Abstract

This study evaluates the land use and land cover (LULC) dynamics that play an indispensable role in the degradation and deterioration of soil and water quality affecting the natural resources throughout the Kulfo watershed in Ethiopia. Directed image classification is initiated for satellite images to study the watershed. The image classification is categorized into ten different LULC classes with validation of ground control points. A Revised Universal Soil Loss Equation (RUSLE) model was used to generate the average soil loss of the watershed. The model involves the Rainfall Erosivity factor (R), Soil Erodibility factor (K), Length and Slope factor (LS), Cover Management factor (C), and Support Practice factor (P). The dynamics of LULC change and rainfall erosivity over the past 30 years have been interpreted using maps from 1990, 2005, and 2020 using the C-factor and R-factor. The remaining factors, like
K-factor, LS-factor, and P-factor, were kept constant over the period. The results reveal that the average annual soil loss rate (A) of the watershed is estimated to be
138.8 t ha^-1, 161.2 t ha^-1, and 173.25 t ha^-1 per year, for the selected period intervals. During the past three decades, the soil loss rate in the watershed has increased by 34.4 t ha^-1 per year. The watershed and sustainable soil and water conservation practices need special attention to mitigate the severity of soil erosion risks to avoid disaster.

1 Introduction

Soil erosion, a gradual process involving the detachment and removal of soil particles by water or wind impact, leads to soil degradation and deterioration (Al-Kaisi 2000). The impact of LULC dynamics on soil erosion has garnered global attention due to its significant role in disrupting Earth's ecosystem (Kidane et al. 2019; Polykretis et al. 2020). Land degradation, soil deterioration, water quality issues, and the decline in natural resources are among the consequences of LULC dynamics (Hasan et al. 2021). Such environmental changes play a crucial role in sustainable natural resources management and land use planning (Kidane and Alemu 2015).

Recent studies have highlighted the predicted impact of land-use changes on soil erosion, particularly in regions experiencing rapid LULC transformations, making them susceptible to erosion risks (Tadesse et al. 2017; Esa et al. 2018; Abdulkareem et al. 2019; Aneseyee et al. 2020). The country's overall soil loss is estimated to be approximately 1.5 billion tons per year (Negese et al. 2021). Over the past few decades, soil loss has varied across different land types, with croplands experiencing an average erosion rate of 42 t ha^-1, and extreme cases reaching up to 300 t ha^-1 (Tsegaye 2019). The high potential for rainfall in many watersheds within the central Rift Valley River basin makes them susceptible to severe soil erosion (Awdenegest and Holden 2007).

Various studies have utilized the Revised Universal Soil Loss Equation (RUSLE) Model to assess the impact of LULC changes on soil erosion in different watersheds (Abdulkareem et al. 2019; Näschen et al. 2019; Kogo et al. 2020). These studies have not only examined the effect of LULC dynamics on soil erosion risks but have also tested the repeatability of the RUSLE model in capturing these dynamics (Desta and Fetene 2020; Kurwadkar et. al. 2022). The model's performance has been evaluated across the entire country, with recommendations for its application in such analyses (Kayet et al. 2018, Kassie et al. 2020). This study aims to apply the RUSLE model to evaluate the impact of LULC dynamics on soil erosion risks in the Kulfo watershed of the Rift Valley River basin, which will provide valuable insights for sustainable land management and conservation practices.

2 Materials and methods

2.1 Description of the study area

The Digital Elevation Model (DEM), with a cell size resolution of 20 m, derived from the Shuttle Radar Topographic Mapper Model (SRTM) and produced by the U.S. Geological Survey (USGS), was utilized to determine the location and boundaries of the Kulfo watershed. The watershed is located between 37^o20′30′′ and 37^o37′0′′E longitude, and 5^o55′0′′ to 6^o15′0′′N latitude, within the southern Ethiopian Rift Valley Lake basin (refer to Figure 1). Encompassing an area of 418 km², the watershed's topography is characterized by an elevation range from 1109 m to 3478 m above mean sea level. The elevation within the Kulfo watershed exhibits a diagonal increasing pattern, with a notable rapid change in elevation observed in the uppermost part of the region. The processed topographic elevation map provides insights into the terrain features and elevation differences present within the watershed, contributing to a better understanding of its landscape characteristics.

Figure 1 Study area map of the Kulfo watershed.

