksan River debris flow descended on settlements. In
the valley of the Baksan River debris flow descended
on the Verkhnii Baksan settlement, destroyed the Tyrn-
yauz – Terskol highway and other communications.
The phenomenon of mass occurence of numerous de-
bris flows of a various order in the Terek River basin
due to extreme rain is indicative. The meteorological
parameters of the debris flow situation of 1967 year
was taken as the reference point between debris flow-
inducing and non-inducing rainfall, according to the
height level of 0°C isotherm position.
In the Baksan River basin the events of 19 July
1983 were the most significant by number of simulta-
neously forming debris flows. Debris flows formed in
83 basins of high mountains tributaries of the Baksan
River. The catastrophic debris flow on the Chegem
River totally washed off the main road connecting the
foothills to the high mountains settlements. In the Adyr-
Su valley extreme debris flow destroyed an alpine hotel
(z
aPoRozHCHenko
, 1985). After 6 day of hot weather,
background image
REGIONAL SHORT–TERM FORECAST OF DEBRIS FLOW INITIATION FOR GLACIATED HIGH MOUNTAIN ZONE OF THE CAUCASUS
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
1007
of debris flows hazard and determination of their im-
portance in complex process of debris flow initiation.
The 38 basic cases of debris flow formation were
used, from the earliest (5 July) to the latest (4 Septem-
ber) seasonal date of debris flow events.
These debris flows occurred at following values
of the air temperature and precipitation (the critical
values of meteorological parameters are designated by
an index “cr” and mean boundary conditions between
situations of debris flow hazard and no hazard):
1. Sum of the daily air temperatures from the date of
steady transition above 0 °C to the date of debris
flow (Q): 1450≥Q≥Q
cr
(Q
cr
=670°C);
2. Sum of the precipitation for the same period (W):
540≥W≥W
cr
(W
cr
=180 mm);
3. Sum of the daily air temperatures for 6 days before
debris flow
4. Daily sum of precipitation (X): 70≥ Х≥ Х
cr
cr
=20
mm);
5. Daily air temperature in the day with precipitation
(T): 14≥ Т≥Т
cr
(Tcr=9.0 °C).
The values of meteorological predictors vary con-
siderably, but they have a certain optimum range. From
data between 1951 and 2005 years, 400 cases with ex-
treme rains have been determined. The debris flows of
glacial–rainfall genesis were formed only if extreme
rains with critical parameters coincided in time with
periods of optimum values of other meteorological pre-
dictors. Such overlap occurred in 38 from 400 cases.
Obviously, this forecast method, which is based
on an expert analysis of critical values of meteorologi-
cal parameters, has some uncertainty. Construction of
a function with use of statistical algorithms was car-
ried out for increasing the efficiency of the forecast of
debris flows of complex genesis.
METHODOLOGY OF DEVELOPING THE FO-
RECAST FORMULA
It is possible to present the generalized descrip-
tion of physical process of debris flow formation as a
formula depending on considered parameters, valid in
the chosen critical range of meteorological predictors.
We used several combinations of meteorological
parameters in developing the forecast formula:
1. Results of multiplying of meteorological parameters.
Product of air temperature and quantity of precipita-
tion is one of versions of the hydrothermal factor,
tively. The debris flow that descended on national re-
sort Dzhily-Su in 2006 and Bulungu settlement in 2007
were catastrophic. The trigger in these cases was ac-
cumulation of big amount of thawed glacial waters dur-
ing periods of extremely hot weather. Cumulative air
temperatures for the 6 days before debris flow reached
the maximum values 90-105°C (data from Terskol me-
teorological station).
During the period of observation there were 17
events of hot weather with the sum of average daily
temperatures for the 6 days more than 90°C. 13 of
these cases were of glacial and glacial-shower genesis.
Among them, the 19 July 1983 event was the most
catastrophic by quantity of debris flows in one day and
those of 19-24 July 2000 at Gerkhozhan-Su River (s
ei
-
nova
et alii, 2003) were the maximum by volume with
3.2 million m3 of sediment released. The probability
of debris flows at the maximum values of meteoro-
logical predictors (daily cumulative temperatures for 6
days from 90 to 105°С and precipitation from 50 to 86
mm) was 80%, sufficient to use for a forecast. But they
constitute only 33% (13 from 38) cases of debris flow
releases for the last 58 years. Often glacial and rainfall-
induced debris flows take place at minimum values of
meteorological predictors, as a result we have some un-
certainty. To improve the forecast, a statistical forecast
function based on the long period of observations data
has been constructed.
