Application of quantitative methods for the assessment of landslide susceptibility of the Aghsuchay river basin

Abstract

Problem statement. Azerbaijan is making a lot of efforts to reduce the impact of dangerous geological processes on natural geosystems, but they still cause huge damage. To a greater extent, the region of the Greater Caucasus, namely the southern slope, is subject to such processes, where the whole range of dangerous geological processes occurs: earthquake (7-8 b and above), landslides, landslides, screes, mudflows, etc. All of them are large-scale processes in terms of damage - they affect large areas and lead to economic losses. Purpose - to identify the main factors of the formation and spread of landslides in the basin of one of the most mudflow-bearing rivers not only in Azerbaijan, but also in the South Caucasus - the Agsuchay river, identify the conditions for their formation, assess the risk of the territory's susceptibility to landslide processes, as well as ways to prevent and protect. Research method. To assess landslide susceptibility and create maps of the potential development of landslides in the basin of the Agsuchay river we used the Frequency Ratio method (FR). Research results. For minimize damage from landslides on the example of the Agsuchay river basin a detailed study of the factors (hypsometry, slope angles (slope steepness) was carried out by us. Also slope exposure, geological structure (lithology), distance from faults, average annual precipitation, distance to the erosion network, distance to roads and land use) that determine the development of landslide processes with taking into account the mechanism of their development, as well as an analysis of the obtained values of landslide susceptibility and their potential development was studied. In the ArcGIS software environment, using the “Raster Calculator” spatial analysis tool, summing up each landslide factor multiplied by its weights, a map of the landslide susceptibility of the Agsuchay river basin was obtained. In the river basin Agsuchay we identified over 120 landslide areas. Most of the landslides were recorded along the Baskal tectonic cover, the Steppe Plateau, as well as on the slopes of the Langyabiz ridge, and also partially on the slopes of the Nialdag ridge. Conclusion. Using the natural boundary classification method in the ArcGIS software environment, the study area was divided into five landslide potential zones: very low, low, medium, high, and very high. The result of the analysis showed that zones with very low, low, medium, high and very high landslide development potential are: 13.75; 24.48; 31.51; 20.51 and 9.74% of the study area, respectively. Ultimately, the reliability of the obtained models was evaluated using AUC ROC (area under the error curve) analysis, which showed high performance of the method used (82%). Due to the high reliability, the method used can be used to assess the landslide susceptibility not only of the territory of Azerbaijan, but of similar regions of the Alpine-Himalayan belt.

Full text

ISSN 2410-7360 Вісник Харківського національного університету імені В.Н. Каразіна
‐257‐
https://doi.org/10.26565/2410-7360-2023-58-20
Received10April2023
UDC 551.3/796.5
Accepted16May2023
Application of quantitative methods for the assessment of landslide
susceptibility of the Aghsuchay river basin
Stara Tarikhazer 1
DSc (Geography), Associate Professor,
1 Institute of Geography named by acad. H.A. Aliyev of MES Azerbaijan,
115 H. Javid Av., Baku, AZ 1143, Azerbaijan,
e-mail: kerimov17@gmail.com, https://orcid.org/0000-0001-5870-1721;
Seymur Mammadov 2
PhD (Geography), Leading Engineer, 2 Production Unit «Azneft», SOCAR,
73 Neftchilar Av., Baku, AZ 1000, Azerbaijan,
e-mail: seymurmq@gmail.com, https://orcid.org/0000-0001-8470-2584;
Zernura Hamidova 1
PhD (Geography), Associate Professor,
e-mail: zernura@gmail.com, https://orcid.org/0000-0002-9605-4884
ABSTRACT
Problem statement. Azerbaijan is making a lot of efforts to reduce the impact of dangerous geological processes on natural
geosystems, but they still cause huge damage. To a greater extent, the region of the Greater Caucasus, namely the southern slope, is
subject to such processes, where the whole range of dangerous geological processes occurs: earthquake (7-8 b and above), landslides,
landslides, screes, mudflows, etc. All of them are large-scale processes in terms of damage - they affect large areas and lead to economic
losses.
Purpose - to identify the main factors of the formation and spread of landslides in the basin of one of the most mudflow-bearing
rivers not only in Azerbaijan, but also in the South Caucasus - the Agsuchay river, identify the conditions for their formation, assess
the risk of the territory's susceptibility to landslide processes, as well as ways to prevent and protect.
Research method. To assess landslide susceptibility and create maps of the potential development of landslides in the basin of
the Agsuchay river we used the Frequency Ratio method (FR).
Research results. For minimize damage from landslides on the example of the Agsuchay river basin a detailed study of the
factors (hypsometry, slope angles (slope steepness) was carried out by us. Also slope exposure, geological structure (lithology), distance
from faults, average annual precipitation, distance to the erosion network, distance to roads and land use) that determine the develop-
ment of landslide processes with taking into account the mechanism of their development, as well as an analysis of the obtained values
of landslide susceptibility and their potential development was studied. In the ArcGIS software environment, using the “Raster Calcu-
lator” spatial analysis tool, summing up each landslide factor multiplied by its weights, a map of the landslide susceptibility of the
Agsuchay river basin was obtained.
In the river basin Agsuchay we identified over 120 landslide areas. Most of the landslides were recorded along the Baskal tectonic
cover, the Steppe Plateau, as well as on the slopes of the Langyabiz ridge, and also partially on the slopes of the Nialdag ridge.
Conclusion. Using the natural boundary classification method in the ArcGIS software environment, the study area was divided
into five landslide potential zones: very low, low, medium, high, and very high. The result of the analysis showed that zones with very
low, low, medium, high and very high landslide development potential are: 13.75; 24.48; 31.51; 20.51 and 9.74% of the study area,
respectively.
Ultimately, the reliability of the obtained models was evaluated using AUC ROC (area under the error curve) analysis, which
showed high performance of the method used (82%). Due to the high reliability, the method used can be used to assess the landslide
susceptibility not only of the territory of Azerbaijan, but of similar regions of the Alpine-Himalayan belt.
Keywords: landslide, mud river, geosystem, tourist and recreational potential, damage, landslide hazard, susceptibility, quanti-
tative methods.
In cites: Tarikhazer Stara, Mammadov Seymur, Hamidova Zernura (2023). Application of quantitative methods for the assessment of landslide
susceptibility of the Aghsuchay river basin. Visnyk of V. N. Karazin Kharkiv National University, series "Geology. Geography. Ecology", (58), 257-
273. https://doi.org/10.26565/2410-7360-2023-58-20
Introduction. Studying hazardous geomorpho-
logical processes and the zoning of developed terri-
tories based on the complexity of engineering and ge-
omorphological conditions and the intensity of devel-
opment of unfavourable geomorphological processes
is regularly conducted and appears relevant for many
years (Tarikhazer, 2022). Slope (landslide) processes
are listed among such dangerous geomorphological
processes. The term landslide refers to the gravitati-
onnal movement of rock masses and fragmental de-
bris, as well as loose soils that have lost their balance
due to a complex of factors under the influence of
their weight down the slope. In contrast with other
natural perils, landslide risk assessment requires a
comprehensive study of the spatial interaction of fac-
tors provoking landslides.
The southern slope of the Greater Caucasus, spe-
cifically the Shamakhi-Ismail region, is a unique re-
©Tarikhazer Stara, Mammadov Seymur, Hamidova Zernura, 2023

This work is licensed under a Creative Commons Attribution 4.0 International License

Серія «Геологія. Географія. Екологія», 2023, випуск 58
‐258‐
gion for the peculiarity of its geographical location.
It should be noted that some factors i.e. the diversity
of landscapes, vegetation, and the animal world, in-
cluding the existing developed infrastructure and
transportation network, determine the region’s role as
one of the largest resort and tourist regions of Azer-
baijan. Vast recreational potential attracts a sheer
number of investors and holiday-makers to places
subject to landslide. In recent decades, the studied
territory has been actively settled down. There have
been done several work on the construction and ex-
pansion of settlements, roads, industrial facilities,
and hotels. Some of these objects pass through or are
in mountainous/ highland areas, requiring increased
attention to the conditions for landslide formation
(Tarikhazer, 2020). The suddenness of emergence,
unpredictability, and close connection with other ge-
ological processes, specifically with mudflow pro-
cesses and phenomena, make landslides a severe,
sometimes even insolvable problem in construction,
requiring the development of landslide prevention
works.
An assessment of the formation and distribution
of landslide processes in a given region will allow
solving many practical and economic problems, pre-
venting undesirable consequences caused by the cat-
astrophic transformation of the primary relief, apply-
ing the results obtained for more rational use of the
territory, and reducing the potential hazard and dam-
age from landslides. All of the above stated reasons
define the relevance of the addressed topic.
The purpose of the study is to determine the
prime drivers in the formation and distribution of
landslides in the Aghsuchay River basin, one of the
most mudflow-bearing rivers not only in Azerbaijan
but also in the South Caucasus, elucidate the condi-
tions conducive to their formation, and assess the
sensitivity risk of the territory for landslide processes,
and provide a way of protection and prevention.
The object of the study is the landslides devel-
oped in the basin of the mud-bearing Aghsuchay
River.
Characteristics of the study region. Aghsu-
chay River basin is located between 40°25'-40°50' N
and 48°15'-48°45' E (Figure 1). In terms of altitude,
the river basin belongs to low-sited water-collecting
headers, where 45.5 km2 or 8% of the basin area falls
to a height above 1500 m. The rest of the river basin
is located at an altitude of less than 1500 m. The av-
erage height is 666 m; the area of the Aghsuchay ba-
sin is 572 km2, while the river length makes 85 km.
The reason for the widespread occurrence of land-
slides here is the complex geological and tectonic
(thick member of clay materials and the existence of
acting tectonic faults) and geomorphological (large
slopes) structure of the territory.
The mountainous part represents the middle al-
titudes, with the absolute elevation reaching 2200 m.
The geographic area is strongly split by the river val-
leys and streams, where the steepness of slopes does
Fig. 1. The geographical location of the basin Agsuchay river