The seasonal migration of the Inter-tropical Convergence Zone (ITCZ) mainly controls the climate of the region. The convergence of northern and southern hemispheric trade winds in the tropics, as well as the associated atmospheric circulation, are also responsible. The rainy season in the area extends from March to May and from July to October, indicating a bimodal rainfall pattern. The mean annual rainfall varies across the river basin from 870 mm to 1420 mm, with a small variation in the mean annual temperature, about 20.2 ^oC, and relatively higher values on the rift floor with a slight decrease in the adjacent highlands (Tsegaye 2019).

2.2 Methods

The study goal was to evaluate the role of LULC dynamics over a set of past periods (1990, 2005, and 2020) on the magnitude of soil erosion risks. The method involved the extraction of three satellite image products and classification of LULC characteristics for different periods, parameterization, and estimation of five universal soil loss equations, in which daily rainfall, soil type, elevation data, and land cover data were obtained from both ground-based and satellite data sources, and the interpretation of LULC dynamics and its role in soil erosion using the spatial distribution over different periods in the Kulfo watershed.

Characterization of Land Use and Land Cover

Three cloud free, land cover satellite images are extracted for the periods 1990, 2005, and 2020 to characterize the LULC of the study area, comprising of conventional supervised satellite image classification (Phinzi et al. 2020). This supervised image classification algorithm involves validation of classified images with existing ground control points and Google Earth images from their respective years. The image classification is carried out using Java and Python scripts over the Google Earth Engine (GEE) interface. This GEE platform is applied and was chosen for its flexible efficiency in capturing the desired land cover class of the given area. More recent studies evaluated the efficiency of GEE for classifying land use land cover changes using the Classification and Regression Tree algorithm (CART). Sidhu et al. (2018) recommended GEE as an excellent tool for accurate classification of land use land cover. Therefore, during this research, GEE was used for classifying the LULC of the Kulfo watershed.

Rate of Soil Loss

The mean annual soil loss rate (A) is determined using a Revised Universal Soil Loss Equation (RUSLE) and expressed in t ha^-1 yr^-1 in a cell-by-cell multiplication of the potential five-layer parameters. RUSLE provides an ideal framework for assessing soil erosion and its factors (Moges and Bhat 2017).

The utilization of empirical models for predicting soil erosion is a widespread practice in the field of environmental research, based on established correlations and models. By employing this model, erosion rates can be accurately estimated, enabling the assessment of the effects of changes in land use and land cover on soil loss. In the specific context of the Kulfo watershed in the Southern Ethiopian Rift Valley, the application of RUSLE for these parameters offers a standardized method for examining erosion processes. This approach not only ensures the reliability of the study's predictions but also enhances the current body of knowledge on soil erosion dynamics in this region. Particularly, RUSLE considers Rainfall (R), Topography (LS), Soil Erodibility (K), Cover Management (C), and Support Practice (P) as potential factors that affect soil erosion (Equation 1) to calculate A (Phinzi et al. 2020).

(1)

Where:

A	=	mean annual soil loss rate (t ha-1 yr-1),
R	=	rainfall erosivity factor (MJ mm ha-1 h-1 year-1), ,
K	=	soil erodibility factor (t h MJ-1 mm-1),
LS	=	slope length and slope steepness factor (dimensionless),
C	=	cover and management factor (dimensionless), and
P	=	support practice factor (dimensionless).

Rainfall Erosivity Factor (R): R is the primary factor causing soil erosion, and accounts for about 85% of degradation. Here, the R-factor developed for the Ethiopian condition is applied, and this technique has also been adopted by several other countries (Belayneh et al. 2019). The Hurni’s equation for estimation of the R-factor is provided in Equation 2.

(2)

Where:

R	=	Rainfall Erosivity Factor (MJ mm ha-1 h-1 year-1), and
AR	=	mean annual rainfall (mm).

The spatial rainfall erosivity of the study area is regionalized by the Kriging technique for corresponding rainfall magnitudes for 1990, 2005, and 2020 (Belay et al. 2019). Due to limited spatiotemporal rainfall data (< 6 years) available from meteorological observations to describe the rainfall dynamics in the past 30 years, alternative satellite rainfall sources are adopted through bias correction (Goshime et al. 2019). Among satellite rainfall products, the Climate Hazards Group InfraRed Precipitation (CHIRP) product was previously verified for its good performance over the Kulfo watershed (Kim et al. 2016). The annual total rainfall for 30 years (1990–2020) is calculated using the bias-corrected satellite rainfall data for every 6 stations.

Soil Erodibility Factor (K): This stands for the inherent susceptibility of soil to erosion by rainwater and runoff as influenced by the biophysical and chemical properties of the soil (Panagos et al. 2015). The K-factor is calculated using William's equation due to its practical significance among other techniques and experimentally verified evidence over different places (Khanchoul et al. 2020; Nut et al. 2021). The equation for K-factor is given in Equation 3.

