Spatial Assessment of Soil Erosion Risk Using RUSLE Embedded in GIS Environment: A Case Study of Jhelum River Watershed

Abstract

The watershed area of the Mangla Reservoir spans across the Himalayan region of India and Pakistan, primarily consisting of the Jhelum River basin. The area is rugged with highly elevated, hilly terrain and relatively thin vegetation cover, which significantly increases the river’s sediment output, especially during the monsoon season, leading to a decline in the reservoir’s storage capacity. This work assesses the soil erosion risk in the Jhelum River watershed (Azad Jammu and Kashmir (AJ&K), Pakistan) using the Revised Universal Soil Loss Equation of (RUSLE). The RUSLE components, including the conservation support or erosion control practice factor (P), soil erodibility factor (K), slope length and slope steepness factor (LS), rainfall erosivity factor (R), and crop cover factor (C), were integrated to compute soil erosion. Soil erosion risk and intensity maps were generated by computing the RUSLE parameters, which were then integrated with physical factors such as terrain units, elevation, slope, and land uses/cover to examine how these factors affect the spatial patterns of soil erosion loss. The 2021 rainfall data were utilized to compute the rainfall erosivity factor (R), and the soil erodibility (K) map was created using the world surface soil map prepared by the Food and Agriculture Organization (FAO). The slope length and slope steepness factor (LS) were generated in the highly rough terrain using Shuttle Radar Topography Mission Digital Elevation Model (SRTM DEM). The analysis revealed that the primary land use in the watershed was cultivated land, accounting for 27% of the area, and slopes of 30% or higher were present across two-thirds of the watershed. By multiplying the five variables, the study determined that the annual average soil loss was 23.47 t ha⁻¹ yr⁻¹. In areas with dense mixed forest cover, soil erosion rates ranged from 0.23 t ha⁻¹ yr⁻¹ to 25 t ha⁻¹ yr⁻¹. The findings indicated that 55.18% of the research area has a low erosion risk, 18.62% has a medium erosion risk, 13.66% has a high risk, and 11.6% has a very high erosion risk. The study’s findings will provide guidelines to policy/decision makers for better management of the Mangla watershed.

1. Introduction

Soil erosion is a significant global environmental issue that has far-reaching impacts. It reduces soil productivity, causes the loss of essential nutrients, and leads to siltation in water bodies, further exacerbating the problem [1]. Additionally, soil erosion has severe effects on public health and the livelihoods of marginalized communities, especially those reliant on agriculture. Therefore, addressing soil erosion is crucial to mitigate its damaging effects on the environment and society [2]. Soil erosion by water is one of the most common problems, second only to climate change, that requires attention [3,4,5]. Soil erosion poses a significant risk to the northern regions of Pakistan, including AJ&K, due to various factors such as heavy rainfall, land susceptibility, overgrazing, and human activities that contribute to the problem. These activities include overpopulation, deforestation for agriculture, high urban demands, as well as the extraction of wood and raw materials [6,7,8,9].

Mangla and Tarbela watersheds are among the most vital water resources in Pakistan [10]. The Upper Jhelum Basin is the catchment area of the Mangla Reservoir, which is made up of high hills and relatively sparse vegetation. Particularly during the monsoon season (July to September), a large amount of silt is generated in the basin. This is due to heavy rainfall during monsoons, which leads to a high silt yield in the river [11]. Because of its morphology, the basin is exposed to landslides, floods, and earthquakes and is extremely prone to soil degradation [12]. The effects of land erosion on Mangla Dam are significant and far-reaching, impacting crucial areas such as agriculture, waterways, and infrastructure in the country [13,14]. In particular, the physical stress caused by land erosion hinders the development of agriculture in the country, as it undermines the ability of farmers to grow crops and support their livelihoods. This underscores the urgent need to address land erosion to ensure the sustainability of agriculture and other vital sectors [15]. Additionally, the erosion leads to a reduction in water-holding capacity and soil nutrient accumulation, which inhibits the regrowth of vegetation [16,17].