ISSN 2410-7360 Вісник Харківського національного університету імені В.Н. Каразіна
‐259‐
not exceed 30-40°. Tertiary sedimentary and volcano-
genic sedimentary rocks of the Cretaceous Period pri-
marily take part in the geologic aspects of the study
area. The Aghsuchay river flows through a billowy
area, a relatively wide plateau, rising in places to a
height of up to 1000 m. The Gurjivan Plateau borders
the basin in the west and the Langabiz Ridge - in the
east. The average density of the river network makes
0.46 km/km2. The main mudflow centres, specifically
landslides and landslides-flows, are located at an al-
titude of over 1200 m.
In recent years, the number of landslide (slope)
processes has increased dramatically in the basin of
the Aghsuchay River (Figure 2). Based on fund and
own field data, a table of manifestations of landslide
processes in this basin over the past 23 years has been
compiled (Table 1). As can be seen from the table, the
main triggers of landslide processes are atmospheric
precipitation (mainly heavy rains) and anthropogenic
activity, expressed by cutting slopes during construc-
tion work. As field studies (observations) show, im-
proper laying of the drainage network during the con-
struction of new roads in this basin causes rapid
destruction of the roadway. If earlier these processes
were mainly developed in the upper parts of the mid-
dle and high mountains, now they are actively devel-
oping in the low-mountain and foothill zones of the
basin. Along with natural factors, anthropogenic ac-
tivities (earthmoving operations on slopes in the
course of road construction) played a significant role
as well (Figure 3).
The territory, through which the Aghsuchay
River flows, is characterised by a complex orotec-
tonic plan. The structure of the relief reflects the fold-
block basis of the territory, representing a combina-
tion of large and small, positive and negative struc-
tures, disjunctive dislocations, and overthrust masses.
Structural features stem from even greater differenti-
ation due to the close relationship between endoge-
nous and exogenous processes.
At the riverhead, the valley of the Aghsuchay
River is represented by a gorge with steep, often
sheer slopes, ravines, and waterfalls. In the middle
reaches, the river valley expands. Gorge with steep
rocky slopes is formed in places where it cuts through
effusive rocks with interlayers of sandstones and li-
Fig. 2. Active landslides in the basin of the Agsuchay river
Fig.3. Retrospective analysis of landslide formation. a) satellite image of 2004, b) satellite image of 2012
The landslide began to form after the construction of the tourist complex

Серія «Геологія. Географія. Екологія», 2023, випуск 58
‐260‐
Table 1
Dates of manifestation of the most dangerous landslide processes
within the basin of the Agsuchay river (for 2000-2022)
№ Date of mani-
festation
Place of
manifestation
Causes of a
landslide The aftermath of a landslide
1 2 3 4 5
1 March 2000 Chaily village Rains
The landslide intensified in the south-west of Chaily vil-
lage at an altitude of 710-860 m. Landslide length 250-
300 m, width 150-160
m
2 April 2000 Muganli village //-//
The landslide is developed in the central part of the Mu-
ganly village. Landslide length 920-930 m, width 85-100
m
3 April 2001 The villages of Shir-
van and Sheradil //-// Several residential buildings are under the threat of a
landslide. Landslide len
g
th 250-300 m, width 200-220
m
4 April 2001 Geylardag village //-//
The landslide occurred in the central part of the village
of Geylardag. The building of the Historical and Cultural
Reserve was under threat. Landslide length 640-670 m,
width 150-170 m
5 April 2001 Dilman villa
g
e //-// Landslide activation in Dilman villa
g
e
6 May 2002 г. Adnaly village //-// Landslide activation in the south of Adnaly village.
Landslide len
g
th 120-130 m, width 30-40
m
7 April 2003 Chabany village
Leakage of
water from
the reservoi
r
Damaged rural road. Landslide length 1200 m, width
300-330 m
8 May 2003 Nuydu village Rains A landslide intensified at an altitude of 720-860 m south-
west of the villa
g
e of N
y
uid
y
u
9 May 2003 Dedegunesh village //-//
The landslide is developed at an altitude of 1000-1150 m.
Several houses are under threat. Landslide length 5400
m, width 200-220
m
10 May 2003 Kalva village //-// Landslide activation in the north-west of Kalva village at
an altitude of 770-880
m
11 May 2004 Sangalan village //-// Landslide activation in the eastern part of Sangalan vil-
la
g
e. Landslide len
g
th 270-280 m, width 300-320
m
12 May 2004 Garavelli village //-//
Landslide activation. Small cracks were recorded in sev-
eral houses. Landslide length 300-320 m, width 120-130
m
13 March 2005 Khazydere village //-//
Activation of a landslide in the north-eastern part of the
village of Khazydere. Landslide length 80-100 m, width
100-120
m
14 April 2005 Sheradil village Rains, slope
cutting
The landslide is developed at an altitude of 750-880 m.
The landslide first appeared in 1985. Landslide length
200-250 m, width 120-150
m
15 February 25,
2010
Villages of Sangan,
Kalva, Surakhani,
Girrar, Guzai
Rains
80 houses and 4 office buildings were damaged. In San-
galan village (38 houses), Kalva village (35 houses), Su-
rakhani village (2 houses), Girrar village (1 house),
Guza
y
villa
g
e (1 house)
16 April 21,
2010 Chagan village //-//
The 700-meter highway Shamakhi-Uncle Gunash was
closed. Cracks appeared in 20 houses of Mirikand and
Mu
g
anli villa
g
es
17 March 20,
2012
Agsu district, Cha-
gan and Dede-
gunesh villages
//-//
The activity of landslide processes. Landslides and sub-
sidence phenomena have developed on the 2nd km of the
Muganly-Ismayilli highway, landslides have intensified
in the villa
g
es of Cha
g
an and Dede
g
unesh
18 March 21,
2012 Muganli village
Melting
snow and
rains
There are 50 residential buildings in the landslide zone.
18 houses, a school, a club, an outpatient facility are se-
verel
y
deforme
d
19 February 27,
2014
147, 149, 150 and
152 km of Shama-
khi-A
g
su hi
g
hwa
y
Rains Cracks and several subsidence of the asphalt pavement
of the road formed
20 March 7,
2015
Agsu-Shamakhi
hi
g
hwa
y
//-// Vehicular traffic blocked