(3)

Where:

K	=	factor is expressed in terms of t h MJ-1 mm-1,
fcsand	=	factor that lowers the K indicator in soils with high coarse- sand content, and higher for soils with little sand,
fci-si	=	low soil erodibility factors for soils with high clay-to-silt ratios,
forgc	=	reduces K values in soils with high organic carbon content, and
fhisand	=	lowers K values in soils with extremely high sand content.

The Digital Soil Map of the World (DSMW) developed by the Food and Agricultural Organization (FAO) for the Kulfo watershed is used to appropriately determine the magnitudes of the parameters (such as sand fraction, silt fraction, clay fraction, and organic carbon).

Length-Slope Factor (LS): LS is a topographic factor and can be defined as the horizontal distance from the beginning of surface flow to the point of deposit, or the joint influence of slope extent (L) and slope steepness (S) on erosion (Renard et al. 1997; Fu et al. 2021). Equation 4 is used to calculate the adopted experimental equation, as follows:

(4)

Where:

LS	=	Length-Slope factor (%)
Flow Accumulation	=	runoff accumulation number obtained using the Digital Elevation Model (DEM),
Cell Size	=	resolution of the Shuttle Radar Topographic Mapper Model (SRTM) DEM (i.e., 20 m x 20 m), and
sin(Slope)	=	sin of Slope gradient (in degrees).

The appropriate magnitudes of the C-factor values for each of the ten land cover classes were assigned (Bastola et al. 2019). Generally, the magnitudes of the C-value range from 0 to 1.

Cover Management Factor (C): C is defined as the ratio of soil loss from land cropped under specific conditions to the corresponding loss from clean-tilled, continuous fallow. Land cover affects the magnitude and spatial distribution of soil erosion in a given area. For instance, vegetation cover reduces soil erosion by trapping the movement of runoff and by decreasing the surface area vulnerable to raindrop influence.

Support Practice Factor (P): This refers to the effects of conservation practices in reducing the quantity and rate of runoff with respect to the amount of soil erosion. In other words, it indicates the impact of management through the control of runoff, with specific reference to how the management practices (e.g., contour tillage, strip cropping, and terraces) reduce and alter the pattern, direction, and speed of that runoff (Phinzi and Ngetar 2019). Based on the various conservation practices, the magnitude of the P-factor could range from 0 to 1 (Aneseyee et al. 2020), where the value 0 indicates a good manmade erosion-resistant facility, and the value 1 indicates the absence of any erosion-resistant facility.

Severity of Soil Erosion

Identification of erosion-prone areas has valuable benefits for efficient land management and planning for decision and policymakers. Appropriate implementation and intervention measures can be applied. Identification of the spatiotemporal distribution of soil erosion-prone areas is crucial to determine using the severity of the average soil loss rate. The severity classes can be categorized into seven soil loss rates (Tadesse et al. 2017).

3 Result and discussion

3.1 Classification and dynamics of land use and land cover

A supervised analysis approach was employed for satellite image classification to identify 10 distinct major LULC classes. Validation of the land cover classification was carried out using control points established through GPS tracking. A total of 80 control points were utilized for this purpose, arranged in an 8x10 grid pattern to ensure comprehensive coverage and accuracy in correcting the land-sat image.

When selecting control points for validation, the process involved strategically placing these points across the watershed to represent various land cover types and features. To achieve a sufficient level of spatial distribution and representation within the study area, 80 control points were selected. Having a higher number of control points enhances the accuracy and reliability of the land cover classification results by providing more data points for comparison and validation. The selection of 80 control points was deemed appropriate to effectively capture the diversity and variability of land cover within the Kulfo watershed. While a lower number of control points may not adequately represent the complexity of the landscape, an excessive number of control points could lead to redundancy and inefficiency in the validation process. Therefore, the decision to use 80 control points strikes a balance between thorough validation and practicality in the classification process (Figure 2).

Figure 2 Ground control points used to validate LULC classification.

The Kappa coefficient was obtained through the analysis of the confusion matrix, which was generated based on ground validation Landsat images and 40 training points. The Kappa coefficient, along with the overall accuracy, was used to assess the accuracy of the classification. The Kappa coefficient measures the agreement between the classification results and the ground truth data, providing a statistical measure of inter-rater reliability. It considers the possibility of agreement occurring by chance, thereby providing a more robust assessment of the classification accuracy. The high Kappa coefficient values obtained (0.94, 0.88, and 0.89 for three periods) indicate a strong agreement between the classified land cover and the ground truth, reinforcing the reliability and precision of the classification process (Figure 3).