Globally, land erosion leads to an estimated $7 billion in damages to farms annually due to increased flooding [18,19]. This damage extends to public and industrial water storage facilities, water-based recreation areas, and more. Sediments also contribute to increased flood risks and infrastructural damage, affecting roads, dams, buildings, and utility networks [20,21,22,23]. Over the past century, land degradation has intensified, resulting in the loss of approximately twenty-four million tons of productive topsoil from agricultural lands worldwide [24]. According to the United Nations Food and Agriculture Organization, 90% of the Earth’s precious topsoil will likely be at risk by 2050 [25].

Traditional soil erosion evaluation methods based on field surveys are labor-intensive, expensive, and time-consuming [26]. Therefore, a numerical evaluation technique is often desirable for developing regional management strategies to estimate the severity of soil erosion and quantify the amount of soil loss, enabling better management strategies to be developed. This provides an alternative way to study and simulate the long-term and short-term consequences of land use activities on the natural system [27] and to assess other land management strategies in gauged and ungauged basins. Effective use of land resources can also be achieved by developing several land use cases and evaluating their results using soil erosion models [28,29,30]. These models have been developed and used for more than seven decades, with examples including the Water Erosion Prediction Project (WEPP) [31,32,33], the Soil and Water Assessment Tool (SWAT) [31,34], the revised model of Morgan and Finney (RMMF) [35], the Soil Erosion Model for Mediterranean Regions (SEMMED) [31], and the European Soil Erosion Model EUROSEM. There are several levels of complexity in these models, with the most popular empirical model used to estimate soil erosion being the Universal Soil Loss Equation (USLE) [26]. The USLE model was revised and digitized in the early 1990s to produce the Revised Universal Soil Loss Equation (RUSLE), which is widely used to predict soil erosion [36,37,38]. The Universal Soil Loss Equation (USLE) and Revised Universal Soil Loss Equation (RUSLE) [39,40] are mostly used worldwide under a variety of circumstances. The three categories of soil erosion models are empirical, conceptual (partially empirical/mixed), and physical, with USLE and RUSLE being classified as empirical and conceptual (partly empirical) based models [41]. The RUSLE model has been used for more than 80 years due to its good reliability, ease of applicability, high degree of flexibility, correctness, and data accessibility [42,43,44,45,46,47,48]. The current study used RUSLE Model and Geo-Information tools to quantify soil erosion.

The aim of this study is to utilize the revised Universal Soil Loss Equation and geographic information techniques for the following purposes: (i) to calculate the potential soil loss in the Jhelum River watershed, (ii) to create maps of soil erosion risk and severity, and (iii) to identify areas with critical soil erosion conditions that necessitate immediate conservation and land management measures. The study shall help the authorities in the adaptation of adequate measures and best management practices for soil and water conservation in the zones with high erosion risk. The method of the applied RUSLE model in combination with GIS can be transferred at the desired rate of complexity to similar catchments in the south Asian region.

2. Study Area

The Jhelum River basin, as shown in Figure 1, extends from 33°3′ N–35°9′ N and 73°8′ E–75°35′ E and originates in the rugged mountain ridge of the upper part of the Kashmir valley [11]. After the River Indus, the upper Jhelum river is the second-largest tributary of the Indus basin [49]. The watershed’s drainage area is 33,563 km², and the altitude varies from 264 to 6304 m. The entire basin ultimately flows into the Mangla Reservoir, which is Pakistan’s second-largest reservoir, constructed in 1967. The primary function of this dam is irrigation, which serves about 6 million hectares of land. Additionally, the dam serves as a hydropower generator, with a current capacity of 1000 MW, contributing to 6% of the country’s overall installed hydropower generation capacity [50].

Due to its morphology, the Jhelum River basin is highly susceptible to soil erosion, in addition to being extremely sensitive to landslides, floods, and earthquakes. Human-caused threats, such as overgrazing, deforestation for agricultural purposes, increasing urbanization, and mining activities, also contribute to the vulnerability of the Jhelum watershed [51]. This study focused on analyzing the Jhelum River basin, which is not only important for the country but is also facing extreme soil erosion due to anthropogenic and climate change activities.