ISSN 2410-7360 Вісник Харківського національного університету імені В.Н. Каразіна
‐261‐
21 April 2, 2015 Shamakhi regions,
Agsu pass //-// Activation of landslides on the Baku-Shamakhi, Mu-
ganly-Ismayilli and Agsu Pass roads
22 October 22,
2015
Roadbed of Baku-
Shamakhi, Agsu
p
ass
//-//
Activation of landslides at 106, 109 and 111 km of the
Baku-Shamakhi highway, at the Agsu pass. Numerous
cracks foun
d
23 November 12,
2015
106-109 and 111 km
of the Shamakhi-Is-
mayilli highway,
132 and 135 km of
the Baku-Agsu
hi
g
hwa
y
//-// Vehicular traffic blocked
24 February 26,
2016 Sheradil village
Rains and
melting
snow
Cracks 1.5 m deep and more than 60 m long formed on
Baku-Shamakhi-Yevlakh main road
25 June 3, 2016
106-107 km of
Baku-Shamakhi-
Yevlakh hi
g
hwa
y
Rains Vehicular traffic blocked
26 September 22,
2016
Agsu and Shamakhi
regions //-//
Numerous landslide processes on the right bank of the
Agsuchay River, at 147-154 km of the Baku-Shamakhi-
A
g
su hi
g
hwa
y
. Difficult
y
in traffic
27 November 14,
2016 Madras village //-// Cracks appeared on the main road. Landslide area 40 ha
28 December 6,
2016
Shamakhi-Gyzmey-
dan hi
g
hwa
y
//-// Blocked communication with 4 villages
29 January 18,
2017 Madras village
Melted wa-
ters of snow
and rains
7 houses in disrepair
30 February 4,
2017 Agsu pass //-//
Activation of landslides on the Agsu pass, on the Baku-
Shamakhi-Yevlakh highway. Landslides damaged con-
crete barriers. Numerous cracks were found on the as-
p
halt pavement. Difficult
y
in traffic
31 February 17,
2017
149 km of the road
of the A
g
su
p
ass Rains The road sank to a depth of 40-50 sm. Landslide length
40
m
32 March 29,
2017
The villages of Cha-
bany, Madras,
Meysari, Birindzhi
Chaily, Muganly,
Ajidere,
Gale
y
budu
g
//-// The road at 4, 6 and 10 km Shamakhi-Gyzmeydan was
destroyed. Blocked traffic
33 June 2, 2017
106-107 km of
Baku-Shamakhi
hi
g
hwa
y
//-// Landslide intensified on 106-107 km of Baku-Shamakhi
highway
34 June 28, 2017
106-107 km of
Baku-Shamakhi
hi
g
hwa
y
//-// Landslide intensified on 106-107 km of Baku-Shamakhi
highway
35 July 21, 2017
106-107 km of
Baku-Shamakhi
hi
g
hwa
y
//-// Landslide intensified on 106-107 km of Baku-Shamakhi
highway. Asphalt collapsed in two places
36 August 28,
2017
106-107 km of
Baku-Shamakhi
hi
g
hwa
y
//-// Landslide intensified on 106-107 km of Baku-Shamakhi
highway
37 September 12,
2017
106-107 km of
Baku-Shamakhi
hi
g
hwa
y
//-// Landslide intensified on 106-107 km of Baku-Shamakhi
highway
38 October 5,
2017
The villages of Dil-
man, Khatman,
Khadzhman
//-//
Cracks appeared in many houses in the villages of Dil-
man, Khatman and Khadzhman. About 70 houses are in
disrepai
r
39 October 16,
2017 Sagyan village //-//
In the village of Sagian, fences were destroyed, ceilings
sagged in the houses, and there were numerous cracks on
the walls. Residents evacuate
d

Серія «Геологія. Географія. Екологія», 2023, випуск 58
‐262‐
40 October 19,
2017
142, 148 and 152
km of Baku-Agsu-
Yevlakh roa
d
//-// The asphalt sank in some places, cracks appeared 20-30
m long, 10-20 m deep, in some places up to 30 sm
41 January 5,
2018
106-111 km of the
Baku-Shamakhi
road and 132-156
km of the Baku-
A
g
su roa
d
Melted wa-
ters of snow
and rains
Landslide activity at km 106-111 of the Baku-Shamakhi
highway, at km 132-156 of the Baku-Agsu highway
42 April 3, 2018
The villages of Sha-
bany and Dede-
g
unesh
Rains Landslides also intensified in the villages of Shabany and
Dedegunesh. Landslide length 150 m, width 100 m
43 April 13,
2018 Adnaly village //-// There are 12 residential buildings in the landslide zone.
Cracks appeared in 4 houses. Residents evacuate
d
44 April 19,
2018
Road Shamakhi-
Pir
g
uli-Damirchi //-// In many places of the Shamakhi-Pirguli-Damirchi road,
asphalt subsided, numerous cracks develope
d
45 April 24,
2018
3 km Gushchu-
Chayly road
On the 3rd km of the Gushchu-Chayly road, landslide-
subsidence phenomena are developed. Overturned road
si
g
ns alon
g
the roa
d
46 April 24,
2018
19-20 km of Shama-
khi-Geylyar-Pa-
darchel roa
d
//-// Numerous cracks developed on the 19-20th km of the
Shamakhi-Goylar-Padarchel road
47 June 15, 2018 Zaratheiberi village //-//
In the village of Zaratkheyberi, a landslide destroyed a
road bridge. Communication of 300 villagers with the re-
g
ional center was interrupte
d
48 December 11,
2018 Gushchu village //-//
In the village of Gushchu, cracks appeared in 15 houses.
Damage was caused to personal plots, vegetable gardens,
orchards. Roads dama
g
e
d
49 April 1, 2019 Agsu-Khanbulag
road //-//
Landslide processes on the Agsu-Khanbulag roadbed.
Blocked communication between the villages of Khan-
bulag Khingar, Girda, Kendahan, Zargava, Yenikend and
Kevludzh with the re
g
ional cente
r
50 January 10,
2020 Agsu pass //-//
Activation of landslide processes in 4 places of the Agsu
pass. Large cracks formed on the roadway, numerous
landslides and subsidence
51 Ma
y
5, 2021 A
g
su
p
ass //-// Threatened destruction of a rural cemeter
y
52 January 13,
2022
22 km of the high-
way Chukhurdyurd-
Sis-Galeybugurd-
Kechmeddin-Gala-
deresi
//-//
A landslide descended on the 22 km of the Chu-
khurdyurd-Sis-Galeybugurd-Kechmeddin-Galaderesi
highway. Several sections of this 16 km road, which was
reconstructed and put into operation at the beginning of
2020, were destroyed, cracks formed on the asphalt sur-
face
mestones. Its tributaries are characterised by their
nesting in folds overturned to the south. The northern
slopes are steep, while the southern slopes are rela-
tively slightly sloping, corresponding to the bedding
of rocks. The valleys of lateral tributaries developed
in thick and relatively rapidly eroding rocks are
deeper. In the lower reaches, the Aghsuchay River
takes the form of a wide box canyon with a high
floodplain and stream terraces.
The sub-mountain region and low-hill terrain,
along which the Aghsuchay River valley passes, are
characterised by the development of a strong debris
cone plume, in the upper parts of which the largest
material of mudflows has accumulated. The middle
altitudes of the Aghsuchay River basin, are character-
ised by the development of landslide-scree slopes in
dense sandstone-limestone deposits of the Creta-
ceous and Jura. The high mountain regions are an
area of nival-denudation impact and gravitational
processes.
Huge landslide massifs, amphitheatres, circuses,
slowly moving streams, and landslide disruptions
cover the middle course of the Aghsuchay River,
mainly complicating the relief and obscuring the
structural features of the relief. These processes and
landforms of gravitational origin have erased the
structural features of low-order morphostructures.
Massive landslide flows with powerful plume cones
fill the bottoms of the basins of their slopes.
According to N. Shirinov (1982), the presence
of intensely dissected slopes of the longitudinal part
of the Aghsuchay River stems from the intense frag-
mentation of the Lower Cretaceous limestones,
marls, and sandstones by a series of longitudinal frac-
tures active in the latest stage. It is also due to the
scaly structure of the slope with the layers falling to