Figure 3 Land use/land cover change map of the Kulfo watershed for the years 1990, 2005, and 2020.

A post-classification change detection algorithm is used to determine changes in LULC. The change detection provides the size and distribution of changed areas (either negative or positive). It also gives the individual percentages of each LULC class. The dynamics of the change on the map and change statistics (gain or loss) were compared to analyze the effect of LULC change on soil erosion. The LULC classes in past decades indicate that there is a drastic reduction in the forest and a radical increase in arable land (Figure 3). On the other hand, there is a significant increase in the riverbank from 1990 to 2005, which could have resulted from an extreme flood event; whereas, from the years 2005 to 2020, there is a high reduction of the riverbank, which might have been related to the expansion of arable lands near the riverbank area.

The percentage change in different land use categories was calculated based on the comparison of land cover proportions between different time periods. In the case of the shrub lands, it was observed that despite already accounting for a significant portion of the watershed (27.5% in 1990, and 30.3% in 2020), there was an increment of 2.8% over the past 30 years. On the other hand, the arable lands experienced substantial increases, with increments of 5.9%, 9.8%, 22%, and 22% from 1990 to 2005 and a notable rise in 2020. This growth in arable land came at the expense of deforestation, reflecting a shifting landscape in the watershed over the years. (Figure 4).

Figure 4 Histogram of percentage changes of LULC classes for the three periods

There was 16.3% coverage of dense forest in 1990; unfortunately, this declined by 3.6% and 3.2% in 2005 and 2020, respectively. Similarly, expansion of urban land cover and bare land areas have shown a cumulative reduction in magnitude in comparison to the forest land cover. There is a linear reduction in the spatial coverage of cropland. Although there has been an increase in spatial coverage of arable land in the past 30 years, cropland reduced from 25.2% in 1990 to 16.7% in 2020. This could be due to an increase in the expansion of pasture land, which was 5% in 1990, and reached 15.4% in 2020.

3.2 Parameterization of soil erosion

The results of soil loss parameters are categorized by Rainfall Erosivity (R), Soil Erodibility (K), Length and Slope (LS), Crop Erosivity (C), and Support Practice (P), and were determined using the RUSLE model.

Rainfall Erosivity (R)

The spatial distribution of Rainfall Erosivity (R) over the Kulfo watershed shows an increasing diagonal (southeast to northwest) pattern for each of the three periods. This diagonal increment pattern compliments the topographical variation as a linear relationship with an increased altitude. This implies that the lower valley land areas had a lower magnitude R-factor than the highland areas for the past 30 years (Figure 5).

Figure 5 Spatial distribution of rainfall erosivity (R) for 1990, 2005, and 2020 (left to right).

On the other hand, the watershed experienced a temporally increasing pattern of R-factor over the past 30 years. This indicates that the intensity of rainfall has an increasing trend over time, and the values range between 360 and 890 (MJ mm ha^-1 h^-1 year^-1).

Soil Erodibility (K)

The Digital Soil Map of the watershed (FAO 1995) contains two major soil types: haplic xerosols (XH), and ochric andosols (TO) (Figure 6). The haplic xerosols soil is dominant over the uppermost part of the watershed, whereas the ochric andosols soil covers the lower part. The spatial pattern of corresponding magnitudes of soil erosivity ranges from 0.016  t h MJ^-1 mm^-1 in the uppermost part, to 0.020 t h MJ^-1 mm^-1 in the lowermost part of the Kulfo watershed. 

Figure 6 Spatial distribution of soil type and the corresponding soil erodibility (K).

Length and Slope (LS)

The slope factor, which is deduced from the digital elevation model, is crucial in determining soil loss. In the study area, the magnitude of LS varies from 0 to 43.50, with a high degree at the escarpment (Figure 7). The temporal variability of the slope factor appears insignificant for the past three decades.

Figure 7 Spatial distribution of length and slope.

Crop Erosivity (C)

Higher magnitude values (0.3–1) of the spatial map are concentrated in the central part of the watershed, which corresponds to the land cover types of pasture land, arable land, and bare land (Figure 8). Dense forests, shrubs, and crop covers have comparatively had lower C-values (0.05, 0.014, and 0.13, respectively). Generally, a relatively smaller magnitude of C-factor is found in the upstream marginal areas, and a larger part in the downstream areas. Additionally, land cover characteristics such as a lake, road, and urban areas have minimum to zero magnitudes. The temporal variations of the C-factor between 1990, 2005, and 2020, and respective magnitude of crop erosivity appear to be increasing.

