Understanding the Factors Contributing to Landslide Hazards in Sheko, Ethiopia

doi:10.21203/rs.3.rs-5281124/v1

This study was conducted in Sheko District, Bench Sheko Zone, and Southwestern Ethiopia Regional State. The region has been experiencing recurrent landslide problems for several years. The primary objective of the study was to evaluate landslide hazard in the study area. To achieve this: the Slope Susceptibility Evaluation Parameter rating scheme was used.

The parameters considered in this study included slope geometry, slope material, structural discontinuities, land-use and land-cover, groundwater depth, seismicity, rainfall, and human activities. In addition, an inventory of past landslides in the study area was prepared.

Facet-wise observations were made thematic layers for past landslides, slope facets, and causative factors were created in a GIS environment using secondary data, field observations, topographical maps, and satellite images. Hazard index value was first calculated for each parameter class and ratings for each parameter class were assigned based on the prevailing conditions on individual facets.

The results showed that the landslide hazard map produced by the Slope Susceptibility Evaluation Parameter approach categorized the study area into Very High hazard, High hazard, Moderate hazard, and Low hazard zones with areal coverage’s of 2.372 km², 48.198 km², 54.75 km², and 0.015 km², respectively.

The LHZ map produced by the SSEP approach demonstrated approximately 100% validation with past landslide activities.

Landslide evaluation

Landslide hazard zonation

Slope instability

Hazard index

Landslide hazard can be defined as the likelihood of a potentially damaging event (landslide) occurring within a specific area and period (Raghuvanshi, 2019). Landslides occur when rocks, debris, and soil move down a slope – also known as landslips – either suddenly or gradually over time due to natural causes or human activities. The driving forces behind slope instability are predominantly rainfall, seismic activity, and human actions (Schuster et al., 1996). Landslides can involve various types of movements such as flowing, sliding, toppling, and falling individually or in combination (Schuster et al., 1996).

Most of the people of the high lands and ragged topographic setting of Ethiopian especially in the countryside are living at the toe and sloppy side of the mountain leaving the plain land for agricultural purposes. In towns also due to the shortage of suitable lands for building construction as a result of the increasing population at an alarming rate, and increased cost of land, people are forced to live in much more hazardous areas. The same is true in the study area; this has increased the risk of Landslide by increasing susceptibility, which means the degree of loss to a given element or elements from the occurrence of natural phenomena of a given magnitude (Anbalagan, 1992).

In the study area landslide event had occurred in different localities. The landslide event damaged the main Gravel road, agricultural land which were used for cultivation crops and Coffee degraded by the event, and residential houses were also affected. Furthermore, this landslide has resulted in considerable economic losses to the local community, infrastructures, and environment. Due to this evaluation of the factors influencing landslide hazard initiation and occurrence and conducting landslide hazard zonation is important to recognize the spatial probability of landslides for safer strategic planning of future developmental activities and to minimize the upcoming impact.

The main objective of the present study was to produce a potential landslide hazard map of the study area. the general methodology includes study area classified into facets (a land unit with more or less uniform slope geometry in terms of slope inclination and slope direction), landslide inventory mapping following, data collection on intrinsic and external causative parameters, hazard index calculation, assigning of rating values of all causative parameters for individual facets, and finally rating summed up to obtain an Evaluated landslide hazard (ELH) value which indicates the net probability of instability and potential landslide hazard map was produced.

2.1. The study area

The study area is located in Southwestern Ethiopia, Southwestern Ethiopia regional state, Bench-Shoko zone, Shoko district and this study is conducted in 105\(\:.335\:{km}^{2}\) areas. The study area is located in the UTM (zone 37) and it is bounded by the geographic coordinates of \(\:{35}^{0}{26}^{{\prime\:}}{0}^{{\prime\:}{\prime\:}}\)E-\(\:{35}^{0}{34}^{{\prime\:}}{0}^{{\prime\:}{\prime\:}}\)E and \(\:{7}^{0}{0}^{{\prime\:}}{0}^{{\prime\:}{\prime\:}}\)N-\(\:{7}^{0}{5}^{{\prime\:}}{0}^{{\prime\:}{\prime\:}}\)N (Fig. 1).

According to the Ethiopian metrological institute, the climate of the study area falls in tropical (Kolama) with GPS reading between 1850M-1056 above sea level. The hottest month is February, with a mean maximum temperature of \(\:{33.3}^{0}\) c and a mean minimum temperature of 12⁰c; the coldest month is August which has a mean monthly temperature of 22⁰c and the corresponding mean monthly maximum and minimum values being 22.65⁰c and 17.5⁰c.

2.2. Methodology

In the present study Slope stability Susceptibility Evaluation Parameter (SSEP) was followed. Slope geometry (relative relief and slope morphology), slope material, structural discontinuity, land use/land cover, and groundwater are intrinsic causative parameters whereas rain induce manifestations, rainfall, seismicity, and manmade activity are the external causative parameters that are considered in this method.