ISSN 2410-7360 Вісник Харківського національного університету імені В.Н. Каразіна
‐263‐
the north.
One of the main factors in the development of
landslide processes is the orotectonic structure,
which is confirmed by the geological and geomor-
phological analysis of the structure of the region un-
der study.
Materials and research methods. The danger
posed by landslides encourages researchers to look
for the most advanced approaches and tools for their
prediction. In recent years, probabilistic-statistical
methods have been increasingly applied for forecast-
ing purposes. As of the current date, a considerable
number of works have been published on this prob-
lem (van Westen CJ, 1997; Suzen, Doyuran, 2004;
Guzetti et al., 2005; Lee, Pradhan, 2007; van Westen
CJ., 2008; Nefeslioglu et al., 2008 ; Ozdemir, 2009;
Castelanos Albella, Cervi et al., 2010; Oh, Lee, 2011;
Pendin V.V., Fomenko I.V., 2015; Ciurleo et al.,
2017; Arabameri et al., 2019; Cantarino et al., 2019;
Mandal, Mondal, 2019; Nahayo, 2019; Shano et al.,
2020; Kharchenko S.V., Shvarev S.V., 2020; Mersha
T., Meten M., 2020; Rocatti et al., 2021, etc.). It is not
possible to list all researchers in this study. However,
the features, often found in these works, could be
generalised.
The most popular methods for forecasting land-
slide processes include the frequency ratio (Akgun et
al., 2008; Akgun and Needet, 2010; Constantin et al,
2011; Ram Mohan et al., 2011; Yalcin et al., 2011;
Pourghasemi et al., 2012; Berhane, Tadesse, 2021;
Duong et al., 2022), logistic regression (Getachew,
Meten, 2021), linear discriminant analysis, and hier-
archy analysis method (Pourghasemi et al., 2013;
Duong et al., 2021). In recent years, new methods
have been increasingly used: region-partitioning ap-
proach (Hong et al., 2018), random forest (Yesilnajar,
2005), decision tree (Saito et al., 2009; Nefeslioglu et
al., 2010), neural networks (Lee et al., 2003; Sezer et
al., 2011; Pradhan, 2013; Xiong et al., 2019), and
Support Vector Machine.
Similar sets of variables are often used to de-
velop models. Positive altitude, angle of gradient,
layout, distance to faults, distance to watercourses,
etc. take the lead among the quantitative variables in
terms of frequency of use (Fell et al., 2008; van
Westen et al., 2008; Gaidzik, Ramirez-Herrera,
2021). Along with this, qualitative variables are also
applied: the composition of a subsurface rock, the
type of vegetation and land management, and rainfall
amount.
Landslide susceptibility is “the spatial probabil-
ity of landslides occurring in a given area, depending
on local conditions, indicating where landslides may
occur” (Sestrash et al., 2019; Kose and Turk, 2019).
In other words, landslide susceptibility is considered
as the likelihood of a landslide occurring in a partic-
ular area, being estimated based on the quantitative
and qualitative interpretation of certain natural and
anthropogenic factors leading to the landslides emer-
gence. Landslide Susceptibility Mapping (LSM) is
“the process of defining the positional relation and
classifying units of territory based on their propensity
to cause landslides. Topography, geology, soil prop-
erty characteristics, climate, vegetation, and anthro-
pogenic environmental impact influence this process
as well” (Fell et al., 2008). Spatial analysis using GIS
“clarifies the connections between various elements
of slope stability and the development of landslide
processes, being an effective method for assessing
landslide susceptibility” (Van Westen et al., 2006;
McColl, 2015).
Landslide susceptibility analysis is the most
commonly used statistical approach. In this analysis,
landslides and their causing factors are used to de-
velop a landslide susceptibility model to forecast fu-
ture landslides (Aleotti and Chowdhury, 1999;
Gokceoglu et al., 2005; van Westen et al., 2006; Ti-
ranti and Cremonini, 2019). Approaches to assessing
landslide susceptibility could be classified as quanti-
tative and qualitative (Guzetti et al, 1999; Yalcin et
al, 2011; Pendin, Fomenko, 2015; Gaidzik, Ramirez-
Herrera, 2021). Recently, the number of quantitative
estimates of landslide susceptibility has been consid-
erably expanded. It stems from the fact that quantita-
tive approaches deliver the most accurate results.
Nevertheless, qualitative approaches remain relevant
in assessing landslide susceptibility in large regions
or in cases where quantitative approaches are not fea-
sible due to a lack of data.
Quantitative methods are widely applied in as-
sessing landslide susceptibility due to the following
advantages:
1) The results can be easily explained owing to an
independent analysis of each of the maps of the
development factors of landslide (slope) pro-
cesses;
2) Analysis can include the expert assessment,
since specific combinations of variables can be
considered and assessed in terms of their sig-
nificance in the occurrence of landslides;
3) The accuracy of the developed maps can be
verified using data on the spatial distribution
of landslides.
The study has used the Frequency Ratio method
(FR) and the Index of entropy method to assess the
landslide susceptibility and compile maps of the po-
tential development of landslides in the
Girdimanchay River basin. When forecasting the
landslide (slope) processes, it is reasonable to assume
that causal factors define their occurrence. In addi-
tion, it is assumed that future landslides can occur un-
der the same conditions as the previous ones.
Frequency ratio as a bivariate statistical method
represents a straightforward and efficient model for

Серія «Геологія. Географія. Екологія», 2023, випуск 58
‐264‐
assessing landslide susceptibility. This method is
based on the observed correlation between the distri-
bution of landslides and each associated factors to re-
veal the correlation between the places of manifesta-
tion of the process and the factors that cause it in the
study area. According to it, each causative factor is
subdivided into several classes. In addition, the fre-
quency rate (FR) for each class of factors is defined
by applying the following equation (1):
𝐹 


 󰇛1󰇜
where Ni is the number of points (pixels) of land-
slides in the class of factors i; N is the total number
of points (pixels) of landslides on the map of the
study area; Pi is the number of points (pixels) in the
class of factors i; and P is the total number of points
(pixels) on the map of the study area.
The Frequency Ratio (FR) model is commonly
practiced in forecasting geological and geomorpho-
logical processes. It was first proposed by Lee, Talib
(2005). FR is a usable geospatial assessment tool,
predicting the probability distributions of occurrence
and non-occurrence of geological processes for each
class of related factors. Classes are assessed based on
the ratio of observed landslides to the entire study re-
gion. FR is one of the best geospatial assessment
tools for defining the spatial correlation between the
record of stocktaking and a class of related factors.
The amount or percentage of stocktaking record in
each class refers to the significance of correlation
with the development of geological processes. FR
shows the correlation between the locations of geo-
logical processes and the factors influencing the oc-
currence of processes in a given area. Process devel-
opment factors could be assessed, considering the ra-
tio of observed processes in the territory under con-
sideration. The correlation among the class factors
could be determined through FR, which is quite a
useful geospatial estimation tool.
As a result of the analysis, five zones of potential
development of the landslide process have been iden-
tified in the study area: very low, low, medium, high,
and very high.
The results of quantitative methods in assessing
landslide susceptibility are verified by AUCROC
(area under the error curve) analysis.
ROC-curve or curve of errors (Receiver Operat-
ing Characteristic) is a graph, allowing estimating the
quality of two-class classification. It shows the de-
pendence of the number of correctly classified posi-
tive examples on incorrectly classified negative ex-
amples. The Area Under the ROC-curve is an aggre-
gated characteristic of the classification quality, inde-
pendent of the ratio of error rates. The higher the
AUC value, the better the resulting classification
model.
Landslide inventory. One of the main elements
of the FR methodology is the compilation of an in-
ventory map of landslides and fall-scree processes. A
map of manifestations of slope processes can be com-
piled based on both field studies and by interpreting
multispectral images obtained using remote sensing.
The sources for the inventory of landslides are
satellite images of 2008-2022 taken from open
sources (Google Earth, Earthexplorer), field surveys
and geomorphological maps of the region.
The Landslide Inventory Mapping (LIP) is a
map, showing the number of active manifestations of
landslide processes. The compiling inventory maps
of landslides (LIM - Landslide Inventory Mapping)
calls attention to the selection of the boundaries of
landslides and ignores the specifics of landslide de-
formations. LIP is a major element in landslide risk
assessment. At the same time, with the image of the
spatial distribution of landslides, the landslide inven-
tory map can contain the following types of infor-
mation, such as the geometrical features of the land-
slide (scale, area, and depth of capture of the slope
mass by landslide deformations), structural style (li-
thology, structure, and soil properties) and hydrogeo-
logical conditions.
Compiling the landslide and avalanchine-scree
inventory maps will be conducted by applying the
contrast enhancement algorithm proposed by Gond
and Brognoli in 2005. This technique is based on a
combination of spectral bands to generate a normal-
ized difference vegetation index (NDVI) and normal-
ized difference water index (NDWI). Afterwards, the
resulting layers are combined with mid-infrared
(MIR) to increase terrain contrasts (Figure 3).
The study has identified over 120 landslide areas
in the Aghsuchay River basin. Most of the landslides
have been observed along the Baskal tectonic cover,
the Steppe plateau, as well as on the slopes of the
Langabiz Ridge and partially on the slopes of the
Nialdagh Ridge. Slope processes are long-term suc-
cessive events, starting from their formation and end-
ing with the results. Sometimes there is a need only
to prevent its destructive consequences. In many
cases, it is not possible to eliminate the primary rea-
son for a landslide. Then it becomes efficient to mit-
igate the impact rather than trying to eliminate the
cause. It is more common for landslides to emerge
under the influence of geological-tectonic, topo-
graphic, hydrological-climatic, and anthropogenic
factors.
Research results. The research has selected ten
factors, associated with landslides, for landslide sus-
ceptibility mapping and compiling maps for the po-
tential development of landslides based on the avail-
able data, the characteristics of landslides, and the re-
lationship between the formation of slope defor-
mations and factors causing landslides. These factors