DEVELOPING OF THE FORECAST FOR-
MULA
We use the long-term observations of debris
flows initiation conditions, which were carried out
using a unified method (R
ukovodstvo
, 1976). For in-
terpretation of meteorological factors, data from Ter-
skol meteorological station, representative for our
territory, have been used. It was established that to
obtain optimum meteorological parameters describ-
ing conditions for debris flow occurrence, a 60 years
period of observations is necessary. The authors have
58-year data series, so it is practically enough for the
formula construction.
ESTIMATION OF OPTIMUM VALUES OF ME-
TEOPREDICTORS
The first task during construction of the forecast
formula was the determination of the range of opti-
mum values of meteorological predictors of situations
background image
I.B. SEINOVA, Y.B. ANDREEV, I.N. kRYLENkO & S.S. CHERNOMORETS
1008
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
which is widely used in climatology.
2. Sum of daily values of meteorological parameters for
a certain time period - for the description of prepara-
tion of potential debris flow materials in this period.
The formula is as follows:
where А, В, С – specific coefficients (as described be-
low); X - daily sum of precipitation; T – projected daily
air temperature in the day with triggering precipitation;
Q – cumulative air temperatures from date of steady
transition above 0 °C to the date of debris flow; w
cumulative precipitation for the same period, ∑T - cu-
mulative sum of air temperatures for 6 days before the
debris flow.
For calculation of coefficients А, В, С the follow-
ing data are used:
1. the mean (optimal) values of meteorological param-
eters Х
opt
,
which have been determined from observation data
as average values of meteorological parameters Х,
2. the values of weighting coefficient for each mete-
orological factor
or for their joint influence on debris flow initiation P
XT
,
P
Qw
, P
XTQw
.
The weight of factors (P
i
) was defined as a relative
frequency of of debris flow cases to the total number
of days with meteorological parameters above critical
values.
where N
r
- quantity of daily observations with de-
bris flow events occurring at achievement or in excess
of critical values on each of meteorological predictors,
N
i
– total number of days with meteorological param-
eters above critical values.
The following values of weighting coefficient
have been received from data of observations from
1951 to 2005: P
X
=0.11; P
T
=0.05; P
Q
=0,14; Pw=0.06;
P
XT
=0,27, P
Qw
=0.07,
The coefficients А, В, С have been defined as nor-
malized weights of meteorological factors and their
combinations:
Authors received the first version of the forecast
formula from observation data for 1953-1983 years
(a
ndReev
& s
einova
, 1985).
The weights of meteorological factors defining
values of coefficients A, B, C possess wide confidence
intervals. These intervals show that our estimations
differ by 1.5 - 2 times from statistical sample, which
can be received from the full set of tests. Using for-
mulas for calculation of errors (k
oRn
& k
oRn
, 1968),
we defined that the relative error due to uncertainty
of coefficients A, B, C reaches 16%, while an inter-
nal integrity error due to errors in definition of values
of parameters is only 2%. Thus, the efficiency of the
forecast formula is defined basically by reliability of
the coefficients estimation.
Later the formula has been transformed on the ba-
sis of a 44-year series of observations (1951–1995).
It is established that values of critical parameters, re-
ceived for a longer series of data, have not changed
and correspond to the 95% confidence interval, except
for the temperature factor for which the confidence
probability increases to 99%. It means that the factor
of cumulative air temperatures for 6 days is the most
informative factor. Its forecast success rate reaches
80% while for the rainfall factor it is 20% only, but
the latter is important for debris flows of polygenic
glacial-rainfall genesis.
Taking into account the new information, the final
version of the formula, has been accepted:
where X – forecasted amount of precipitation, mm, T
- forecasted air temperature in the day with precipita-
tion, °C,
– cumulative air temperatures for 6 days,
(1)
(2)
(3)
(4)
(5)
background image
REGIONAL SHORT–TERM FORECAST OF DEBRIS FLOW INITIATION FOR GLACIATED HIGH MOUNTAIN ZONE OF THE CAUCASUS
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
1009
ed under the formula.
FSR = 77% for 1984–1995 – the first period of valida-
tion.
There are 17 coincidences from 22 cases predict-
ed under the formula.