3. Methodology

3.1. RUSLE Equation

Soil erosion is a complex phenomenon that can be studied through various models, including the widely used universal soil loss equation (USLE). The empirical nature of the revised universal soil loss equation (RUSLE) makes it a suitable approach for analyzing soil erosion. Using this approach, cartographic maps can be created to understand soil erosion in specific regions, with several parameters directly or indirectly linked to soil erosion [52,53]. Directly linked parameters such as slope length and steepness, rainfall intensity, and soil erodibility factor impact soil loss, while indirectly linked parameters such as crop management practices and soil conservation practices impact soil loss by influencing the direct parameters [54]. By predicting USLE parameters and creating cartographic maps, the model gains insights into erosion in vulnerable areas [55]. In this study, the RUSLE model was applied using a land cover map created from remotely sensed satellite data, soil types, and agricultural practices. One of the model’s main characteristics is that it can be easily incorporated into GIS for improved analysis, making it a valuable option for this study. The methodology followed during this study is depicted in Figure 2, with the RUSLE equation (Equation (1)) developed by combining five input parameters that are interrelated and subject to variations in both space and time. The input data for each element depends on one another and can change over time and space, leading to the calculation of erosion at a pixel level.

$$\mathrm{A}=\mathrm{R}\times \mathrm{K}\times \mathrm{L}\mathrm{S}\times \mathrm{C}\times \mathrm{P}$$

(1)

where A, average annual soil erosion (t∙ha⁻¹∙y⁻¹); R, rainfall erosivity factor (MJ∙mm∙ha⁻¹∙h⁻¹∙y⁻¹); K, erodibility of soil (t∙ha⁻¹∙h∙MJ⁻¹∙ha⁻¹∙mm⁻¹); the rest of parameters are dimensionless as shown in Figure 2.

All these five contributing factors were incorporated individually in ArcGIS, and then the soil erosion was estimated empirically using the raster map function. The boundaries of the study area were set by using the spatial function of ArcGIS. The entire spatial and remotely sensed data were processed using ArcGIS Pro 2.8.0 software. The RUSLE parameters were calculated using satellite imagery, and digital elevation model data are shown in

Table 1. Sources of remotely sensed data.

Sr. No.	Type of Data	Description	Source
1	DEM	SRTM DEM 1 Arc-Second Global with spatial resolution of 30 m	USGS Earth Resources Observation and Science [56]
2	Soil Data	Digital soil Map of the World by FAO spatial resolution of 5 × 5′	Food and Agriculture Organization of the United Nations [57]
3	Rainfall Data	CRU TS Monthly High-Resolution Gridded Multivariate Climate Dataset Spatial resolution 0.5°	Climatic Research Unit (University of East Anglia) and Met Office [58]
4	Satellite Data	Landsat 8-9 OLI/TIRS Collection 2 with spatial resolution of 30 m	USGS Earth Resources Observation and Science [56]

.

3.2. Rainfall Erosivity Factor-R

R-factor plays a main role in RUSLE. This factor is a major cause of soil erosion and represents the erosivity of rainfall over time. It measures the amount of soil erosion in response to precipitation at a specific place. For soil erosion evaluation, the R-factor is crucial when land use and climate scenarios are changing [59,60,61]. It determines the effect of rainfall and runoff rate [62,63].

In order, to compute the R-factor, the Annual and monthly rainfall data for the Jhelum Basin watershed were used. After the data were imported into a GIS system, the yearly rainfall data from the netCDF raster were converted to point data, and then the Kriging interpolation method was applied to generate the Precipitation map. The monthly R-factor was computed by Equation (2), developed by Smith and Wischmeier [39] and revised by Arnoldus [64].

$$\mathrm{R}={\sum }_{\mathrm{i}=1}^{12}1.735\times 10\times \{1.5\times {\log }_{10}\left\{\frac{{\mathrm{P}}_{\mathrm{i}}^2}{\mathrm{P}}\right\}-0.08188\}$$

(2)

where, R (MJ∙mm∙ha ⁻¹∙h⁻¹∙y ⁻¹) shows the rainfall erosivity, P_i for the rainfall (mm) occurring each month, and P stands for the rainfall (mm) occurring annually. Mean annual rainfall and distribution of rainfall erosivity (R-factor) are shown in Figure 3a,b, respectively. For the study area, the value of R ranges between 544.8 and 766.8, with a minimum value of 267.26 and a maximum of 2043.