ISSN 2410-7360 Вісник Харківського національного університету імені В.Н. Каразіна
‐265‐
Fig. 4. Map of the distribution of landslides in basin Agsuchay river
include hypsometry, slope angles (steepness of
slopes), slope exposure, topographic moisture indi-
ces, geological structure (lithology), distance from
faults, average annual precipitation, distance to ero-
sion network, distance to roads, and land use (Fig. 5).
Maps for hypsometry, slope and aspect were
compiled using a digital elevation model (DEM). To
create a digital elevation model were used SRTM
(resolution 30m) and Alos Palsar (resolution1 2.5m)
data. Fault distance, road distance, and streams dis-
tance were estimated using the “Euclidean distance”
tool in Spatial Analyst Toolbox in ArcGIS Desktop
(ArcMap). A map of average monthly precipitation
was developed by interpolating rainfall data for
nearby settlements (Figure 6). The Land use and land
cover map was developed based on the classification
training in the ArcGIS software environment.
Data on the lithology and tectonic disturbances
were digitised from the geological map of the Moun-
tainous Shirvan economic region at a scale of
1:200,000. Afterwards, to conduct analysis, these
maps were transformed into a raster format to calcu-
late the weights of classes and factors and compile
maps of landslide susceptibility.
The slope angle is considered one of the key
triggering agents in landslide events and is widely
utilised in landslide susceptibility mapping; this pa-
rameter was obtained from the DTM using spatial
analysis tools and reclassified into five classes based
on the natural boundaries (intervals) algorithm.
Aspect is the most important source of changes
in soil properties. It determines the direction of the
slope and is measured clockwise in degrees from 0
(North) to 360 (North again), forming a full circle.
Flat areas have a value of -1. The effect of slope ex-
posure is reflected in differences in temperature and
humidity between polar and equatorial exposure. Ac-
cordingly, slopes facing south and west are warmer
than slopes exposed to the east and north. On the
other hand, the southern slopes in the mountainous
area make it possible to determine the places where
the snow can melt first. Differences in impact on
slopes determine changes in soil properties due to
their effect on microclimatic and vegetation condi-
tions. Such parameters related to the orientation of
the slope as exposure to sunlight, dry wind, rainfall,
and discontinuities, can influence the occurrence of
landslides. The slope exposure was split into nine cat-
egories: flat (–1°), northern (0°-22.5°; 337.5°-360°),
northeastern (22.5°-67.5°), eastern (67.5°-112.5°),
southeast (112.5°-57.5°), south (157.5°-202.5°),
southwest (202.5°-247 .5°), western (247.5°-292.5°-
292.5°-337.5°), and northwestern (292.5°-337.5°).
The distance to rivers is one of the factors de-
termining the stability of the slope as they destroy the
toe of the slope, resulting in erosion processes. Con-
sequently, the risk of soil slippage decreases due to
an increase in the distance from watercourses.
Based on the Distance to Rivers parameter, the
research has identified three classes using the ArcGIS
Euclidean distance tool: up to 200 m, within 200-500
m, and over 500 m.
The Topographic Wetness Index (TWI) is a ro-
bust index of wetness. It is mainly used for quantitive
evaluation of the topographic control of hydrological
processes. TWI is an indicator of the hydromor-

Серія «Геологія. Географія. Екологія», 2023, випуск 58
‐266‐
phism of the soil continuum, commonly defined by
the relief features of the studied territory. Large val-
ues of the index correspond to the wet deposition, and
its increased content in the soil. TWI is estimated us-
ing the following formula (2), where a is the total run-
off into the cell (non-dimensional value); tan b is the
surface slope within the cell (in radians):
𝑇𝑊𝐼  𝑙𝑛 𝑎
tan𝑏 󰇛2󰇜
Based on the Distance to roads parameter, the
research has identified three classes using the
ArcGIS: distance up to 500 m, within 500-1000 m,
and over 1000 m.
Discussion. The article has analysed the rela-
tionship between the factors causing landslides and
the actual occurrence of landslides in the study area.
The frequency ratios (Fij) and the weight factor (Wj)
were calculated using the FR models, respectively
(Table 2).
Fig. 5. Maps of landslide factors in the basin of the Agsuchay river

ISSN 2410-7360 Вісник Харківського національного університету імені В.Н. Каразіна
‐267‐
Fig. 6. Distribution map of annual precipitation in the basin Agsuchay river
Table 2
Analysis of the relationship between factors causing landslides and landslide distribution
Factor Name Factor classes Class area, Pi Landslide area, Ni FRij Weight
factor a, Wi
Hypsometry,
in meters
14-284 306323400 4993,08222 0,005907426
3,83
284-616 203414900 1314290,198 2,341629738
616-973 258086400 828851,6485 1,16391537
973-1444 117177100 49930,8222 0,154431429
1444-2237 54171620 393343,9215 2,631542192
Slope, in de-
grees
0-6 330906600 66019,64268 0,072005394
3,32
6-12 227688500 841611,7475 1,334037325
12-19 217790500 1162278,583 1,926054587
19-28 121181000 483219,4015 1,439153503
28-69 37695560 38280,29702 0,366507096
Aspect
flat areas 50809610 74341,44638 0,528058947
1,00
north 26740180 110957,3827 1,497576918
northeast 64591080 272955,1613 1,525162171
east 82213330 207490,3056 0,910862518
southeast 131276500 357282,7722 0,982250699
south 173426400 601389,014 1,251519271
southwest 184496600 494869,9267 0,968054829
west 131472300 309016,3107 0,848290108
northwes
t
90236100 163107,3525 0,652364743
Geology*
1 29494140 97087,70982 1,193028445
2,86
2 140260100 0 0
3 120623500 0 0
4 139786350 0 0
5 51286170 155895,1226 1,101677652
6 61790515 164216,9263 0,963204235
7 33116900 42718,59232 0,467508526
8 108513000 760612,8581 2,540412095

Серія «Геологія. Географія. Екологія», 2023, випуск 58
‐268‐
9 11266060 221914,7653 7,138984167
10 43814306 848269,1903 7,01681515
11 5965070 4438,295306 0,269663974
12 59204091,1 12760,09901 0,078113232
13 98113539,1 128155,777 0,473403567
14 31160724,2 155340,3357 1,806753797
Volcanic breccia 4806120 0 0
Distance to
faults, in me-
ters
0-500 347749900 1543971,98 1,597895539
3,15
500-1000 161212200 662415,5744 1,47879718
1000-1500 87413350 316783,3275 1,30424907
1500< 336260300 68238,79033 0,073035113
Distance to
streams, in
meters
<200 469655500 1388076,857 1,067794446
2,61
200-400 277143300 763016,9347 0,994679918
400-600 140608000 338604,946 0,870035199
600-800 44409040 101710,9341 0,827465345
>800 4429420 0 0
Topographic
wetness index
(TWI)
<6 332261900 759503,2843 0,824985585
3,73
6-12 563107100 1698757,529 1,088773744
12-18 36750760 133148,8592 1,307580313
>18 3142314 0 0
Distance to
roads, in me-
ters
<300 278845400 242441,8811 0,310793604
2,28
300-600 165115200 315673,7536 0,683406856
600-900 116582400 367823,7235 1,127806665
900-1200 84719300 378364,6748 1,596452995
>1200 281064000 1287105,683 1,636955589
Land use and
land cover
water 2350633 554,786913 0,085643955
5,26
forests and for-
est shrubs
194466200 249654,111 0,46585411
crops 290814400 13869,67283 0,017306357
built area 50804620 21081,9027 0,150578142
bare ground 9854127 101526,0051 3,738647165
rangeland 392064100 2204723,193 2,040573919
Precipitation,
mm per year
468-515 120732192,2 0 0
8,77
515-552 177959026 0 0
552-586 223405512,7 64355,28194 0,104399699
586-613 417074996,9 2527054,39 2,195883873
* geological classes: 1 - Holocene, modern alluvial deposits - pebbles, gravel, sands, sandy loams, loams; 2 - Holo-
cene, modern deluvial-proluvial deposits - pebbles, loams, sandy loams, clays; 3 - Middle and Upper Pleistocene, allu-
vial-proluvial deposits - clays, loams, sandy loams, pebbles with layers of volcanic ash; 4 - Eopleistocene, Absheron
marine sediments - clays, sands, sandstones, limestones with layers of volcanic ash, loams, marls, conglomerates; 5 -
Upper Pliocene, Akchagyl sedimentary deposits - clays, volcanic ash, breccias, sands, sandstones, pebbles, limestones;
6 - Lower Pliocene, Balakhani sedimentary deposits - clays, loams, sands, sandstones, pebbles, gravelstones, conglom-
erates; 7 - Upper Miocene, Pontian sedimentary deposits - sands, sandstones, clays, limestones, conglomerates, volcanic
ash; 8 - Lower Miocene, Upper Maikop sedimentary deposits - shale clays with interlayers of clayey siderite nodules,
volcanic ash, sands, sandstones, gravelstones, conglomerates; 9 - Oligocene and Lower Miocene Maikop sedimentary
deposits - clays, mudstones, sandstones, marls; 10 - Eocene (Govundag Formation) sedimentary deposits - clays, marls,
sandstones, clayey dolomites, volcanic ash, conglomerates; 11 - Middle and Upper Paleocene (Sumgait Formation) sed-
imentary deposits - clays, marls, sandstones; 12 - Lower Paleocene sedimentary deposits - limestones, sandstones, marls,
argellites, clays; 13 - Upper Cretaceous volcanogenic-sedimentary and sedimentary deposits - tuff sandstones, basaltic
andesites, porphyrites, limestones, sandstones, marls, mudstones, clays; 14 - Lower Cretaceous volcanogenic-sedimen-
tary and sedimentary deposits - tuff sandstones, sandstones, marls, limestones, mudstones, tuff conglomerates, tuffs, tuf-
fites, porphyrites, andesites.