FSR = 80% for 1995–2005 –the second period of vali-
dation.
There are 16 coincidences from 19 cases predict-
ed under the formula.
Average forecast success rate of the formula dur-
ing 1953–2005 is 78%. Practical verification of the
forecast function has shown satisfactory results. De-
bris flow events on 15 and 25 July 1996 and on 12 Au-
gust 1996, catastrophic debris flows of glacial genesis
on 19 July 2000 and 7 August 2006 were all predicted
THE ANALYSIS OF A LONG-TERM SERIES OF
FORECAST FUNCTION VALUES
Analyzing structure of the formula it is possible
to present forecast function F as the hydrothermal in-
dicator, whose fluctuations reflect fluctuations of the
climate, virtually included in it. It has changed within
last 58 years in the range from 1 to 1.3 for local debris
including the day of weather forecast release, °C
The forecast for debris flows is given in the case
when F exceeds 1:
ASSESSMENTS OF FORECAST FORMULA EF-
FICIENCY AND RECOMMENDATIONS FOR
REAL-TIME FORECAST
Efficiency of the formula was checked by authors
on dependent (1951-1995) and independent (1995-
2005, Tab. 1) observation data and also through its
practical use in Hydrometeorological Center of the
Republic Kabardino-Balkaria in 1995-1996 and 2000-
2009. To estimate forecast success rate FSR (the re-
lation of correctly predicted cases of debris flow to
the total number of forecasts) statistical processing of
the calculated values forecast function from 1953 for
2005 was carried out.
The following values have been received for all
periods of observations:
FSR = 79% for 1953–1983 – a so-called theoretical
estimation.
There are 23 coincidences from 29 cases predict-
(6)
Tab. 1 - Data for verification of the forecast formula
Symbols of the table: 0 – debris flow absence, 1 – debris flow event (1l- local, 10- 20 cases; 1m – mass occurrence, more
than 20 cases)
background image
I.B. SEINOVA, Y.B. ANDREEV, I.N. kRYLENkO & S.S. CHERNOMORETS
1010
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
temperatures reaches 100-105°C at the height of
2000 m, corresponding to height of 0°C isotherm
during many days above 4000 m) the outbursts of
glacial lakes and debris flows of glacial genesis are
possible. In the case of invasion of moisture-laden
air when the level of 0°C temperature line is above
3500 m, the local or mass occurrence (depending
on quantity of precipitation) of debris flows of gla-
cial –rainfall genesis is possible.
3. The final forecast of the time of the debris flow
event is given by the formula on the basis of me-
teorological forecast of values of air temperature
and precipitation. In the case of F≥1 the short-term
regional forecast of debris flow release is given.
CONCLUSION
The basis of the proposed method of the regional
debris flow forecast of glacial-rainfall genesis is the
long-term series of observations of debris flows and me-
teorological predictors. The verification of the forecast
formula efficiency has shown forecast success rate from
70 to 80%. This result has been reached by using the
complex systemic approach to the solution of the fore-
cast problem. The physical - statistical modelling which
considered multifactor processes of debris flows forma-
tion, has allowed to simplify the methodology consider-
ably and to raise the forecast success rate by 30%.
It has been established, that for this method it is
sufficient to use observation data for one relatively
short natural cycle of meteorological parameters. It is
important for regions, where long series of observa-
tions of debris flow activity are not available.
ACKNOWLEDGEMENTS
Authors are grateful to Irina Malneva (All-Rus-
sian Institute of Hydrogeology and Engineering Geol-
ogy), Nina Kononova (Institute of Geography, Russian
Academy of Sciences), Evgeniy Bogachenko, Irina
Feoktistova, Nadezhda Beituganova (Roshydromet),
Dmitry Petrakov, Olga Tutubalina, Artur Kapitanov
(Moscow State University), Tatiana Sidorova, and
Valeriy Perekrest. This study has been supported by
the Russian Foundation for Basic Research (grants No.
10-05-01127, 09-05-00934, NSC 10-05-92001, NSFC
08-05-92206), NATO Science for Peace Programme
(grants SFP No. 982143 and CLG No. 983784) and
the grant of the President of the Russian Federation of
young Russian scientists (MK-7722.2010.5).
flows and from 1.3 to 1.5 for catastrophic events. Ob-
viously, the maximum and minimum testify the inten-
sification and decreasing of the debris flow activity.