3.3. Soil Erodibility Factor-K

The K-factor represents the typical long-term response of soil to rainfall and runoff. This concept can be further supported by the fact that rainfall detaches soil particles, which are then carried away by runoff. The repeated occurrence of this process over time results in soil erosion, which is numerically estimated using the K-factor. This factor specifies the soil loss for a specific soil type with a unit rainfall erosion index. The USLE model employs Wischmeier’s method to calculate the K-factor, which takes into account various soil characteristics such as soil size, level of organic matter, structure, and porosity of particles [27,28]. In order to prepare a soil profile for the study area, the FAO digital soil map for the world was utilized. The soil map for the study area was obtained using the “clip features” function in ArcGIS Pro 2.8.0. The corresponding K-factors were calculated for each soil type using Equations (3)–(7) proposed by Williams [65]:

$$K_{usle}=f_{csand}\times f_{cl-si}\times f_{orgc}\times f_{hisand}$$

(3)

$$f_{csand}=(0.2+0.3\times \mathrm{e}\mathrm{x}\mathrm{p}(-0.256\times m_s(1-\frac{m_{silt}}{100})))$$

(4)

$$f_{cl-si}=(m_{silt}/(m_c+m_{silt}){)}^{0.3}$$

(5)

$$f_{orgc}=(1-\frac{0.25\times orgc}{orgc+\mathrm{e}\mathrm{x}\mathrm{p}(3.72-2.95\times orgc)})$$

(6)

$$f_{hisand}=(1-\frac{0.7\times (1-\frac{m_s}{100})}{(1-\frac{m_s}{100})+\mathrm{e}\mathrm{x}\mathrm{p}(-5.51+22.9\times (1-m_s/100\left)\right)})$$

(7)

where, $K_{usle}$ represents the factor of erodibility, $m_s$ represents the percent sand, $m_{silt}$ represents the percent silt, $m_c$ represents the percent clay, and $f_{orgc}$ reduces K values in soil with high organic carbon content, $orgc$ shows organic carbon content, $f_{csand}$ is a factor that lowers the K indicator in soils with high coarse-sand content and higher for soil with little sand, $f_{cl-si}$ gives low soil erodibility factors for soils with high clay-to-silt ratios, $f_{hisand}$ lowers the K value for soils with extremely high sand content. The corresponding K-factor was computed for each soil type, and the values were entered in the raster table of ArcGIS. By using these values, a K-factor map was generated. The values calculated for each soil type are provided in

Table 2. K-factor values for different soil types of the study area.

Sr. No.	Soil Sample	FAO Soil Class	K-Factor
1	BE	Eutric Cambisols	0.15
2	I	Lithosols	0.14
3	LO	Orthic Luvisols	0.13
4	RC	Calcaric Regosols	0.16
5	GL	Glacier	1
6	WR	Inland water or ocean	0

.

The map in Figure 4a,b illustrates the Jhelum watershed’s soil type and K factor, respectively; typical K values range from 0.131 to 0.156.

3.4. LS-Factor

In USLE, the L and S factors represent the effects of topography on the erosion rate. Higher slope length and slope steepness factors represent more overland flow and more soil erosion [66]. Moreover, slope variations have a much more significant impact on gross soil loss than slope length variations [67]. Topography is an important factor, especially when the ground slope increases beyond the critical angle. In order to obtain the LS factor for the Mangla watershed, the digital elevation model (DEM) was used in the ArcGIS environment. This is because the DEM provides the slope length and slope steepness of the terrain, which are necessary inputs for calculating the LS factor. It is necessary to account for variables such as flow accumulation and slope steepness while computing LS. The slope steepness and flow accumulation parameters were incorporated through the digital elevation model through the ArcGIS Spatial analyzer extension. From this digital elevation model, runoff accumulation and slope were computed. The LS-factor was computed using Equation (8) as suggested by Moore and Burch [68,69].

$$\mathrm{L}\mathrm{S}=\frac{{(\mathrm{F}\mathrm{l}\mathrm{o}\mathrm{w}\mathrm{a}\mathrm{c}\mathrm{c}\mathrm{u}\mathrm{m}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\times \mathrm{C}\mathrm{e}\mathrm{l}\mathrm{l}\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e})}^{0.4}}{22.13}\times \frac{{\left(\mathrm{S}\mathrm{i}\mathrm{n}\mathrm{S}\mathrm{l}\mathrm{o}\mathrm{p}\mathrm{e}\right)}^{1.3}}{0.0896}$$

(8)

where, flow accumulation represents the cumulative upslope supporting area for a cell, the LS factor represents the factor of slope length along with slope steepness, Cell Size represents the size of the grid cell, and the slope degree is represented by ‘Sin slope’ whose value is in sin. The slope classification map and value of the LS factor is illustrated in Figure 5a,b, respectively.