ISSN 2410-7360 Вісник Харківського національного університету імені В.Н. Каразіна
‐269‐
For the studied territory, the LSI values were de-
termined using the following formula, when compi-
ling maps of the landslide susceptibility index:
𝐿𝑆𝐼  󰇛Hypsometry  𝑊
  Angle of gradient  𝑊
Exposure𝑊
  Geologic aspects  𝑊

Distance to faults𝑊
  Distance to watercourses  𝑊

󰇛Тopographic Wetness Index 𝑊
󰇜
󰇡Distance to roads  𝑊
󰇜  Land management and vegetation cover  𝑊

 Amount of precipitation  𝑊
󰇢 󰇛7󰇜
In this connection, a map of the landslide sus-
ceptibility of the Aghsuchay River basin was com-
piled in the ArcGIS software environment by
summing up each landslide factor multiplied by its
weights, using the Raster Calculator spatial analysis
tool (Fig. 7).
Fig. 7. Map of landslide susceptibility of the basin Agsuchay river
Using the natural boundary classification
method in the ArcGIS software environment, the
studied area was divided into five potential landslide
zones: very low, low, moderate, high, and very high.
The analysis result showed that zones with very low,
low, medium, high, and very high landslide occur-
rence potential make 13.75; 24.48; 31.51; 20.51, and
9.74% of the study area, respectively. Areas with high
and very high landslide susceptibility mainly cover
the regions towards the Nialdagh Ridge, the slopes of
the Steppe (tertiary) Plateau, and Langabiz Ridge. Al-
most the entire highland part of the basin is located in
a zone of very high and high landslide susceptibility.
The areas of low and very low susceptibility cover
the flat part of the basin, as well as flat areas. The
AUC value (from ArcSDM tool in ArcGIS Desktop)
is 82%, showing the effectiveness of the method used
for landslide susceptibility mapping and landslide de-
velopment potential in the study area (Fig. 8).
The areas of distribution of Upper Cretaceous
and Maikop clays, clay shales, and limestones have a
high and very high potential for the development of
landslides. Landslides are more common in areas
with sparse vegetation, subalpine meadows, and
partly in agricultural areas. Excavation work on the
slopes, together with a large amount of precipitation,
increases the risk of a landslide in the study area.
Conclusion. Landslide hazard assessment is an
essential component of the national disaster preven-
tion and mitigation strategy in Azerbaijan. The zon-
ing of the territory, according to the potential for the
development of a landslide process, is the basis for

Серія «Геологія. Географія. Екологія», 2023, випуск 58
‐270‐
Fig. 8. Graph showing the validity of the model
assessing the landslide hazard associated with this
risk and designing early warning systems.
Based on this idea, the study has made an anal-
ysis of the development potential of a landslide pro-
cess in the Aghsuchay River basin on the southern
slope of the Greater Caucasus. Landslide susceptibil-
ity mapping was compiled using statistical models
(based on GIS), making it possible to determine the
significance of each parameter influencing the devel-
opment of landslide processes. Subsequently, suscep-
tibility assessment was conducted by aggregating the
result of the analysis of selected factors using spatial
analytical equations. The study area was classified
into five zones based on the degree of potential de-
velopment of landslides: very low, low, medium,
high, and very high. The accuracy of the obtained
models was estimated using the AUC ROC (area un-
der the error curve) analysis, which showed the high
performance of the method used.
The results of the research conducted are of ma-
jor importance for estimating landslide hazards and
risks, planning sustainable land use, and reducing
damage from landslides in the area under considera-
tion. By virtue of its high reliability, the method can
be used to assess the landslide susceptibility of vari-
ous regions of Azerbaijan.
References
1. Akgun, A., Dag, S., Bulut F. (2008). Landslide susceptibility mapping for a landslide-prone area (Findikli, NE of
Turkey) by likelihood frequency ratio and weighted linear combination models. Environmental Geology. 54(6), 1127–
1143.
2. Akgun, A., Needet, T. (2010). Landslide susceptibility mapping for Ayvalik (Western Turkey) and its vicinity by multi
criteria decision analysis. Environmental Earth Science. 61, 595–611.
3. Aleotti, P., Chowdhury, R. (1999). Landslide hazard assessment: summary review and new perspectives. Bulletin of
Engineering Geology and Environment. 58, 21-44.
4. Arabameri, A., Pradhan, B., Rezaei, K., Lee, C.-W. (2019). Assessment of landslide susceptibility using statistical-
and artificial intelligence-based FR–RF integrated model and multiresolution DEMs. Remote Sensing. 11 (9). URL:
https://doi.org/10.3390/rs11090999.
5. Baynes, F.J., Lee, I.K., Stewart, I.E. (2002). A study of the accuracy and precision of some landslide risk analyses.
Australia: Geomech. 37, 149-156.
6. Berhane, G., Tadesse, K. (2021). Landslide susceptibility zonation mapping using statistical index and landslide sus-
ceptibility analysis methods: a case study from Gindeberet district, Oromia Regional State, Central Ethiopia. Journal
of African Earth Sciences. 180. URL: https://doi.org/10.1016/j.jafrearsci.2021.104240.
7. Brardinoni, F., Slaymaker, О., Hassan, М.А. (2003). Landslide inventory in a rugged forested watershed: a compar-
ison between air-photo and field survey data. Geomorphology. 54, 179-196.
8. Cantarino, I., Carrion, M.A., Goerlich, F., Martinez Ibañez, V.A (2019). ROC analysis-based classification method
for landslide susceptibility maps. Landslides. 16, 265–282. URL: https://doi.org/10.1007/s10346-018-1063-4.
9. Castellanos Abella, E.A., Van Westen, C.J. (2008). Qualitative landslide susceptibility assessment by multicriteria
analysis: A case study from San Antonio del Sur, Guantánamo, Cuba. Geomorphology. 94, 453–466.
10. Cervi, F., Berti, M., Borgatti, L., Ronchetti, F., Manenti, F., Corsini, A. (2010). Comparing predictive capability of
statistical and deterministic methods for landslide susceptibility mapping: a case study in the northern Apennines
(Reggio Emilia Province, Italy). Landslides. 7, 433–444.
11. Chacon, J., Irigaray, C., Fernandez, T. et al. (2006). Engineering geology maps: landslides and geographical infor-
mation systems. Bulletin of Engineering Geology and the Environment. 65 (4), 341-411.