In this study potential landslide hazard of the study area was carried out in the following order: First, the area was classified into a land unit with more or less uniform slope geometry in terms of slope inclination and slope direction. Secondly, thematic map of each causative factor has to be created. Secondly, conversion of thematic maps into raster-based data was carried out to get the pixel counts of past landslide occurred areas and sliding does not occur areas. Thirdly, based on the class distribution and the landslide density, respective weights derived. Fourthly, based on the results the hazard index was calculated and normalized by dividing by the highest hazard index value; this step is essential because it helps to know better in what condition sliding can occur and the possibility of nonoccurrence.

After the hazard index value calculation, the rating values of all causative parameters for individual facets are summed up to obtain an Evaluated landslide hazard (ELH) which indicates the net probability of instability. Raghuvanshi et al., (2014) proved the rationality of considered governing parameters, the adopted Slope stability Susceptibility Evaluation Parameter (SSEP) technique, tools, and procedures in developing the landslide hazard zonation mapping. The advantage of using this for landslide hazard evaluation and zonation is that it uses detailed facet-wise field-based data as input, it is simple, applicable, and time and cost-effective for larger study areas. Finally LHZ map was prepared.

2.3. Data collection and analysis

In this study secondary data and required field materials were gathered and prepared. The following materials were gathered from their respective organizations: Topographic map at a scale of 1:50,000 an aerial photograph from an Ethiopian maps agency, climate data (rainfall and temperature) from an Ethiopian metrological institute, and field equipment including GPS, compass, and geological hammer from Mizan Tepi university department of Geology. Thematic maps of the following three causative factors were derived from DEM (digital elevation model) and Blender GIS 2.3 data of the study area, i.e. slope map, elevation map, and aspect map. Landslide inventory map, soil and rock mass map, and land use/land cover maps were modified and downloaded from previous works on the area and with the help of Google Earth and satellite images respectively. Before field data collection begins Facet-wise classification of the study area was carried out with the help of ARGIS10.3 and Blender2.93 software.

The overlapped map that contains a topographic map together with a facet demarcation map was helpful to easily demarcate facets in the field area. Facet-wise observation and measurement of field data of slope material, land-use and land cover, groundwater-surface manifestations, structural discontinuity, rain-induced manifestations, manmade activities, and past landslide inventory, soil samples from every facet, were collected. Topographic map, Measuring Meter, Clinometers, Geological hammer, GPS, and Compass were very helpful to identify facets, to take measurements of structural discontinuities, knowing the dipping direction of the slope, measuring the strength of rock units, demarcating the exact location of an area during inventory landslide mapping, and for direction respectively.

2.4. Landslide Inventory mapping

The causative factors that led to previous slope failures regularly give clues to the areas and prerequisites of future slope failures; therefore, stock maps provide beneficial statistics about the possible future land sliding (Dai & Lee, 2001). During this study inventory maps were prepared in particular by using geomorphic analysis of Google Earth images, aerial photographs, distinctive subject surveying had been used to demarcate landslides and associated mass moves in the current study area.

Google Earth images with 60 − 30 meter resolution, satellite image analysis up to 10 meter resolution, and using area investigation at a scale of 1:10,000. In the study a total of 14 landslide events were identified (a) two Debris flow, (b) 4 Complex (combination of earth flow and rock flow, (c) 1 Earth flow (soil creep), (d) 6 Earth spread, (e) 1 Rock fall had been identified (Fig. 3).

2.5. Landslide Triggering Parameters and their distribution

The factors that affect slope stability are classified into two classes; intrinsic and external factors Raghuvanshi et al. (2014). Intrinsic factors are inherent parameters that affect slope stability, whereas External parameters are parameters from the external environment.

2.5.1 External Causative Parameters and their Distribution

2.5.1.1 Seismicity

According to the Ethiopian Seismic Risk Map created by Laike Mariam in 1986, the study area is classified as less than five (< 5) on the moment magnitude (M.M) scale, which indicates likely low ground acceleration with a return period of 100 years and a probability of 0.99.

2.5.1.1. Rain Induced manifestations on slope

Gully erosion, no rain-induced manifestations, slope toe erosion, and stream bank erosion are the sub-classes that were identified during the fieldwork and through different secondary data sources. As it is indicated in the table (Table 1) No Rain induced manifestation on a slope, Stream bank erosion, Slope Toe erosion, and Gully erosion have the highest to lowest areal coverage with 64.28%, 24.46%, 8.80%, 2.44%, respectively (Fig. 2).