3.5. C-Factor (Crop Cover Factor)

The C-factor ranges from 0 to 1 [11], and it characterizes the vegetation cover percentage and crop’s influence on soil erosion. Vegetation coverage is considered a key factor in preventing water erosion. In order to quantify vegetation cover indices, the Normalized Difference Vegetation Index (NDVI) is used [11,49]. NDVI establishes a relation between soil erosion on land with no vegetation and soil erosion under certain vegetation. It also highlights cover type and density’s impacts on preserving soil. A C-factor map was generated by using Equation (9).

$$\mathrm{N}\mathrm{V}\mathrm{D}\mathrm{I}=\frac{\mathrm{N}\mathrm{I}\mathrm{R}-\mathrm{R}\mathrm{e}\mathrm{d}}{\mathrm{N}\mathrm{I}\mathrm{R}+\mathrm{R}\mathrm{e}\mathrm{d}}$$

(9)

NDVI stands for normalized difference vegetation index, NIR represents the near-infrared band, and Red represents the red band of the satellite image. Landsat 8-9 OLI/TIRS images with a spatial resolution of 30 m were utilized to estimate vegetative density. A Normalized Difference Vegetation Index (NDVI) was computed from Landsat images to approximate greenery. C factor was estimated by using Equations (10) and (11) [70,71], shown below.

$$\mathrm{C} \mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}=0.450-0.805\times \mathrm{N}\mathrm{D}\mathrm{V}\mathrm{I}$$

(10)

$$\mathrm{C} \mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}=\frac{(-\mathrm{N}\mathrm{D}\mathrm{V}\mathrm{I}+1)}{2}$$

(11)

More than 60% of the watershed is covered by vegetation and trees, resulting in a lower C-factor. Conversely, the C-factor is higher in the northeastern parts of the area, which consists of clean land and glaciers. The land use and corresponding C-factor values, as shown in Figure 6a,b, illustrate the variation in vegetation cover and its influence on erosion across the study area.

3.6. P-Factor (Conservation Practice Factor)

The P factor governs the soil loss ratio caused by uphill and downhill tillage, and it is used to adjust the USLE estimate for management and tillage operations to prevent soil erosion [40]. The map of this factor was obtained from the percent slope and land use, and land classification was performed using supervised classification in the ArcGIS environment. Once the land was classified, the slope classes were combined, and the P-conservation factor was assigned to the defined slope classes [39,72].

The typical values of the P factor range between 0.1 and 1, as shown in Figure 7. Non-conservation practices result in higher values, while areas with plantations result in lower values. Based on the results, the core zone of the study area has lower p values, while larger values were found in the top part of the catchment. A conservation value of 0.1 characterizes durable safety capability against erosion resulting from human intervention, and 1 represents no capability for protection [73].

4. Results and Discussion

4.1. Annual Average Soil Loss

This study finds the annual soil erosion for the Jhelum River catchment based on rainfall data for the year 2021 at 1 km² spatial resolution. We used all the geospatial data available for our study area and selected the RUSLE model as a cost-effective and easily applicable method for soil erosion estimation. Soil erosion risk maps were produced by consolidating all five parameters (LS, K, P, R, C) using Equation (1) in the spatial analyst tools, i.e., raster calculator ArcGIS. The map was calibrated and analyzed for the severity of soil risk in the watershed. A higher A value indicates a higher erosion rate, while a lower value indicates a softer sediment yield.

Figure 8 demonstrates all six categories created using the soil erosion and severity map based on previous studies in the same area [74]. 55% of the watershed has very low soil erosion, followed by 19% for low, 14% for moderate, 8% for moderate-high, and 4% for high erosion risk, according to the study’s categories for soil erosion risk, as shown in

Table 3. Classification of soil erosion of Jhelum watershed concerning the area.