On the basis of a 5-year moving average, it has been
established that the obtained values fluctuate on a si-
nusoid with the period about 16-17 years. Real cycles
of change of regime of debris flow activity depend on
long-term fluctuations of climate and have periods
from 18 (1967-1984) to 23 years (1985-2007) (s
ei
-
nova
et alii, 2007). As a result of the further statisti-
cal processing it is received that the obtained forecast
function values are quasinormally distributed around
the average value of 1.16 with deviation in ±0.17. So
there was no significant trend of debris flow activity
during period of observations. Thus it was established
that meteorological predictors of debris flows have
a natural cycles of about 16-17 years. Therefore the
developed technique of the forecast formula construc-
tion can be applied for shorter series of observations.
Observations data during one cycle of meteorological
parameters change (16-17 years in our case) can be
enough for the forecast formula construction.
RECOMMENDATIONS FOR REAL-TIME FORE-
CAST OF DEBRIS FLOwS
The following recommendations have been de-
veloped for the central Caucasus.
The forecast of time of the debris flow release
is carried out stage-by-stage on the basis of the rep-
resentative meteorological station Terskol data and
weather forecast.
1. The forecast of the beginning of debris flow hazard
period consists in the summation of current daily
average air temperatures from date of steady tran-
sition above 0°C and the precipitation for the same
period. On reaching the critical values Q ≥ 670°C
and W ≥ 180mm the debris flow hazard period is
declared
2. The second step is the forecast of debris flow haz-
ard situation. The real-time observation by tem-
perature regime of weather and summation of air
temperatures for 6 day before the forecast day is
carried out. If the critical sum
is above 70°C, the high probability of debris flow
releases is possible in the following conditions:
–in the cases of extremely hot weather (the sum of
background image
REGIONAL SHORT–TERM FORECAST OF DEBRIS FLOW INITIATION FOR GLACIATED HIGH MOUNTAIN ZONE OF THE CAUCASUS
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
1011
REFERENCES
a
ndReev
Y.B. & s
einova
I.B. (1984) - Model-statistical approach to development of forecast of shower mudflow by the example
of high-mountain Central Caucasus. Bulletin of Moscow University. Series 5. Geography, 4: 86-92 (In Russian).
G
oRsHkov
S.P. (1982) - Exodynamic processes in developed territories. Nedra, Moscow. (In Russian).
k
oRn
G.A. & k
oRn
T.M. (1968) - Mathematical Handbook for Scientists and Engineers. McGraw-Hill Company. N.Y., S.-Fr.,
Toronto, L., Syd.
R
ukovodstvo
Po
izuCHeniyu
selevykH
Potokov
. [Manual for study of debris flows] (1976) - Gidrometeoizdat. Leningrad. (In
Russian).
s
einova
I.B., s
idoRova
T.L. & CHERNOMORETS S.S. (2007) - Processes of debris flow formation and the dynamics of glaciers
in the Central Caucasus. C
Hen
C.-l. & m
aJoR
J.J. (
eds
). Debris-Flow Hazards Mitigation: Mechanics, Prediction, and As-
sessment. Rotterdam: Millpress: 77-85.
s
einova
I.B. & Tatyan L.V. (1977) - Critical values of meteorological parameters of debris flow dangerous situation of Central
Caucasus. Meteorology and Hydrology. Moscow, 12: 74-82 (In Russian).
s
einova
I.B., P
oPovnin
V.V. & z
olotaRev
Y.A. (2003) – Intensification of glacial debris flows in the Gerkhozhan River basin,
Caucasus, in the late 20
th
century. Landslide New,. 14/15: 39-43.
s
einova
I.B & z
olotaRev
E.A. (2001) - Glaciers and debris flows in the Elbrus region. - Evolution of glaciation and debris flow
activity. Nauchny Mir, Moscow. (In Russian with English summary).
t
alanov
E.A. (2007) - Regional assessment of ecological and economic risk of water erosion and debris flows. Almaty. (In
Russian)
w
ateR
R
esouRCes
of
tHe
ussR (1973) - Volume 8, North Caucasus. Leningrad, Russia. (In Russian).
z
aPoRozHCHenko
E.V. (1985) - The extraordinary debris flow in kullumkol-su River. Meteorology and Hydrology. 12: 102-108.
(In Russian).
Statistics