Figure 8 Spatial distribution of Crop Erosivity (C-factor).

Support Practice (P)

The appropriate magnitude of P-values is derived from the arrangement of the cultivated land, and the observed pattern for most of the cultivated land during the field visit (Figure 9) shows contour farming, strip cropping, or terracing scheme. Thus, corresponding P–values of contour farming are adopted for this watershed.

Figure 9 Images of cultivated land arrangements over five selected points of the Kulfo watershed: Dida, Ako, Wusamo, Zigity Abo, and Bakole (red line indicates the boundary of the watershed).

Similarly, the distributed Support Practice (P) over the Kulfo watershed indicates that a relatively lower value (0.55) is dominantly found over the southern most part (Figure 10). The largest northern most parts have higher values (> 0.8). Moreover, the spatial distribution of the P-factor follows the general slope pattern.

Figure 10 Spatial distribution of Support Practice Factor (P) over the Kulfo watershed.

3.3 Soil Erosion

The estimated average annual soil loss of the watershed is estimated to be 5.5, 6.45, and 6.93 t ha^-1 yr^-1 for 1990, 2005, and 2020, respectively (Figure 11). The soil loss has temporally increased over the past decades. In 1990, the maximum soil loss estimated was about 2002 t ha^-1 yr^-1, and this value increased to 2258 t ha^-1 yr^-1 in 2005, whereas in 2020, it reached up to 2514 t ha^-1 yr^-1.

Figure 11 Spatial distribution of soil loss (t ha^-1 yr^-1), over the Kulfo watershed for the years 1990, 2005, and 2020.

The increasing trend in soil erosion is primarily related to the rainfall erosivity magnitude and the spatial expansion of arable land and pasture land, and a significant reduction of forest lands. The crop erosivity dependency on LULC spatially increased from 1990 to 2020, whereas soil loss of 25 to 60 t ha^-1 yr^-1 is recorded in most parts of the upper portion of the watershed.

The areal average soil loss magnitude ranges from 138.85–173.25 t ha^-1 yr^-1 over the Kulfo watershed. This result indicates that in the past 30 years, the LULC dynamics have predominantly impacted average soil loss up to 34.4 t ha^-1 yr^-1, with a maximum of 512.34 t ha^-1 yr^-1.

Table 1 presents important statistics of soil loss rates over the Kulfo watershed, which include pixel size, average annual soil loss, and maximum and total soil losses with standard deviation for the corresponding three periods.

Table 1 Statistical values of soil loss rate over the Kulfo watershed.

Soil loss rate statistics (t ha-1 yr-1)	Study period				Change between the study period (in 30 years)
	1990	2005	2020
Pixel average Soil loss rate	5.55	6.45	6.93		+1.38
Areal average soil loss	138.85	161.25	173.25		+34.4
Maximum soil loss rate	2002	2258.04	2514.38		+512.34
Total soil loss rate	161,365	187,346.78	201,506		+40,141.17
Standard deviation	24.8	34.11	36.52

N.B: Where (+) indicates the increment of the magnitude between three decades.

The estimated result from the RUSLE model is compared to other similar studies in nearby watersheds, and the areal average soil of the study results are comparable, which confirms the model's reputability in capturing the soil loss.

3.4 Severity of soil erosion prone areas

The magnitude of erosion-prone areas based on the severity of annual average soil loss over the Kulfo watershed are given in Table 2. About 38.7–42.9% of the watershed area is susceptible to low soil loss with an average of 5–15 t ha^-1 yr^-1, whereas 31–32% experience moderate soil loss (with 15–30 t ha^-1 yr^-1). This indicates that nearly 72% of the watershed experiences a severity of slight to moderate soil loss, which indicates a relatively lower priority in terms of land management and conservation practices.

Table 2 Priority area based on soil loss severity over the Kulfo watershed.

No.	Soil Loss Rate (t ha-1 yr-1)	Severity Class	Priority Class	1990 coverage (%)	2005 coverage (%)	2020 coverage (%)
1	0–5	Slight	VII	0.02	0.02	0.02
2	5–15	Low	VI	42.9	40.9	38.7
3	15–30	Moderate	V	31.7	32.2	32.4
4	30–50	High	IV	14.4	11.4	11.5
5	50–100	Very High	III	7.13	10.56	11.14
6	100250	Severe	II	3.22	3.84	4.76
7	> 250	Extreme	I	0.62	1.11	1.47

The remaining 28% of the watershed is highly vulnerable to extreme soil loss, ranging between 30 and 250 t ha^-1 yr^-1. Therefore, the areas identified as severe priority (III, II, and I, respectively) should get attention for land conservation management (Figure 12).