Range	Cell Count	Area km2	Area %	Classes
0–10	1,852,243	18,522.43	55%	0–10 tons per hectare
10–25	625,111	6251.11	19%	10–25 tons per hectare
25–50	458,754	4587.55	14%	25–50 tons per hectare
50–100	281,068	2810.68	8%	50–100 tons per hectare
100–250	111,059	1110.59	3%	100–250 tons per hectare
250.00	28,149	281.49	1%	above 250 tons per hectare

.

4.2. Comparison with Global and Local Studies

The present study’s findings were compared to those of similar studies conducted in nearby areas. Specifically, the first point of comparison was Borrelli et al.’s global soil ero-sion map, which utilized RUSLE 2017. This map was resampled to a scale of 25 km² by Gilani et al. to create soil erosion maps for various regions of Pakistan [75,76]. The re-ported soil erosion rate for AJ&K was taken into consideration and compared to the results of this study. To facilitate this comparison, the soil erosion map generated in the present study was resampled to a scale of 25 km² and then compared to past studies. Global soil erosion map showed significantly higher values than those estimated in this study (23.47 t ha⁻¹ yr⁻¹). This fact stands true, especially with lower erosion values where our study re-ports 0–1 t ha⁻¹ yr⁻¹ erosion rates, the global map reports 4–5 t ha⁻¹yr⁻¹.

Due to similar topography and rainfall patterns, the soil erosion values reported in this study are within the range of soil erosion rates reported by other researchers in the adjoining areas and other countries.

Table 4. Comparing soil erosion estimates in neighboring areas with similar topography using both global and local studies.

Reference	Study Area	Soil Erosion Rate (t ha−1 yr−1)
		Calculated by the Respective Local Study
Ashraf et al. [78]	Rawal watershed	10.3
	Ghabbir watershed	22
Aslam et al. [77]	Jhelum watershed	14
SRC’s. [77]	Jhelum watershed	17.51
Borrelli et al. [75,76]	Azad Jammu and Kashmir	40.88
Gilani et al. [76]	Azad Jammu and Kashmir	22.25
Current Study	Jhelum watershed	23.47

highlights a comparison of the average soil erosion rates across various studies conducted at the watershed level. The analysis reveals that the soil erosion rates reported in Sediments rating curve SRC by Surface Water Hy-drology Department (2005) [77], and Rawal watershed were lower than those observed in the present study [78]. In contrast, the Ghabbir watershed and national erosion map for Azad Jammu and Kashmir reported almost equal soil erosion in this watershed.

The drainage area of the Rawal watershed is 268.69 km², while that of the Mangla watershed is 33,563 km². In the Rawal watershed, only 23% [NO_PRINTED_FORM] of the area has a slope of more than 30%, while in the Mangla watershed, the slope of most of the area ranges from 17% to 57% and reaches even 510% in some areas [79]. Because of these facts, the Rawal watershed has an annual average soil loss of 10.3 t ha⁻¹ yr⁻¹, while the Mangla watershed has 23.47 t ha⁻¹ yr⁻¹ of annual average soil loss [78]. The Ghabir watershed drainage area is 417 km², with 85% of the area having gills and nullah [78]. Steep slopes, excessive hilly areas, and less vegetation are the causes of soil loss which is comparable with that of the Mangla watershed.

Furthermore, in previous studies, the soil loss for the Mangla watershed, as shown in Table 4, was calculated using USLE [26,77], whereas we used the RUSLE model. The discrepancy in soil loss values may be due to the fact that USLE does not consider land use and management practices, while RUSLE does. Additionally, differences in data accuracy and completeness and the time elapsed between both studies could also be driving factors for this difference.

Moreover, the soil loss in Azad Jammu and Kashmir is higher than in our study area due to the more rugged mountains and steep slopes, as well as the wetter climate with frequent rainfalls.

4.3. Primary Causes for Soil Loss, Land Use, and Land Cover Changes

From the research conclusions, the estimated soil erosion reported in earlier studies based on the data from 2005 and 2017 was 15 t ha⁻¹ yr⁻¹ which has increased to 23.47 t ha⁻¹ yr⁻¹ in 2021 [80].

The earthquake that occurred in the northern regions of Pakistan on 8 October 2005 resulted in a manifold increase in soil erosion in the area. This earthquake of magnitude 7.6 Mw (Moment Magnitude Scale) and its aftershocks were the sources of many landslides and soil erosion in the area [81]. Analysis shows that soil erosion has increased significantly near the epicenter of the earthquake. Moreover, soil erosion has also escalated due to the high average annual population growth rate of AJ&K (1.61%) [82] and the poor development of infrastructure [83]. The instability in the mountainous regions is further compounded by the fact that Pakistan has a forest cover of only 2.5%, and the annual deforestation rate is 2.1% [84]. Soil conservation methods face various challenges, including socioeconomic differences, violations of the law and court rulings, as well as a lack of awareness and education.

Due to poor rock weathering, increased surface runoff, and natural disasters, soil erosion increased manifold. Additionally, landslides and forest degradation significantly contributed to soil erosion at higher altitudes throughout the sixteen-year study period (2005–2015) [76].

The results indicate that soil erosion values are higher in areas with steep slopes near drainage lines such as streams and rivers. This could be due to runoff carrying large amounts of damaged materials. The steeper sections and higher slopes are also impacted by rills, gullies, and significant amounts of silt, which were easily observable during field surveys. These features are common in the research area due to their delicate geomorphology.

Moreover, the study found that areas with farming and dense mixed forests had the lowest erosion rates, whereas areas covered in snow and glacier patches had the highest rates of erosion [79].

4.4. Strengths and Limitations

The most applicable soil loss estimation model is RUSLE. Because it can predict soil loss using little information, this model is ideal for developing countries with scarce data. The results of other models are elaborate but difficult to interpret and validate due to their complex nature. Most other models are based on empirical rules, leaving much room for error [85]. This model identifies high-risk areas, and management actions are required immediately. It provides an easy-to-use, adaptable, and physical foundation for determining the relative soil loss pattern.

Despite its numerous advantages, the RUSLE model has certain limitations. This equation was established via regression analysis with little field data of similar nature regarding rain patterns, soil types, land cover, and topography [86]. In addition, since it was developed on datasets of the US, its application on a catchment in Pakistan can be questioned. Additionally, to use it for location outside the datasets on which the model was calibrated requires long-term data to calibrate and calculate its coefficients. Moreover, it is also known that gully or stream-channel erosion is not considered while calculating soil loss which is the governing factor in soil erosion. Finally, calculating soil erosion by merely multiplying various factors cannot achieve full accuracy because factors such as sediment deposition, routing, and sediment yield at the catchment outlet are not considered, resulting in inaccurate estimations [87].

5. Conclusions

This study attempted to produce the yearly soil erosion map for the Jhelum watershed using the RUSLE model in a GIS-embedded environment. Various data sources in ArcGIS were used to create RUSLE input parameters. By merging the model RUSLE with GIS techniques, the results of each factor contributing to the erosion process are displayed. Regarding the many elements contributing to land degradation, GIS allows the management of huge amounts of data relatively simply. It’s obvious from this study that the GIS approach is straightforward, has a low cost for modeling and mapping soil erosion, and is very valuable for assessing soil erosion.

The combination of heavy rainfall and steep slopes in the study area results in significant erosion runoff. The primary cause of the huge soil erosion in our study area is the outcome of high runoff and soil particle dislodging. This study shows that 55.18% of this watershed has a low erosion risk, 18.62% of the area has a moderate erosion risk, 13.66% has a high risk, and 16% has a very high erosion risk. Moreover, this study finds that the annual average erosion for the Mangla watershed is 23.47 t ha⁻¹ yr⁻¹.

Although soil loss in this area is comparatively low, effective management policies and strategies must be adopted to further reduce it, with a special focus on C and P. This is because the LS factor is influenced by topography, which cannot be easily changed on a larger scale, while the other two factors, R and K, cannot be changed [88]. This shall increase the life of the Mangla DAM to a greater extent, positively impacting Pakistan’s agricultural and power sectors.

• Proper supervision of terraces will condense the length of the slope and, subsequently, the soil loss [89].
• By adopting innovative and sound cultivation methods, the C factor can be reduced, and erosion can be reduced.
• Conservation practices (P), such as curvature against the slope and trees on the field’s boundary, will ensure that sediments do not leave their source [90].