Table 1

, Hazard index value and corrected normalized value of all causative factors
Type	Count	Landslide occurred (a)		Landslide not occurred (b)		Hazard index (a/b)	Hazard Normalized to 1 (N)	SSEP Original rating	Revised value	Corrected Value
Type	Count	Count	Ratio	Count	Ratio	Hazard index (a/b)	Hazard Normalized to 1 (N)	SSEP Original rating	Revised value	Corrected Value
Discontinuity orientation with respect to slope
Soil Mass	25647	2417	24.17	23230	232.3	0.1	0.59	0.25	0.15	0.15
More than one discontinuity dipping into the hill	3186	459	4.59	2727	27.27	0.17	1	0.25	0.25	0.25
At list discontinuity dipping into the hill	434	12	0.12	422	4.22	0.03	0.18	0.12	0.02	0.02
Total	29267	2888
Land-Use Land-Cover
Thickly Vegetated	11053	1880	65.1	9173	34.77	1.87	0.27	0.4	0.11	0.11
Sparsely vegetated	4792	716	24.79	4076	15.45	1.61	0.23	1.2	0.27	0.27
Cultivated Land	8257	0	0	8257	31.3	0	0	0.4	0	0
Moderately Vegetated	4956	201	6.96	4755	18.03	0.39	0.05	0.75	0.04	0.04
Barren land	209	91	3.151	118	0.447	7.04	1	1.5	1.5	1.5
Total	29267	2888	100	26379	100
Man Made Developmental Activity
No Activity	12000	0	0	12000	120	0	0	0	0	0
Moderately cultivated land	205	89	0.89	116	1.16	0.77	0.22	0.15	0.03	0.03
Sparsely cultivated land	7599	14	0.14	7585	75.85	0	0	0.1	0	0
Dense cultivated land	5170	0	0	5170	51.7	0	0	0.25	0	0
Rock mass cut into gentle slope	33	0	0	33	0.33	0	0	0.5	0	0
Soil mass cut into gentle slope	118	0	0	118	1.18	0	0	0.75	0	0
Moderate steep rock mass cut	55	0	0	55	0.55	0	0	0.75	0	0
Steep soil mass cut	2454	1898	18.98	556	5.56	3.41	1	1.25	1.25	1.25
Steep rock mass cut	1539	871	8.71	668	6.68	1.3	0.38	1	0.38	0.38
Moderate steep soil mass cut	94	16	0.16	78	0.78	0.21	0.06	1	0.06	0.06
Total	29267	2888

2.5.1.2. Manmade Developmental Activities

No Activity land, Moderately cultivated, Sparsely cultivated, Dense cultivated, Rock mass cut into gentle slope, Soil mass cut into gentle slope, Moderate steep rock mass cut, Steep soil mass cut, Steep rock mass cut, and Moderate steep soil mass cut, have highest to lowest areal coverage with 46.04, 0.7, 32.8, 17.79, 0.113, 1.117, 0.188, 0.714, 0.222, and 0.321 respectively (Fig. 3).

2.5.1.3.. Rain fall

Most of the rainfall in the study area is concentrated starting from January up to October. Thirty years of data from Ethiopian metrological agency indicates that the study area receives 2356.1 mm annual rainfall which is classified as high according to the SSEP rating parameters (Dinku et al., 2014). Rainfall has a significant influence over the slope stability (Woldearegay, 2013).

2.5.2. Internal Causative Parameters and their Distribution in the study area

Intrinsic parameters are the inherent or static causative parameters that define the favorable or adverse balance situations inside the slope (Keefer, 2000; Raghuvanshi, 2019; Raghuvanshi et al., 2014).

2.5.2.1 Surface traces of groundwater

Surface traces of groundwater such as damp, moist, dripping, flowing, and dry are the parameters that are provided through the SSEP technique. By using these parameters and their respective ratings, a detailed study was conducted to create a surface groundwater manifestations map of the study area (Fig. 4). As indicated in the figure below, the majority of the study area is Flowing 24.71%, Dry 51.77%, Dripping 4.64%, Damp 13.64%, and Wet 5.21%.

2.5.2.2. Slope Morphology

2.5.2.2.1. Relative relief

The slope may be greater inclined for instability if the relative consolation is higher; the essential top of the slope depends on shear strength, density, and bearing functionality of the slope basis. Slope balance generally decreases with the increase in the height of the slope (Anbalagan, 1992; Raghuvanshi et al., 2014).

If slope increase the shear pressure interior toe of the slope will increase as a result of the brought weight. Shear strain is likewise related to the mass of the fabric and the slope attitude. With increasing slope angle, the tangential strain increase which brings about an increase in shear strain because of this reducing its balance (Hoek & Brown, 1997) (Fig. 5).

2.5.2.2.2. Slope

On steeper slopes, the tangential factor of the weight of a mass and the shear stress will grow at the identical time because the perpendicular aspect of the burden decreases (Anbalagan, 1992). The resistance to the downslope motion is preferred by the frictional resistance and concords a quantity of the particles that make up the object (Dai & Lee, 2001).