ISSN 2410-7360 Вісник Харківського національного університету імені В.Н. Каразіна
‐271‐
12. Ciurleo M., Cascini L., Calvello M. (2017). A comparison of statistical and deterministic methods for shallow land-
slide susceptibility zoning in clayey soils. Engineering Geology. 223 (7). 71–81. URL:
https://doi.org/10.1016/j.enggeo.2017.04.023.
13. Constantin, M., Bednarik, M., Jurchescu, M.C., Vlaicu M. (2011). Landslide susceptibility assessment using the biva-
riate statistical analysis and the index of entropy in the Sibiciu Basin (Romania). Environmental Earth Science. 63,
397-406. https://doi.org/10.1007/s12665-010-0724-y.
14. Corominas, J., van Westen, C.J., Frattini, P. et al. (2014). Recommendations for the quantitative analysis of landslide
risk. B. Eng. Geol. Environ. 73, 209-263.
15. Duong, V.B., Fomenko, I.K., Nguyen, T.K., ThiHong, L.Vi., Zerkal, O.V., Hong, D.Vu. (2022) Application of GIS-
based bivariate statistical methods for landslide potential assessment in Sapa, Vietnam. Bulletin of the Tomsk Poly-
technic University [TPU Bulletin]. Geo Assets Engineering. 333 (4), 126-140. URL: http://earchive.tpu.ru/han-
dle/11683/70765.
16. Fell, R., Corominas, J., Bonnard, C., Cascini, L., Leroi, E., Savage, W.Z. (2008). Guidelines for landslide suscepti-
bility, hazard and risk zoning for land use planning. Engineering Geology. 102 (3), 85–98. URL:
https://doi.org/10.1016/j.enggeo.2008.03.022.
17. Froude, M.J., Petley, D.N. (2018). Global fatal landslide occurrence 2004 to 2016. Natural Hazards and Earth System
Sciences. 18 (8), 2161–2181. URL: https://doi.org/10.5194/nhess-18-2161-2018.
18. Gaidzik, K., Ramírez-Herrera, M.T. (2021). The importance of input data on landslide susceptibility mapping. Scien-
tific Reports. 11 (1). URL: https://doi.org/10.1038/s41598-021-98830-y.
19. Getachew, N., Meten, M. (2021). Weights of evidence modeling for landslide susceptibility mapping of Kabi-Gebro
locality, Gundomeskel area, Central Ethiopia. Geoenvironmental Disasters. 8 (1). URL:
https://doi.org/10.1186/s40677-021-00177-z.
20. Gokceoglu, C., Sönmez, H., Nefeslioglu, H.A., Duman, T.Y., Can, T. (2005). Kuzulu landslide (Sivas, Turkey) and
landslide susceptibility map of its near vicinity. Engineering Geology. 81 (1), 65-83.
21. Guzzetti, F., Reichenbach, P., Cardinali, M., Galli, M., Ardizzone, F. (2005). Probabilistic landslide hazard assess-
ment at the basin scale. Geomorphology. 72 (1), 272–299. URL: https://doi.org/10.1016/j.geomorph.2005.06.002
22. Haque, U., Da Silva, P.F., Devoli, G., Pilz, J., Zhao, B., Khaloua, A., Wilopo, W., Andersen, P., Lu, P., Lee, J.,
Yamamoto, T., Keellings, D., Wu, J.-H., Glass, G.E. (2019). The human cost of global warming: deadly landslides
and their triggers (1995–2014). Science of The Total Environment. 682, 673–684. URL: https://doi.org/10.1016/j.sci-
totenv.2019.03.415.
23. Hong, H., Pradhan, B., Sameen, M.I. et al. (2018). Improving the accuracy of landslide susceptibility model using a
novel region-partitioning approach. Landslides. 15 (4), 753–772.
24. Kose, D.D., Turk, T. (2019). GIS-based fully automatic landslide susceptibility analysis by weight-of-evidence and
frequency ratio methods. Physical Geography. 40 (5), 481–501. URL:
https://doi.org/10.1080/02723646.2018.1559583.
25. Lee, S., Pradhan, B. (2007). Landslide hazard mapping at Selangor, Malaysia using frequency ratio and logistic
regression models. Landslides. 4, 33–41.
26. Lee, S., Ryu, J.H., Min, K.D., Won, J.S. (2019). Landslide susceptibility analysis using GIS and artificial neural net-
work. Earth Surface Process Landforms. 27, 1361–1376.
27. Mandal, S., Mondal, S. (2019). Statistical approaches for landslide susceptibility assessment and prediction. Switzer-
land: Springer International Publishing, 200.
28. McColl, S.T. (2015). Chapter 2. Landslide causes and triggers // Landslide hazards, risks and disasters (J.F. Shroder,
T. Davies. Eds.). Boston: Academic Press, 17–42. URL: https://doi.org/10.1016/B978-0-12-396452-6.00002-1.
29. Mersha, T., Meten, M. (2020). GIS-based landslide susceptibility mapping and assessment using bivariate statistical
methods in Simada area, northwestern Ethiopia. Geoenvironmental Disasters. 7 (1). URL:
https://doi.org/10.1186/s40677-020-00155-x.
30. Nahayo, L., Mupenzi, C., Habiyaremye, G., Kalisa, E., Udahogora, M., Nzabarinda, V., Li, L. (2019). Landslides
hazard mapping in Rwanda using bivariate statistical index method. Environmental Engineering Science. 36 (8), 892–
902. URL: https://doi.org/10.1089/ees.2018.0493.
31. Nefeslioglu, H.A., Duman, T.Y., Durmaz, S. (2008). Landslide susceptibility mapping for a part of tectonic Kelkit
Valley (Eastern Black Sea region of Turkey). Geomorphology. 94 (3), 401–418. URL: https://doi.org/10.1016/j.geo-
morph.2006.10.036.
32. Nefeslioglu, H.A., Gokceoglu, C., Sonmez H. (2008). An assessment on the use of logistic regression and artificial
neural networks with different sampling strategies for the preparation of landslide susceptibility maps. Engineering
Geology. 97, 171–191.
33. Nefeslioglu, H.A., Sezer, E., Gokceoglu, C., Bozkir, A.S., Duman, T.Y. (2010). Assessment of landslide susceptibility
by decision trees in the Metropolitan Area of Istanbul, Turkey. Mathematical Problems in Engineering. Article ID
901095, 15. https://doi.org/10.1155/2010/901095.
34. Oh, H.-J., Lee, S. (2011). Cross-application used to validate landslide susceptibility maps using a probabilistic model
from Korea. Environmental Earth Science. 64, 395-409.
35. Oh, H.-J., Lee, S., Hong, S.-M. (2017). Landslide susceptibility assessment using frequency ratio technique with iter-
ative random sampling. Journal of Sensors. URL: https://doi.org/10.1155/2017/3730913.

Серія «Геологія. Географія. Екологія», 2023, випуск 58
‐272‐
36. Ozdemir, A. (2009). Landslide susceptibility mapping of vicinity of Yaka Landslide (Gelendost, Turkey) using condi-
tional probability approach in GIS. Environmental Geology. 57, 1675-1686.
37. Pourghasemi, H.R., Mohammady, M., Pradhan, B. (2012). Landslide susceptibility mapping using index of entropy
and conditional probability models in GIS: Safarood Basin, Iran. CATENA. 97, 71–84.
38. Pourghasemi, H.R., Moradi, H.R., Aghda, S.F. (2013). Landslide susceptibility mapping by binary logistic regression,
analytical hierarchy process, and statistical index models and assessment of their performances. Natural Hazards. 69
(1), 749–779.
39. Pradhan, B. (2013). A comparative study on the predictive ability of the decision tree, support vector machine and
neuro-fuzzy models in landslide susceptibility mapping using GIS. Computer Geosciences. 51, 350–365.
40. Ram Mohan, V., Jeyaseelan, A., Naveen Raj, T., Narmatha, T., Jayaprakash, M. (2011). Landslide susceptibility map-
ping using frequency ratio method and GIS in south eastern part of Nilgiri District, Tamilnadu, India. International
Journal Geomatics and Geoscience. 1 (4), 951-961.
41. Reichenbach, P., Rossi, M., Malamud, B.D., Mihir, M., Guzzetti, F. (2018). A review of statistically-based landslide
susceptibility models. Earth-Science Reviews. 180, 60–91. URL: https://doi.org/10.1016/j.earscirev.2018.03.001.
42. Roccati, A., Paliaga, G., Luino, F., Faccini, F., Turconi, L. (2021). GIS-based landslide susceptibility mapping for
land use planning and risk assessment. Land. 10, (2). URL: https://doi.org/10.3390/land10020162.
43. Saito, H., Nakayama, D., Matsuyama, H. (2009). Comparison of landslide susceptibility based on a decision-tree
model and actual landslide occurrence: The Akaishi Mountains, Japan. Geomorphology. 109, 108–121.
44. Schlögel, R., Doubre, C., Maletl, J.-P. et al. (2015). Landslide deformation monitoring with ALOS/PALSAR imagery:
a DInSAR geomorphological interpretation method. Geomorphology. 231, 314-330.
45. Sezer, E.A., Pradhan, B., Gokceoglu, C. (2011). Manifestation of an adaptive neuro-fuzzy model on landslide suscep-
tibility mapping: Klang valley, Malaysia. Expert Systems with Applications. 38 (7), 8208-8219.
46. Sestraș, P., Bilașco, Ș., Roșca, S., Naș, S., Bondrea, M.V., Gâlgău, R., Vereș, I., Sălăgean, T., Spalević, V., Cîmpeanu,
S.M. (2019). Landslides susceptibility assessment based on GIS statistical bivariate analysis in the hills surrounding
a metropolitan area. Sustainability. 11 (5). URL: https://doi.org/10.3390/su11051362.
47. Shano, L., Raghuvanshi, T.K., Meten, M. (2020). Landslide susceptibility evaluation and hazard zonation techniques–
a review. Geoenvironmental Disasters. 7 (1). URL: https://doi.org/10.1186/s40677-020-00152-0.
48. Shannon, C.E. (1950). Prediction and entropy of printed English. The Bell System Technical Journal. 30, 50-64.
49. Shaw, S.C., Vaugeois, L.M. (1999). Comparison of GIS-based Models of Shallow Landsliding for Application to Wa-
tershed Management. Seattle: State of Washington Timber/Fish/Wildlife Publication 118, TFW-PR10-99-001, 132.
50. Süzen, M.L., Doyuran, V. (2004). A comparison of the GIS based landslide susceptibility assessment methods: multi-
variate versus bivariate. Environmental Geology. 45 (5), 665–679. URL: https://doi.org/10.1007/s00254-003-0917-8.
51. Tarikhazer, S.A. (2020). The geographical prerequisites for the identification and prevention of dangerous geomor-
phological processes in the mountain geosystems of the Alpine-Himalayan belt (on the example of the Major Caucasus
of Azerbaijan). Bulletin of Dnipropetrovsk University. Geology, Geography and Geoecology, 1. 176-187. DOI
https://doi.org/10.15421/112016
52. Tarikhazer, S.A. (2022). Assessment of ecological strength and risk of geosystems of the north-eastern slope of the
Great Caucasus (within Azerbaijan). Bulletin of V.N. Karazin Kharkiv National University, series Geology. Geogra-
phy. Ecology. 56. Pp. 264-276 https://doi.org/10.26565/2410-7360-2022-56-20
53. Tiranti, D., Cremonini, R. (2019). Editorial: landslide hazard in a changing environment. Frontiers in Earth Science.
7 (3). URL: https://doi.org/10.3389/feart.2019.00003.
54. Van Westen, C.J. (1997). Statistical landslide susceptibility analysis. ILWIS 2.1 for Windows application guide. En-
schede: ITC Publ., 73–84.
55. Van Westen, C.J., Van Asch, T.W.J., Soeters, R. (2006). Landslide hazard and risk zonation – why is it still so difficult?
Bulletin of engineering geology and the environment. 65 (2), 167–184. URL: https://doi.org/10.1007/s10064-005-
0023-0.
56. Van Westen, C.J., Castellanos E., Kuriakose S.L. (2008). Spatial data for landslide susceptibility, hazard, and vulner-
ability assessment: An overview. Engineering Geology. 102, 112-131.
57. Xiong, J., Li, J., Zhang, H., Sun, M., Cheng, W. (2019) Quantitative Hazard Assessment of Landslides Using the
Levenburg–Marquardt Back Propagation Neural Network Method in a Pipeline Area. Geosciences. 9 (10), 1-23.
URL: https://doi.org/10.3390/geosciences9100449.
58. Yalcin, A., Reis, S., Aydinoglu, A.C., Yomralioglu, T. A (2011). GIS-based comparative study of frequency ratio, an-
alytical hierarchy process, bivariate statistics and logistics regression methods for landslide susceptibility mapping
in Trabzon, NE Turkey. CATENA. 85 (3), 274–287. URL: https://doi.org/10.1016/j.catena.2011.01.014.
59. Yesilnacar, E.K. (2005). The application of computational intelligence to landslide susceptibility mapping in Turkey
[Candidate’s dissertation]. Department of Geomatics the University of Melbourne, 423.
60. Belov A.V., Sokolova L.P. (2012). Nekotorye aspekty ekologicheskih riskov prirodopol'zovaniya na yuge Bajkal'skoj
Sibiri [Some aspects of environmental risks of nature management in the south of Baikal Siberia]. Geografiya i pri-
rodnye resursy [Geography and natural resources], 4, 90-97 [in Russian].
61. Budagov B.A., Lilienberg D.A., Shirinov N.Sh. (1960). Istoriya razvitiya gidrograficheskoj seti Yugo-Vostochnogo
Kavkaza [History of the development of the hydrographic network of the South-Eastern Caucasus]. Izv. AN Azeb.
SSR, Seriya geol.-geograf. nauk [Proceedings of AS of Azerbaijan. SSR, Series geol.-geographer. Sciences], 1, 3-7
[in Russian].
62. Budagov, B.A. (1993). Gravitacionnaya morfoskul'ptura [Gravitational morphosculpture]. V knige: "Rel'ef
Azerbajdzhana" [In book: Relief of Azerbaijan]. Baku: Elm, 22-28 [in Russian].