(I, II, III, and IV stand for First, Second, Third, and Fourth priority ranks, respectively)

Figure 12 Priority area based on soil loss severity of sub-watersheds, corresponding to the dominant land use / land cover characteristics.

The spatial distribution of priority areas in terms of erosion susceptibility corresponds to the dominant LULC classes. The uppermost part of the watershed, which is dominated by agricultural land (i.e., arable land, crop land, and pasture land) experiences an estimated average loss rate of 7.81 t ha^-1 yr^-1 (Table 3).

Table 3 Prioritization of erosion-prone sub-watersheds for appropriate management activities.

Sub watersheds	Dominant LULC class	Sub- watershed area (km2)	Minimum soil loss rate (t ha-1 yr-1)	Maximum soil loss rate (t ha-1 yr-1)	Pixel average soil loss rate (t ha-1 yr-1)	Priority rank for management
1	Agricultural	358.2	0	2514.38	7.81	I
2	Urban	16.4	0	267	3.06	III
3	Forest	33.2	2.03	520.17	3.3	II
4	Lake	10.8	0.13	35.2	0.98	IV

Sub-watersheds dominated by forest and urban land covers experience soil loss at a rate of 3.3 t ha^-1 yr^-1 and 3.06 t ha^-1 yr^-1, respectively. The lowermost part of the watershed, dominated by the lake, has the lowest soil loss rate (0.98 t ha^-1 yr^-1). Therefore, this area is less likely to be an erosion prone area. Thus, soil and water conservation practices are highly recommended for the upper part of the watershed (class-I). It can be concluded that soil and water conservation practices can be used as control measures with professional suggestions.

4 Conclusion

The expansion of agricultural practices, overgrazing, and deforestation are mainly driven by population pressure, which determines the spatiotemporal distribution of land use and land cover (LULC). LULC dynamics play a major role in soil loss at global, regional, and watershed scales. This research was conducted to estimate the role of LULC on mean annual soil loss through application of an empirical soil erosion model (such as RUSLE), with integrated GIS in the Kulfo watershed, in the southern Ethiopian Rift valley. The extracted satellite images of the study area were classified into ten different land use classes. A significant increase in arable land was observed over the past three decades, with a drastic reduction in forest cover. Consequences of LULC dynamics resulted in watershed degradation that led to an average soil loss rate of 138.8, 161.25, and 173.25 t ha^-1 yr^-1 in 1990, 2005, and 2020, respectively. The soil loss rate was calculated to illustrate an increase of 34.4 t ha^-1 yr^-1 over the last 30 years. About 28% of the watershed is highly vulnerable to extreme soil loss rates (>30 t ha^-1 yr^-1). The undulating topographic nature and steep slopes in the escarpment and plateau areas partly contributed to severe erosion problems due to rainfall erosivity. Therefore, it is important to prioritize erosion-prone areas (the upper part of the watershed) which require significant soil and water conservation practices. The findings of this study provide sufficient scientific information to regulate land-use design and planning for sustainable development.

References

1. Abdulkareem, J.H., B. Pradhan, W.N.A. Sulaiman, and N.R. Jamil. 2019. “Prediction of spatial soil loss impacted by long-term land-use/land-cover change in a tropical watershed.” Geoscience Frontiers 10 (2): 389–403. https://doi.org/10.1016/j.gsf.2017.10.010
2. Al-Kaisi, M. 2000. “Soil erosion: an agricultural production challenge.” Integrated Crop Management 484 (19): 141–143.
3. Aneseyee, A.B, E. Elias, T. Soromessa, and G.L. Feyisa. 2020. “Land use/land cover change effect on soil erosion and sediment delivery in the Winike watershed, Omo Gibe Basin, Ethiopia.” Science of the Total Environment 728, 138776. https://doi.org/10.1016/j.scitotenv.2020.138776
4. Awdenegest, M., and N.M. Holden. 2007. “Farmers’ perceptions of soil erosion and soil fertility loss in southern Ethiopia.” Land Degradation and Development 18, (5): 543–554. https://doi.org/10.1002/ldr.795
5. Bastola, S., J.Y. Seong, H.S. Lee, Y. Shin, and Y. Jung. 2019. “Assessment of Soil Erosion Loss by Using RUSLE and GIS in the Bagmati Basin of Nepal.” Journal of the Korean Geo-Environmental Society 20 (3): 5–14. https://doi.org/10.14481/jkges.2019.20.3.5
6. Belay, A.S., A.A. Fenta, A. Yenehun, F. Nigate, S.A. Tilahun, M.M. Moges, M. Dessie, et al. 2019. “Evaluation and Application of Multi-Source Satellite Rainfall Product CHIRPS to Assess Spatio-Temporal Rainfall Variability on Data-Sparse Western Margins of Ethiopian Highlands.” Remote Sensing 11 (22): 2688. https://doi.org/10.3390/rs11222688
7. Belayneh, M., T. Yirgu, and D. Tsegaye. 2019. “Potential soil erosion estimation and area prioritization for better conservation planning in Gumara watershed using RUSLE and GIS techniques.” Environmental Systems Research 8 (20). https://doi.org/10.1186/s40068-019-0149-x
8. Desta, H., and A. Fetene. 2020. “Land-use and land-cover change in Lake Ziway watershed of the Ethiopian Central Rift Valley Region and its environmental impacts.” Land Use Policy 96, 104682. https://doi.org/10.1016/j.landusepol.2020.104682
9. Esa, E., M. Assen, and A. Legass. 2018. “Implications of land use/cover dynamics on soil erosion potential of the agricultural watershed, northwestern highlands of Ethiopia.” Environmental Systems Research 7, 21. https://doi.org/10.1186/s40068-018-0122-0
10. Food and Agriculture Organization (FAO). 1995. “FAO/UNESCO Digital Soil Map of the World.” Land and Water Development Division, FAO, Rome, Italy. Available online: http://www.fao.org/geonetwork (accessed on 21 August 2021)
11. Fu, A., Y. Cai, T. Sun, and F. Li. 2021. “Estimating the Impact of Land Cover Change on Soil Erosion Using Remote Sensing and GIS Data by USLE Model and Scenario Design.” Scientific Programming 2021, 6633428. https://doi.org/10.1155/2021/6633428
12. Goshime, D.W., R. Absi, and B. Ledésert. 2019. “Evaluation and Bias Correction of CHIRP Rainfall Estimate for Rainfall-Runoff Simulation over Lake Ziway Watershed, Ethiopia.” Hydrology 6 (3): 68. https://doi.org/10.3390/hydrology6030068
13. Hasan, M.F., M. Nur-E-Alam, M.A. Salam, H. Rahman, S.C. Paul, A.E. Rak, B. Ambade, et al. 2021. “Health risk and water quality assessment of surface water in an urban river of Bangladesh.” Sustainability 13 (12), 6832.
14. Kassie, A.E., D.G. Amare, and G.G. Sulla. 2020. “Soil Erosion Estimation and Watershed Prioritization in Kulfo Watershed, Ethiopia.” Journal of Natural Sciences Research 10 (1). https://doi.org/10.7176/JNSR/10-1-01
15. Kayet, N., K. Pathak, A. Chakrabarty, and S. Sahoo. 2018. “Evaluation of soil loss estimation using the RUSLE model and SCS-CN method in hillslope mining areas.” International Soil and Water Conservation Research 6 (1): 31–42. https://doi.org/10.1016/j.iswcr.2017.11.002
16. Khanchoul, K., F. Balla, and O. Othmani. 2020. “Assessment of soil erosion by RUSLE Model using GIS: A case study of Chemorah Basin, Algeria.” Malaysian Journal of Geosciences 4 (2): 70–78. https://doi.org/10.26480/mjg.02.2020.70.78
17. Kidane, D., and B. Alemu. 2015. “The Effect of Upstream Land Use Practices on Soil Erosion and Sedimentation in the Upper Blue Nile Basin, Ethiopia.” Research Journal of Agriculture and Environmental Management 4 (2): 55–68.
18. Kidane, M., A. Bezie, N. Kesete, and T. Tolessa. 2019. “The impact of land use and land cover (LULC) dynamics on soil erosion and sediment yield in Ethiopia.” Heliyon 5 (12). https://doi.org/10.1016/j.heliyon.2019.e02981
19. Kim, K.B., M. Bray, and D. Han. 2016. “Exploration of optimal time steps for daily precipitation bias correction: a case study using a single grid of RCM on the River Exe in Southwest-England.” Hydrological Sciences Journal 61 (2): 289–301. https://doi.org/10.1080/02626667.2015.1027207
20. Kogo, B.K., L. Kumar, and R. Koech. 2020. “Impact of Land Use/Cover Changes on Soil Erosion in Western Kenya.” Sustainability 12 (22): 9740. https://doi.org/10.3390/su12229740
21. Kurwadkar, S., S.S. Sethi, P. Mishra, and B. Ambade. 2022. “Unregulated discharge of wastewater in the Mahanadi River Basin: Risk evaluation due to occurrence of polycyclic aromatic hydrocarbon in surface water and sediments.” Marine Pollution Bulletin 179, 113686. https://doi.org/10.1016/j.marpolbul.2022.113686
22. Moges, D.M., and H.G. Bhat. 2017. “Integration of geospatial technologies with RUSLE for analysis of land use/cover change impact on soil erosion: a case study in Rib watershed, north-western highland Ethiopia.” Environmental Earth Sciences 76, 765. https://doi.org/10.1007/s12665-017-7109-4
23. Moisa, M.B., D.A. Negash, B.B. Merga, and D.O. Gemeda. 2021. “Impact of Land Use and Land Cover Change on Soil Erosion Using Rusle Model and GIS: A Case of Temeji Watershed, Western Ethiopia.” Journal of Water and Climate Change 12 (7): 3404–3420. https://doi.org/10.2166/wcc.2021.131
24. Näschen, K., B. Diekkrüger, M. Evers, B. Höllermann, S. Steinbach, and F. Thonfeld. 2019. “The impact of land use/land cover change (LULCC) on water resources in a tropical catchment in Tanzania under different climate change scenarios.” Sustainability 11 (24): 7083. https://doi.org/10.3390/su11247083
25. Negese, A., E. Fekadu, and H. Getnet. 2021. “Potential Soil Loss Estimation and Erosion-Prone Area Prioritization Using RUSLE, GIS, and Remote Sensing in Chereti Watershed, Northeastern Ethiopia.” Air, Soil, and Water Research 14 (1): 1–17. http://doi.org/10.1177/1178622120985814
26. Nut, N., M. Mihara, J. Jeong, B. Ngo, G. Sigua, P.V. Prasad, and M.R. Reyes. 2021. “Land Use and Land Cover Changes and Its Impact on Soil Erosion in Stung Sangkae Catchment of Cambodia.” Sustainability 13 (16): 9276. https://doi.org/10.3390/su13169276
27. Panagos, P., P. Borrelli, and K. Meusburger. 2015. “A new European slope length and steepness factor (LS-Factor) for modeling soil erosion by water.” Geosciences 5 (2): 117–126. https://doi.org/10.3390/geosciences5020117
28. Phinzi, K., and N.S. Ngetar. 2019. “The assessment of water-borne erosion at catchment level using GIS-based RUSLE and remote sensing: A review.” International Soil and Water Conservation Research 7 (1): 27–46. https://doi.org/10.1016/j.iswcr.2018.12.002
29. Phinzi, K., N.S. Njoya, O. Ebhuoma, and S. Szabó. 2020. “Comparison of RUSLE and supervised classification algorithms for identifying erosion-prone areas in a mountainous rural landscape.” Carpathian Journal of Earth and Environmental Sciences 15 (2): 405–413. https://doi.org/10.26471/cjees/2020/015/140
30. Polykretis, C., D.D. Alexakis, M.G. Grillakis, and S. Manoudakis. 2020. “Assessment of Intra-Annual and Inter-Annual Variabilities of Soil Erosion in Crete Island (Greece) by Incorporating the Dynamic “Nature” of R and C-Factors in RUSLE Modeling.” Remote Sensing 12 (15): 2439. https://doi.org/10.3390/rs12152439
31. Renard, K.G., F.G. Foster, G.A. Weesies, D.K. McCool, and D.C. Yoder. 1997. “Predicting soil loss erosion by water: A guide to conservation planning with the Revised Universal Soil Loss Equation (RUSLE).” Washington, DC: U.S. Department of Agriculture, Agriculture Handbook No. 703, 404 pp.
32. Sidhu, N., E. Pebesma, and G. Câmara. 2018. “Using Google Earth Engine to detect land cover change: Singapore as a use case.” European Journal of Remote Sensing 51 (1): 486–500. https://doi.org/10.1080/22797254.2018.1451782
33. Tadesse, L., K.V. Suryabhagavan, G. Sridar, and G. Legesse. 2017. “Land use and land cover changes and soil erosion in Yezat Watershed, North Western Ethiopia.” International Soil and Water Conservation Research 5 (2): 85–94. https://doi.org/10.1016/j.iswcr.2017.05.004
34. Tsegaye, B. 2019. “Effect of Land Use and Land Cover Changes on Soil Erosion in Ethiopia.” International Journal of Agricultural Science and Food Technology 5 (1), 26–34. https://doi.org/10.17352/2455-815X.000038