When the shear strain increased than the combination of forces retaining the item on a slope, the item will tend to move downslope. Thus, the slope of a terrain is the most essential element related to the evaluation of landslides (Hoek & Brown, 1997). The slope map of the study area is extracted using ArcMap10.8 GIS from the digital elevation model prepared using a sentinel 2 data set with a pixel size of 10m (Fig. 5).

2.5.2.3. Slope material

Inside the present study area, as it is revealed in Fig. 6, below a slope material map was once organized using field investigation, satellite images, and Google Earth maps. The slope material map suggests that forty 26.84% of the full area is covered via residual soil, 46.61% with residual expansive soil, 13.45% by Very Weak Rock, 12.77% by Collapsible Soil, and 0.32% Medium strong rock.

2.5.2.4. Soil Cover Depth

Probably the parameter with the best uncertainty in these models is soil depth. At the same time as most studies are renowned for their importance, few studies encompass spatial variability of soil depth within the predictions. Further, in mountainous regions landslides are one of the important approaches shaping the reduction and consequently have a good sized effect on the spatial distribution of soils. The soil cover map of the study area indicates that out of the total area, 33.79% is covered by 20m -15 m thick soil, 0.68% of the area is covered by 14m − 10m, 1.47% of the study area is covered by < 5m thick soil mass, 3.52% is covered with 9m − 5m and 60.53% of the area is covered by > 20m soil mass (Fig. 7).

2.5.2.5. Structural discontinuities and Rock mass condition

The strength of rock mass can be understood as the strength of a whole rock block on the slope, as well as the strength of the entire rock slope. At the outcrop scale, the strength of rock mass can be determined by testing representative rock blocks in the laboratory.

The data from such tests are used in numerical models to estimate, although with some uncertainties, the stability of specific slopes (Clarke & Burbank, 2010; Hoek & Brown, 1997). Defining the quantitative strength of the entire rock slope is challenging. Based on the SSEP approach and using the on-field estimation parameters of the slope stability Susceptibility approach the study area is covered by 99.28% soil mass, and Blocky/ Disturbed material with 0.71% out of the total landmass (Fig. 8).

2.5.2.6. Land Use/Land Cover

The land-use/land-cover map of the study area was prepared by field investigation, and largely with the help of Google earth. Using those sources of information the land-use/land-cover map of the study area is modified and created. The land-use/land-cover of the study area is characterized as Thickly Vegetated, Sparsely vegetated, Cultivated Land, Moderately Vegetated, Barren land there relative distribution in the study area is 37.7%, 16.4, 28.2%, 16.9%, 0.7% respectively (Fig. 9).

2.6. Causative Factors and Past Landslide Distribution

Table 1, Table 2, Table 3, Table 4, and Table 5 displays the percentage of area coverage for each parameter class related to landslide occurrences. For instance, within the Land use Land cover class, the highest occurrence of past landslides is observed in the Barren land subclass, 43.5%. Similar calculations are made for the distribution of past landslides across for each sub-class.

Table 2

, Hazard index value and corrected normalized value of all causative factors
Type	Count	Landslide occurred (a)				Landslide not occurred (b)				Hazard index (a/b)		Hazard Normalized to 1 (N)		SSEP Original rating		Revised value			Corrected Value
Type	Count	Count		Ratio		Count		Ratio		Hazard index (a/b)		Hazard Normalized to 1 (N)		SSEP Original rating		Revised value			Corrected Value
Rain Induced manifestations on slope
Stream bank erosion	7164		313		3.13		6851		68.51		0.05		0.29		0.15		0.04	0.25
Slope Toe erosion	2578		25		0.25		2553		25.53		0.01		0.06		0.25		0.02	0.09
Gully erosion	715		0		0		715		7.15		0		0		0.1		0	0
No Rain induced manifestation	18810		2550		25.5		16260		162.6		0.16		1		0		0	0
Total	29267		2888
Soil cover Depth
20 _ 15 m	9890		981		9.81		8909		89.09		0.11		0.81		1.8		1.45	1.77
14 _ 10 m	199		0		0		199		1.99		0		0		1.4		0	0
< 5m	430		0		0		430		4.3		0		0		0.5		0	0
9 _ 5 m	1031		124		1.24		907		9.07		0.14		1		1		1	1.22
> 20 m	17717		1783		17.83		15934		159.3		0.11		0.82		2.5		2.05	2.5
Total	29267		2888				26379		263.8
Surface traces of groundwater
Flowing	7234		1891		18.91		5343		53.43		0.35		0.41		2		0.83	1.1
Dry	15152		14		0.14		15138		151.4		0.01		0		0		0	0
Dripping	1359		627		6.27		732		7.32		0.86		1		1.5		1.5	2
Damp	3995		49		0.49		3946		39.46		0.01		0.01		0.6		0.01	0.01
Wet	1527		307		3.07		1220		12.2		0.25		0.29		1		0.29	0.39
Total	29267		2888		0		26379
Relative Relief
Very High	6996		1750		17.5		5246		52.46		0.33		0.32		1		0.32	0.61
Low	12712		0		0		12712		127.1		0		0		0.1		0	0
Moderate	5636		12		0.12		5624		56.24		0		0		0.2		0	0
High	2758		1118		11.18		1640		16.4		0.68		0.65		0.8		0.52	1
Medium	1165		8		0.08		1157		11.57		0.01		0.01		0.6		0	0.01
Total	29267		2888		28.88		26379		263.8		0.11		0.1