ISSN 2410-7360 Вісник Харківського національного університету імені В.Н. Каразіна
‐273‐
63. Leonova, A.V., Strokova, L.A., Nikitenkov, A.N. (2021). Ocenka opolznevyh processov na territorii g. Tomska s
ispol'zovaniem GIS-tekhnologij [Assessment of landslide processes on the territory of Tomsk by using GIS Technolo-
gies]. Vestnik Voronezhskogo Gosudarstvennogo Universiteta. Seriya: Geologiya [Proceedings of Voronezh State
University. Series: Geology], 1, 94–103. DOI: https://doi.org/10.17308/geology.2021.1/3341[in Russian].
64. Osipov, V.I. (2001). Prirodnye katastrofy na rubezhe XXI veka. [Natural disasters at the turn of the XXI century].
Vestnik Rossijskoj Akademii nauk [Bulletin of the Russian Academy of Sciences]. M., 71 (4), 291-302 [in Russian].
65. Pozdeev, V.B. (1998). Ob opredelenii geoekologii. [On the definition of geoecology]. Geografiya i prirodnye resursy.
[Geography and natural resources]. Novosibirsk, 1, 150-155 [in Russian].
66. Rustamov, S.G., Mardanov, I.E. (1986). Ob opolznevyh selyah Yugo-Vostochnogo Kavkaza [About landslide mud-
flows in the South-Eastern Caucasus]. V sb.: Problemy protivoselevyh meropriyatij [In: Problems of anti-mudflow
measures]. Alma-Ata, 90-94 [in Russian].
67. Fomenko, I.K., Pendin, V.V., Nguyen, Ch.K. (2017). Ocenka ushcherba, opasnosti i riska ot opolznevyh processov (na
primere Severo-Zapadnogo V'etnama) [Assessment of damage, danger and risk from landslide processes (on the ex-
ample of Northwest Vietnam)] Materialy dokladov XIII Obshcherossijskoj nauchno-prakticheskoj konferencii i
vystavki «Perspektivy razvitiya inzhenernyh izyskanij v stroitel'stve v Rossijskoj Federacii» [Proceedings of the XIII
All-Russian scientific-practical conference and exhibition "Prospects for the development of engineering surveys in
construction in the Russian Federation"], 27-34 [in Russian].
68. Chalkova, Yu.S., Cherepanov, B.M. (2007). Opolznevye processy, ih prognozirovanie i bor'ba s nimi [Landslide pro-
cesses, their forecasting and control]. Polzunovskij Vestnik [Polzunovsky Bulletin], 1–2, 80-89 [in Russian].
69. Shirinov, N.Sh. (1982). Morfostrukturnye osobennosti rajona Ismaillinskogo zemletryaseniya (Azerbajdzhanskaya
SSR) [Morphostructural features of the area of the Ismayilli earthquake (Azerbaijan SSR)] Izv. AN Azerb. SSR, seriya
nauk o Zemle [Proceedings of AS of Azerbaijan. SSR, Earth Sciences Series], 5, 9-13 [in Russian].
70. Ekologicheskij risk [Ecological risk]. Materialy Vtoroj Vseros. Konferencii [Materials of the Second All-Russian.
Conferences]. Irkutsk: Publishing House of the Institute of Geography SB RAS, 2001, 262 [in Russian].
AuthorsContribution:Allauthorshavecontributedequallytothiswork
Застосування кількісних методів для оцінки
стійкості до зсувів басейну річки Агсучай
Стара Тарiхазер 1,
д. геогр. н., доцент, гол. наук. співробітник,
1 Інститут географії імені академіка Г.А. Алієва МОН Азербайджану,
просп. Г. Джавіда, 115, Баку, AZ 1143, Азербайджан;
Сеймур Мамедов 2,
к. геогр. н., пров. інженер, 2 ВП «Азнефть», SOCAR,
пр. Нефтяників, 73, Баку, AZ 1000, Азербайджан;
Зернура Гамідова 1,
д. філософії (географія), доцент, пров. наук. співробітник
Азербайджан робить багато зусиль для зменшення впливу небезпечних геологічних процесів на природні
геосистеми, проте вони все ще завдають величезної шкоди. Більшою мірою до таких процесів схильний регіон
Великого Кавказу, а саме південний схил, де зустрічається весь спектр небезпечних геологічних процесів: земле-
трус (7-8 б і вище), обвали, зсуви, осипи, селеві потоки та ін. Усі вони є масштабними за збитками процесами –
впливають на значні площі, призводять до економічних втрат. Мета дослідження – виявити основні чинники фо-
рмування та поширення зсувів у басейні однієї з найбільш селеносних річок не тільки Азербайджану, а й Півден-
ного Кавказу – р. Агсучай, виявити умови їх утворення, дати оцінку ризику вразливості території до зсувних
процесів, а також способи запобігання та захисту. Для оцінки зсувної вразливості та створення карт потенційного
розвитку зсувів у басейні р. Агсучай нами було використано метод співвідношення частотностей (Frequency Ratio
method – FR). На прикладі басейну р. Агсучай для мінімізації збитків від зсувів було проведено детальне вивчення
факторів (гіпсометрія, кути нахилу (крутість схилів), експозиція схилів, геологічна будова (літологія), відстань
від розломів, середньорічна кількість опадів, відстань до ерозійної мережі, відстань до доріг та землекористу-
вання), що визначають розвиток зсувних процесів з урахуванням механізму їх розвитку, а також аналіз отриманих
значень зсувної вразливості та потенційного їх розвитку. Для цього в програмному середовищі ArcGIS за допо-
могою інструменту просторового аналізу «Калькулятор растру» підсумувавши кожен фактор утворення зсувів,
перемножені на свою вагу, була отримана карта зсувної вразливості басейну р. Агсучай. Використовуючи метод
класифікації природних кордонів у програмному середовищі ArcGIS, район дослідження був поділений на п'ять
зон за потенціалом розвитку зсувів: дуже низький, низький, середній, високий та дуже високий. В кінцевому
підсумку достовірність отриманих моделей була оцінена із застосуванням AUC ROC (площа під кривою помилок)
аналізу, який показав високу результативність (82%) методу, що використовується.
Ключові слова: зсув, селеносна річка, геосистема, туристично-рекреаційний потенціал, збитки, зсувна не-
безпека, вразливість, кількісні методи
Внесокавторів:всіавторизробилирівнийвнесокуцюроботу   Надійшла10квітня2023р.
Прийнята16травня2023р.