Table 3

, Hazard index value and corrected normalized value of all causative factors
Type	Count	Landslide occurred (a)		Landslide not occurred (b)		Hazard index (a/b)	Hazard Normalized to 1 (N)	SSEP Original rating	Revised value	Corrected Value
	Count	Count	Ratio	Count	Ratio	Hazard index (a/b)	Hazard Normalized to 1 (N)	SSEP Original rating	Revised value	Corrected Value
SLOPE
Steep slope	5554	1500	15	4054	40.54	0.37	0.35	1.7	0.6	0.6
Very gentle slope	12514	31	0.31	12483	124.8	0	0	0.3	0	0
Gentle slope	5775	220	2.2	5555	55.55	0.04	0.04	0.6	0.02	0.02
Moderately steep slope	3948	382	3.82	3566	35.66	0.11	0.1	1	0.1	0.1
Escarpment/cliff	1476	755	7.55	721	7.21	1.05	1	2	2	2
Total	29267	2888		26379
Slope Material
Residual Soil	7856	0	0	7856	78.56	0	0	0.2	0	0
Residual Expansive Soil	13641	46	0.46	13595	136	0	0	0.6	0	0
Very Weak Rock	3938	1110	11.1	2828	28.28	0.39	0.45	1	0.45	0.45
Collapsible Soil	3737	1732	17.32	2005	20.05	0.86	1	1	1	1
Medium strong rock	95	0	0	95	0.95	0	0	0.5	0	0
Total	29267	2888		26379	263.8
Structural discontinuities and Rock mass condition
Soil mass	25647	2084	20.84	23563	235.6	0.09	0.3	0.25	0.07	0.09
Disintegrated	3525	803	8.03	2722	27.22	0.3	1	0.2	0.2	0.25
Blocky/ Disturbed	95	1	0.01	94	0.94	0.01	0.04	0.15	0.01	0.01
Total	29267	2888	0	26379	263.8	0

Table 4

, past landslide and causative factor distribution
Causative factor type	Pixel Count	%	Past slides	%	Slide not occurred	%
Land use Land cover
Thickly Vegetated	11048	37.7	88	0.8	10960	99.2
Sparsely vegetated	4790	16.4	16	0.33	4774	99.7
Cultivated Land	8265	28.2	0	0	8265	100
Moderately Vegetated	4956	16.9	1	0.02	4955	100
Barren land	209	0.71	91	43.5	118	56.5
Total	29268	100	196
Surface traces of Ground water
Flowing	7234	24.7	60	0.83	7174	99.2
Dry	15153	51.8	103	0.68	15050	99.3
Dripping	1359	4.64	0	0	1359	100
Damp	3995	13.6	33	0.83	3962	99.2
Wet	1527	5.22	0	0	1527	100
Total	29268	100	196
Relative Relief
Very High	6996	23.9	175	2.5	6821	97.5
Low	12712	43.4	1	0.01	12711	100
Moderate	5637	19.3	12	0.21	5625	99.8
High	2758	9.42	0	0	2758	100
Medium	1165	3.98	8	0.69	1157	99.3
Total	29268	100	196
Slope
Steep slope	5554	19	39	0.7	5515	99.3
Very gentle slope	12514	42.8	8	0.06	12506	99.9
Gentle slope	5775	19.7	4	0.07	5771	99.9
Moderately steep slope	3949	13.5	1	0.03	3948	100
Escarpment/cliff	1476	5.04	144	9.76	1332	90.2
Total	29268	100	196
Slope Material
Residual Soil	7856	26.8	8	0.1	7848	99.9
Residual Expansive Soil	13642	46.6	8	0.06	13634	99.9
Very Weak Rock	3938	13.5	115	2.92	3823	97.1
Collapsible Soil	3737	12.8	64	1.71	3673	98.3
Medium strong rock	95	0.32	1	1.05	94	98.9
Total	29268	100	196
Soil Cover deepth
20 − 15 m	9890	33.8	87	0.88	9803	99.1
14 − 10 m	198	0.68	0	0	198	100
< 5m	430	1.47	0	0	430	100
9 − 5 m	1033	3.53	46	4.45	987	95.5
> 20 m	17717	60.5	63	0.36	17654	99.6
Total	29268	100	196

Table 5

, past landslide and causative factor distribution
Causative factor type	Pixel Count	%		%	Slide not occurred	%
Discontinuity orentation with respect to slope
Soil Mass	25648	87.6	84	0.33	25564	99.7
More than one discontinuity dipping into the hill	3186	10.9	111	3.48	3075	96.5
Discontinuity dipping into the hill	434	1.48	1	0.23	433	99.8
Total	29268	100	196
Man made developmenttal activities
No Activitiy	13475	46	103	0.76	13372	99.2
Moderately cultivated land	205	0.7	0	0	205	100
Sparsely cultivated land	9599	32.8	16	0.17	9583	99.8
Dense cultivated land	5206	17.8	0	0	5206	100
Rock mass cut into gentle slope	33	0.11	0	0	33	100
Soil mass cut into gentle slope	327	1.12	0	0	327	100
Moderate steep rock mass cut	55	0.19	16	29.1	39	70.9
Steep soil mass cut	209	0.71	61	29.2	148	70.8
Steep rock mass cut	65	0.22	0	0	65	100
Moderate steep soil mass cut	94	0.32	0	0	94	100
Total	29268	100	196
Rain Induced manifastations on slope
Stream bank erosion	7160	24.5	6	0.08	7154	99.9
Slope Toe erosion	2578	8.81	25	0.97	2553	99
Gully erosion	715	2.44	27	3.78	688	96.2
No Rain induced manifestation on slope	18815	64.3	138	0.73	18677	99.3
Total	29268	100	196
Slope material as rain fall parameter
Soil mass	25648	87.6	84	0.33	25564	99.7
Disintegrated	3525	12	111	3.15	3414	96.9
Blocky/ Disturbed	95	0.32	1	1.05	94	98.9
Total	29268	100	196
Structural discontinuities and Rock mass condition
Soil mass	29058	99.3	166	0.57	28892	99.4
Blocky/ Disturbed	210	0.72	30	14.3	180	85.7
Total	29268	100	196

2.7. Hazard index computation

The hazard index for each parameter class reflects its relative importance in past landslides. It is calculated as the ratio between the pixel count for areas where landslides occurred and areas where they did not occur within a respective parameter class. Hazard index values greater than 1 indicate a higher probability of landslide occurrence, while values less than 1 indicate a lower probability (Keefer, 2000; Korup, 2008).

The thematic map depicting past landslide distribution is transformed into raster data to obtain pixel counts for each class and subclass of included causative factors. These pixel counts are utilized to assess landslide distribution of causative factor distribution across the study area.

Computed hazard index for each parameter are normalized using Eq. (4.1), the corrected normalized value derived from this process is utilized for evaluating landslide hazard zonation within the study area.

\(\:Hin=\frac{Hi}{{H}_{maxj}}*1\) ……………………………………………………………..Eq. (4.1)

Where; Hin is the normalized hazard index, Hi_maxj is maximum hazard index value within parameter (j).

2.8. Landslide hazard risk zonation in the study area

For the present study ten prominent causative factors were considered for the evaluation of LHZ (landslide hazard zonation); six from internal causative factors and four from external causative factors. The internal causative factors are; (1) Slope material (Soil and Rock), (2) Relative Relief, (3) Slope, (4) Structural Discontinuity, (5) Land use/land cover, and (6) Groundwater surface traces. From the external causative factors; (7) manmade developmental activities, (8) Rain Induced Manifestations, (9) seismicity, and (10) Rainfall (Anbalagan, 1992; Ayalew & Yamagishi, 2004; Clarke & Burbank, 2010; Korup, 2008; Larsen, 2008; Raghuvanshi et al., 2014) The basic assumption here is that the combination of these factors might have resulted in the triggering of past landslides in the present study area. The causative factors have their relative contribution to the instability of the slope.

During the post-field investigation all the causative factors thematic maps were produced, and converted into raster format. Using the raster format of every causative factor and integrating it with past landslides in the study area the density count for every thematic map was carried out. By using the pixel count of the causative factors and by integrating with past landslides in the study area calculation of the hazard index for every causative factor carried out.

Finally, by using the ratings that are calculated using with the SSEP approach landslide hazard zonation map was prepared. For this, each of the intrinsic and causative maps geo-processed with the facet map in a GIS environment; so that within each facet it can be deduced what parameter classes are present. Further, within each facet, all SSEP ratings for various parameter classes were sum up to get an evaluated landslide hazard value. Based on these values every facet was classified into various landslide hazard classes as proposed by Raghuvanshi et al (2014).

Based on the evaluation of landslide hazard (ELH) values derived from the combination of intrinsic and external factors, the landslide hazard in the area can be classified into five zones. These include a very high hazard (VHH) zone with an evaluation of landslide hazard value greater than 12, a high hazard (HH) zone with value ranging from [8 to 12), a moderate hazard (MH) zone with value ranging from (8 to 5], a low hazard (LH) zone with value ranging from (5 to 2], and finally a very low hazard (VLH) zone with an ELH value less than 2 (Raghuvanshi et al., 2014).

The landslide hazard map produced has delineated the entire study area into four landslide hazard zones; Low hazard, Moderate hazard, High hazard, and Very high hazard Zones. About 0.014% of the area is covered by Low hazard zones whereas, 51.98% is covered by Moderate hazard zone, 45.756% High Hazard Zone, and the remaining 2.25% very high hazards Zone (Raghuvanshi et al., 2014).

This study evaluates landslide hazard in Sheko district using the SSEP approach. The area was divided into four zones based on their hazard levels: Low Hazard (LH), Moderate Hazard (MH), High Hazard (HH), and Very High Hazard (VHH) (Fig. 10).

Low Hazard (LH), covering 0.014% of the study area, is characterized by moderate vegetation, dry and residual soil; very gentle slopes, low relative relief, and no rain-induced manifestations. Moderate Hazard (MH) 51.98% encompasses areas with thick vegetation, cultivated land, and slopes ranging from very gentle to moderately steep, collapsible soil or weak rock, and generally no rain-induced manifestations. High Hazard (HH) covering approximately 45.756% of the study area, this zone is prevalent in central, northern, southern, and western regions. It is characterized by high to very high relative relief, steep slopes (often due to human activities), and surface indications of groundwater. Very High Hazard (VHH) this zone, comprising about 2.25% of the study area, is primarily located in southern and western parts. It exhibits high to very high relative relief, steep slopes and cliffs, and strong evidence of groundwater.

The majority of the study area falls within the Moderate Hazard zone and High Hazard zone, indicating a significant risk of landslides. Human activities such as soil mass cutting contribute to increased landslide hazard in certain areas. Groundwater conditions play a crucial role in determining landslide susceptibility. Understanding the characteristics of each hazard zone is essential for effective risk management and mitigation efforts (Fig. 10).

4.1 Validation of LHZ map prepared by SSEP approach

To check the validation of the LHZ map prepared by the SSEP approach an overlay analysis in a GIS environment was performed. The past landslide activity map was overlaid on the LHZ map prepared by the SSEP approach. The analysis revealed that out of 14 past landslides 11 (78.57%) past landslides fall within the high-hazard zone and about 3 (21.43%) fall within the very high-hazard zone (Fig. 10). The results gave fitting validation of 100% of the past landslides fall within the high and very high hazard zones.

The objective of this study was to evaluate landslide hazard risk within the Sheko district, an area situated in the Southwestern Ethiopia Regional State. The study area, encompassing 105.335 square kilometers, is characterized by a rugged topography with a maximum elevation of 1850 meters. The region experiences a tropical rainforest climate, with February as the warmest month (mean maximum 33°C, mean minimum 1°C) and August as the coldest (mean maximum 22.6°C, mean minimum 17°C).

A comprehensive methodology incorporating literature review, field surveys, data acquisition, and analysis of contributing factors was employed to assess slope stability. To create landslide hazard zone maps, the area was divided into 303 slope facets, and data for ten (10) causative parameters classes including land use/land cover, slope morphometric, slope material, surface traces of groundwater, mean annual rainfall (mm), rain-induced manifestation on a slope, manmade developmental activities, structural discontinuity and rock mass condition, soil cover depth, seismicity activities were collected through field observations and other sources.

The landslide inventory mapping was created through field investigations and by using Google Earth images. The distribution of landslides across each factor map was analyzed within a GIS environment. Modified ratings for each class within these factor maps were derived using the hazard index value as a guide.

During the field investigation, a total of 14 past landslide events were recorded and categorized into five types. The inventory provided information on the location, dimensions, type of failure, and materials involved in the landslides.

By establishing a density relationship between past landslide occurrences and each parameter class hazard index value was calculated. Ratings were then assigned based on observations for each parameter on each facet to determine the evaluated landslide hazard (ELH) value which formed the basis for generating Landslide hazard zone maps using both the SSEP approach.

The SSEP method classified a very small portion of the area (0.014%) as low-hazard zones, 51.98% of the area as a moderate hazard zone, 45.756% of the area as a high-hazard zone and 2.25% as a very high-hazard zone.

The analysis of the overlap between past landslides and the LHZ map created using the SSEP approach showed that out of 14 past landslides, 11 (78.57%) past landslides fall within the high-hazard zone and about 3 (21.43%) fall within the very high-hazard zone.

The results gave satisfactory validation as 100% of the past landslides fall within the high and very high hazard zone.

DEM: Digital Elevation Model

EIA: Environmental Impact Assessment

ELH: Evaluated Landslide Hazard

ENE-WSW: East North East – West South West

GIS: Geographic information system

GPS: Global positioning system

HH: high hazard

Km: Kilometers

LH: low hazard

LHZ: landslide hazard zonation

MER: Main Ethiopian Rift

MH: moderate hazard

SSEP: Slop Stability Susceptibility Evaluation Parameter

UTM: Universal Transfer Mercator

VHH: very high hazard

VLH: very low hazard

Acknowledgement

The author would like to express his sincere gratitude to Mizan Tepi University for providing important field material support that made this research possible. Additionally, i extend our thanks to the Sheko District Administration for their invaluable assistance during our fieldwork. I am also deeply grateful to the local community of Sheko District for their generous support and cooperation throughout the research process.

Disclosure of Funding

“The author declares that no funds, grants, or other support were received during the preparation of this manuscript.”

Declaration of interest’s
The author declare that he has no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Anbalagan, R. (1992). Landslide hazard evaluation and zonation mapping in mountainous terrain. Engineering Geology, 32(4), 269–277. https://doi.org/10.1016/0013-7952(92)90053-2
Ayalew, L., & Yamagishi, H. (2004). Slope failures in the Blue Nile basin, as seen from landscape evolution perspective. Geomorphology, 57(1–2), 95–116. https://doi.org/10.1016/S0169-555X(03)00085-0
Clarke, B. A., & Burbank, D. W. (2010). Bedrock fracturing, threshold hillslopes, and limits to the magnitude of bedrock landslides. Earth and Planetary Science Letters, 297(3–4), 577–586. https://doi.org/10.1016/j.epsl.2010.07.011
Dai, F. C., & Lee, C. F. (2001). Frequency–volume relation and prediction of rainfall-induced landslides. Engineering Geology, 59(3–4), 253–266. https://doi.org/10.1016/S0013-7952(00)00077-6
Dinku, T., Hailemariam, K., Maidment, R., Tarnavsky, E., & Connor, S. (2014). Combined use of satellite estimates and rain gauge observations to generate high‐quality historical rainfall time series over Ethiopia. International Journal of Climatology, 34(7), 2489–2504. https://doi.org/10.1002/joc.3855
Hoek, E., & Brown, E. T. (1997). Practical estimates of rock mass strength. International Journal of Rock Mechanics and Mining Sciences, 34(8), 1165–1186. https://doi.org/10.1016/S1365-1609(97)80069-X
Keefer, D. K. (2000). Statistical analysis of an earthquake-induced landslide distribution—The 1989 Loma Prieta, California event. Engineering Geology, 58(3–4), 231–249. https://doi.org/10.1016/S0013-7952(00)00037-5
Korup, O. (2008). Rock type leaves topographic signature in landslide‐dominated mountain ranges. Geophysical Research Letters, 35(11), 2008GL034157. https://doi.org/10.1029/2008GL034157
Larsen, M. C. (2008). Rainfall-triggered landslides, anthropogenic hazards, and mitigation strategies. Advances in Geosciences, 14, 147–153. https://doi.org/10.5194/adgeo-14-147-2008
Moore, D. R., Robinson, M. J., Fry, J. L., Tang, J. E., Glover, E. I., Wilkinson, S. B., Prior, T., Tarnopolsky, M. A., & Phillips, S. M. (2009). Ingested protein dose response of muscle and albumin protein synthesis after resistance exercise in young men. The American Journal of Clinical Nutrition, 89(1), 161–168. https://doi.org/10.3945/ajcn.2008.26401
Raghuvanshi, T. K. (2019). Plane failure in rock slopes – A review on stability analysis techniques. Journal of King Saud University - Science, 31(1), 101–109. https://doi.org/10.1016/j.jksus.2017.06.004
Raghuvanshi, T. K., Ibrahim, J., & Ayalew, D. (2014). Slope stability susceptibility evaluation parameter (SSEP) rating scheme – An approach for landslide hazard zonation. Journal of African Earth Sciences, 99, 595–612. https://doi.org/10.1016/j.jafrearsci.2014.05.004
Schuster, R. L., NietoThomas, A. S., D. O’Rourke, T., Crespo, E., & Plaza-Nieto, G. (1996). Mass wasting triggered by the 5 March 1987 ecuador earthquakes. Engineering Geology, 42(1), 1–23. https://doi.org/10.1016/0013-7952(95)00024-0
Tucker, G. E., & Slingerland, R. (1996). Predicting sediment flux from fold and thrust belts. Basin Research, 8(3), 329–349. https://doi.org/10.1046/j.1365-2117.1996.00238.x
Woldearegay, K. (2013). Review of the occurrences and influencing factors of landslides in the highlands of Ethiopia: With implications for infrastructural development. Momona Ethiopian Journal of Science, 5(1), 3. https://doi.org/10.4314/mejs.v5i1.85329

No competing interests reported.

Understanding the Factors Contributing to Landslide Hazards in Sheko, Ethiopia

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods