Modeling Earth Systems and Environment (2019) 5:725–743

https://doi.org/10.1007/s40808-018-0562-6

ORIGINAL ARTICLE

Comparison of the erosion prediction models from USLE, MUSLE

and RUSLE in a Mediterranean watershed, case of Wadi Gazouana
(N-W of Algeria)
Omar Djoukbala1   · Mahmoud Hasbaia2 · Oussama Benselama1 · Mohamed Mazour1

Received: 19 October 2018 / Accepted: 1 December 2018 / Published online: 10 December 2018
© Springer Nature Switzerland AG 2018

Abstract
Water erosion is one of the most serious problems of soil degradation in the world, the north of Africa region is particularly
exposed to this phenomenon. In fact, the phenomenon gets worse with the climate changes and the adverse anthropogenic
environmental interventions. In recent decades, the estimation of soil erosion using empirical models has been a promising
research topic. Nevertheless, their application over a large and ungauged areas remains a real challenge due to the availabil-
ity and quality of the required data. Using the GIS environment, this study aims to estimate and compare the water erosion
rates by the three models of Universal Soil Loss Equation (USLE), Modified Universal Soil Loss Equation (MUSLE) and
Revised Universal Soil Loss Equation (RUSLE) in Wadi Gazouana North-West of Algeria. The estimated specific erosion
in the entire wadi Ghazouana watershed surface is 9.65, (t/ha/year), 9.90 (t/ha/year) and 11.33 (t/ha/year) by USLE, RUSLE
and MUSLE models, respectively. We can also conclude that USLE, RUSLE and MUSLE soil erosion models produced
relatively similar results, however, the MUSLE model showed a higher spatial dispersion of the erosion risk compared to
the others. The rain factor in this model was more effective; which explain its higher erosion rate.

Keywords  Remote sensing · Soil erosion · GIS · USLE/RUSLE/MUSLE · Algeria

List of symbols Introduction

A The computed average soil loss (t/ha/year)
Q Volume of runoff in ­(m3) Soil erosion is a serious threat to the environment in differ-
qp Peak flow rate in ­(m3/s) ent region of the world, it causes onsite problems such as
R Rainfall erosivity in (MJ/ha mm/h) the deterioration of soil physical, chemical and biological
Pi The monthly precipitation (mm) properties (Lal et al. 2000). Soil erosion is a diffuse process
P The annual precipitation (mm) that varies spatially and temporally over a landscape. This
S Surface ­(km2) phenomenon is caused by detachment and entrainment of
soil particles by runoff from land surface.
The influences of climate change on soil erosion rates
have been noticed since the 1940s (Bryan and Albrit-
* Omar Djoukbala ton 1943; Leopold 1951; Ruhe and Scholtes 1956). They
omar.djoukbala@cuniv‑aintemouchent.dz
include direct impacts which are mainly caused by changes
Mahmoud Hasbaia in quantity of precipitation and runoff (Chiew et al. 1995;
hasbaia_moud@yahoo.fr
Bangash et al. 2013), rainfall intensity (Zhang 2012; Tang
Oussama Benselama et al. 2015a) and rainfall spatio-temporal distributional pat-
benselama.oussama@gmail.com
terns (Maeda et al. 2010).
Mohamed Mazour Algeria, following the like of other Mediterranean coun-
mohamed_mazour@yahoo.fr
tries, is classified among the countries most subject to the
1
LHYDENV Laboratory, Belhadj Bouchaib University scourge of land degradation. In effect, several physical fac-
Center of Ain Temouchent, Ain Temouchent, Algeria tors, climatic, environmental and anthropogenic are respon-
2
VESDD Laboratory, University of M’sila, 166 Ichebilia, sible to considerably reduce the area of agricultural land and
28000 M’sila, Algeria therefore their productivity. In view of the food challenges

13
Vol.:(0123456789)

726 Modeling Earth Systems and Environment (2019) 5:725–743

facing the country, the management of the resources for Management Systems (GLEAMS) model of (Davis et al.
sustainable agricultural development is primordial and 1990), the Water Erosion Prediction Project (WEPP)
unavoidable. model of (Nearing et al. 1989), and the Soil Erosion Model
Soil erosion by water is a major soil degradation problem for Mediterranean Areas (SEMMED) model of (Davis
in semi-arid Mediterranean landscapes. Its negative impacts et al. 1990).
are tremendous, including reduction of soil productivity; The empirical models are hydrological models funded
about six million hectares are experiencing a strong to very from applied mathematical laws and validated with labo-
strong degradation (Demmak 1982). The water erosion pre- ratory or field experiments. The simplest and wide used
sents also a real threat to all dams in Algeria, the annual loss model, which links soil loss to either precipitation or runoff
of the Algerian dams capacity is estimated at about 20 mil- is the Universal Soil Loss Equation (USLE) conducted by
lion m3 (Remini 2000). Moreover, we can cite other impacts (Wischmeier and Smith 1965, 1978). The modified version
like: pollution of watercourses, deficits in water availability, of USLE, proposed by (Williams 1975), estimates the sedi-
serious damages to properties by soil-laden runoff which ment transport of each storm taking into account the runoff
increasing the refund demands and desertification of natural volume instead of the erosivity of the rain.
environments. Water erosion is a natural phenomenon con- With further research, experiments, data and available
trolled by climatic characteristics, topography, soil proper- resources, researchers continue to improve the USLE and to
ties, vegetation, and land management. developed new soil loss equations, which led to the devel-
In Algeria, several natural and anthropogenic factors opment of the new universal soil loss equation (RUSLE:
favoring the onset and the development of the processes of Revised Universal Soil Loss Equation), which follows the
erosion: a fragile ecosystem following the aggressiveness of terms of the USLE by correcting some inaccuracies (Renard
climate and the irregularity of precipitation, a hilly topog- et al. 1997) and by providing some improvements in the
raphy and fragile mountainous and of geological substrates. determination of factors.
The human impact manifest in the destruction of a natural It remains the mathematical models, most commonly
plant cover, already low, by overgrazing. used to predict losses due to surface erosion. It predicts the
This degradation of land is a scourge which is wide- annual average rate of long-term erosion based on the factors
spread. At the national level, an average annual specific responsible for the phenomenon: rainfall, soil type, topogra-
erosion ranging between 2000 and 4000 (t/km/year) (Dem- phy, the system of culture and conservation tillage.
mak 1982). The specific erosion changes from one area to The USLE equation was initially proposed for selected
another, the western region is the most affected, i.e., 47% of cropping systems, but it is also applicable to non-agricul-
the whole area, similarly, (26%) and (27%) for the eastern tural conditions. Recently, it has been used successfully at
and the central regions, respectively (Planning Ministry of regional, national and watershed level (Bera 2017; Elaloui
the Environment and Spatial 2000). et al. 2017).
Evaluating the factors controlling erosion and their char- However, the extension of the use of USLE/RUSLE to
acteristics, as well as the detection of eroded areas, are com- study erosion on larger scales than that of the parcel necessi-
plex tasks that can be solved with the integration of several tated the use of geographic information systems and remote
data sources (spatial data, measurements and field surveys, sensing. They have led to great progress in the search for soil
and satellite images) in geospatial processing systems such erosion and soil and water conservation, since the late 1980s.
as geographic information systems. Thus, remote sensing and GIS have become of enormous use
In the scientific literature, different approaches are pro- in assembling, process, analyze and overlay spatial infor-
posed to estimate, quantify and predict soil erosion and sedi- mation that describes the watershed environment, since all
ment transport. These approaches are based on field observa- factors can be mapped. This, made it possible to determine
tions, many modeling concepts, and sometimes both. the values of each factor of erosion per determined spatial
Recently, many models are proposed and developed unit, which is the pixel (Kinnell 2001).
based on the physical aspects of the erosion process, we These techniques also allowed the evaluation of the soil
quote : the model Areal Non-point Source Watershed erosion and its spatial distribution at a reasonable cost,
Environment (ANSWERS) of (Beasley et al. 1980), Annu- reduced time and better accuracy over large areas.
alized Agricultural Non-Point Source Pollutant Loading The main purpose of this study aims at comparing the
(AnnANPSPL), the Erosion-Productivity Impact Calcula- predictions of soil erosion rates between of the universal
tor (EPIC) model of Williams et al. (1985), the Chemicals soil loss equation (USLE) as well as its modified version
Runoff and Erosion from Agricultural (CREAMS) model (MUSLE) and its revised (RUSLE) models integrated under
of (Knisel 1980), the Simulator for Water Resources a Geographic Information System (GIS), in the Mediter-
in Rural Basins (SWRRB) model of (Williams et  al. ranean watershed Wadi Ghazouana (north-west, Algeria).
1985), the Groundwater Loading Effects of Agricultural Wadi Ghazouana presents a typical case of the watersheds

13
Modeling Earth Systems and Environment (2019) 5:725–743 727

of this region with the advantage of the data availability summit less than 1.5%. But the flanks are steep slopes,
required by predictions models. varying between 10 and 12% (Table 1).

Materials and methods Methodology and used data

The study area Details of various parameters used in the soil erosion mod-
eling are given in the Table 2.
The Wadi Ghazouana watershed drains an area of
284.68 km2 with a perimeter of 103.51 km, it is limited
to the North by the Mediterranean Sea; at the South and
South East by the Tafna basin. The watershed is geographi-
cally located between 34°18′ and 34°29′ north latitude and Table 1  Morphometric characteristics of the Wadi El-Ham watershed
between 2°50′ and 3°8′ east longitude (Fig. 1). Parameters Unit Value
The Wadi Ghazouana watershed is located on the eastern
Aria A ha 28468.62
fringe of the chain mountainous Traras and opens on the
Perimeter P km 103.51
Mediterranean Sea, it is characterized by a rugged terrain
Index of compactness IC – 1.72
with steep slopes and the altitudes culminate in the south
Maximum altitude H ­ max m 1120
more than 1100 m at Djebel Fllaoucene.
Minimum altitude H ­ min m 0
The slopes in the Wadi Ghazouana watershed are rela-
Mean basin elevation ­Hmean m 437.30
tively strong, they reach 10–12%. The zone of slight slopes
Altitude to 95% m 210
(I < 2%) is represented by the lower zone of the watershed
Altitude to 50% m 430
at the mouth of the Oued.
Altitude to 5% m 870
At the level of the North-East, the plateau of Sidi Amar
Length of the rectangle L km 45.50
overhangs the city. It is in form triangular whose base
Width of the rectangle l km 6.26
reaches 1500  m. The plateau is flat with slopes at the

Fig. 1  Location of the study area

13

728 Modeling Earth Systems and Environment (2019) 5:725–743

Table 2  Description of the data used

Models Parameter Materials Resolution Year Source

USLE Satellite image Landsat 8 30 m 2018 United States Geological Survey (http://earth​explo​rer.usgs.gov/)
MUSLE Rainfall data Monthly/ 30 m 1985–2015 Agence Nationale des Ressources Hydrauliques (ANRH)
RUSLE annual rain-
fall data
Soil properties H.W.S.D – – Harmonized world soil database (HWSD) version 1.2 (http://webar​
chive​.iiasa​.ac.at/Resea​rch/LUC/Exter​nal-World​-soil-datab​ase/
HTML/)
Topographic data A.S.F 12.5 m 2011 Alaska Satellite Facility: https​://verte​x.daac.asf.alask​a.edu

The USLE model of rainfall-runoff erosivity (R) multiplied by the resistance

of the environment, which includes the soil erodibility factor
The empirical formula of Wischmeier and Smith (1965, (K), the topography factor (LS), the anti-erosion practices
1978) is proposed to estimate the rate of soil loss in many (P) and the vegetation cover factor (C). This model is written
area scales. It is often combined with GIS techniques, similarly to the USLE as:
and widely used around the world to quantify soil losses.
Depending on rainfall patterns, land use, topography, soil
ARUSLE = R × K × LS × C × P. (3)
erodibility and anti-erosion practices, this model is written: The Universal Soil Loss Equation is an empirical basic
equation. Revised Universal Soil Loss Equation (RUSLE)
AUSLE = R × K × LS × C × P , (1) model is only ones of many modification of USLE, espe-
where A is the computed average soil loss (t/ha/year); R is cially for more complex situations of rill and interrill erosion
rainfall-runoff erosivity factor (MJ mm/ha/h/y); K is the soil in conservation planning and land uses. Both erosion-prone
erodability factor (t ha h/ha/MJ/mm); LS is the slope length models calculate detachment capacity and soil loss.
(L) and slope gradient (S) factor (dimensionless); C is the The RUSLE has been applied to many different water-
cropping management factor and P is the supporting con- sheds about the world (Jiang et al. 2015; Tang et al. 2015b;
servation practice factor (dimensionless, ranging between Fernández and Vega 2016; Markose and Jayappa 2016; Tet-
0 and 1). ford et al. 2017; da Cunha et al. 2017; Toubal et al. 2018;
Wijesundara et al. 2018; Djoukbala et al. 2018).

The MUSLE model

Factors of the models
The (MUSLE) model (Williams 1975), is a modified ver-
sion of the USLE model (Wischmeier and Smith 1965). This R‑factor
model replaced the rainfall factor (R) with instantaneous
peak flows and total runoff factor to predict soil erosion. This factor represents the energy with which rain droplets
The average soil loss for a flood is a multiplicative function impact the soil at certain intensity by breaking up surface
of the volume of the flood (V in m ­ 3), the peak discharge of aggregates into particles of transportable size. Many indi-
3
the flood (Qp in ­m /s), the erodibility of the soil (K), the ces have been designed in erosion risk prediction models to
index slope (S), the slope length (L), the vegetation cover estimate the rainfall-runoff effect, the best known of which
(C) and the cultural practices (P). This formula is written in is the R factor.
the following form: The R factor is the erosion index of rainfall and includes
the influence of the kinetic energy of rain on erosion, the
AMUSLE = 11.8(Q × qp )0.56 × K × LS × C × P. (2) disaggregation of soil particles and the compaction of their
surface, as well as their maximum intensity, determining the
appearance of surface runoff infiltration capacity.
The RUSLE model
Rainfall erosivity factor (R) according to MUSLE
The RUSLE model estimates the average annual loss of soil.
It is a revised version of the universal soil loss equation The rainfall erosivity factor was developed by Williams
(USLE) (Wischmeier and Smith 1965, 1978). According to and Berndt (1977). It’s estimated from t the runoff volume
Roose and Noni (2004), erosion is a multiplicative function and the peak flow. Peak flow is calculated using a rational

13
Modeling Earth Systems and Environment (2019) 5:725–743 729

equation, which assumes that precipitation, has a uniform Database Version 1.2 to determine the required soil
intensity over the entire watershed area. parameters over the entire Ghazouana watershed. This
database is the result of a collaboration between FAO and
R = 11.8 × (Q × qp )0.56 . (4) IIASA, ISRIC-World Soil Information, the Institute of
Soil Sciences, Chinese Academy of Sciences (ISSCAS),
With Q is volume of runoff in ­(m3); ­qp is peak flow rate and the Joint Research Center (JRC).
in ­(m3 s− 1). The harmonized database of soil data of the world is a
In our case study, the measured runoff volume and the database with a frame of 30-s arc and more than 15,000
peak flow are recorded at the BESBES gauging station situ- characteristic soil-mapping units. The resulting database
ated at the outlet of Wadi Ghazouana watershed. consists of 21,600 rows and 43,200 columns, which are
linked to harmonized soil property data. The use of a
Rainfall erosivity factor (R) according to RUSLE and USLE standardized structure allows data linkage with the card
frame to display or query the structure in terms of soil
In the original form of USLE, according to the formula of units and description of selected soil parameters (salinity,
Wischmeier and Smith (1965), the estimation of the factor R organic carbon, storage capacity of water, soil profundity,
requires the knowledge of the kinetic energies (Ec) and the pH, cation exchange capacity of the soil, clay fraction,
average intensity over 30 min (I30) of the raindrops of each sum of exchangeable nutrients, lime and gypsum content,
precipitation episode. In our study area, the only available percentage of sodium exchange, textural class and granu-
precipitation data are the monthly and the annual averages lometry) (FAO and ISRIC 2012).
values. In this study, the value of the Soil erodibility was cal-
In the Algerian context, the only available data of precipi- culated using the following formulas proposed by (Neitsch
tation are usually measured at the monthly and the annually et al. 2011).
scale. An alternative formulas, which only involve monthly
and annual precipitation to determine the R factor, are pro- KUSEL = Kw = fcsand .fcl−si .forgc .fhisand , (6)
posed by Arnoldus (Cormary and Masson 1964). Which is where ­(fcsand) is a factor, that brings down the K display in
presented in the form: soils with high coarse-sand content and higher for soils with
little sand; (­ fcl–si) gives low soil erodibility factors for soils
∑
12
P2i with high clay-to-silt ratios; (­ forgc) diminishes K values in
log R = 1.74 log + 1.29, (5) soils with higher organic matter content; (­ fhisand) lowers K
i=1
P
values for soils with extremely high sand content.
where R: rainfall erosivity in (MJ/ha mm/h); Pi is the ( [ ( m )])
monthly precipitation (mm) and P is the annual precipita- fcsand = 0.2 + 0.3 ⋅ exp −0.256 ⋅ ms ⋅ 1 − silt , (7)
100
tion (mm).
In our study watershed, the used data of rainfall are ( )
msilt
recorded from 10 rainfall stations situated inside and around fcl−si = , (8)
mc + msilt
the watershed of Wadi Ghazouana, for a measurement period
variable from 25 to 32 years. The R-values were inserted ( )
over the entire watershed territory utilizing the geostatistical 0.25 ⋅ orgC
forgc = 1− [ ] , (9)
model of Kriging. orgC + exp 3.72 − 2.95 ⋅ orgC

K‑factor � �
⎛ m ⎞
⎜ 0.7 ⋅ 1 − 100s ⎟
The K factor refers to the erodibility of the soil, that is to say fhisand= ⎜1 − � � � � �� ⎟,
the resistance of the soil against the aggressiveness of rain- ⎜ 1−
ms ms
+ exp −5.51 + 22.9 ⋅ 1 − ⎟
⎝ 100 100 ⎠
drops, runoff or both. This factor is related to the integrated
effect of precipitation, runoff and seepage and relies on the (10)
influence of soil properties on soil loss. where ms is the percent (%) sand content (0.05–2.00 mm
This factor is determined according to four necessary diameter particles); msilt is the percent (%) of silt content
parameters: the texture parameter (% silt, sand, clay) and (0.002–0.05 mm diameter particles); mc is the percent (%)
the percentage of organic matter. of clay content (< 0.002 mm diameter particles); orgC is the
Due to the lack of accurate soil map of our study percent organic of carbon content of the layer (%) (Table 3).
region, we have used the Global Harmonized Soil

13

730 Modeling Earth Systems and Environment (2019) 5:725–743

Table 3  Estimation of K factor in Wadi Ghazouana watershed

Soil sample ms (Sand) msilt (Silt) mc (Clay) orgC Oraganic Fcsand Fcl–si Forgc Fhisand KUSLE K
Top soil % Top soil % Topsoil % carbon %

XK 48.7 29.9 21.6 0.64 0.200 0.849 0.977 0.999 0.1658 0.0218
X 72.8 10.5 16.8 0.36 0.200 0.751 0.994 0.918 0.1370 0.0180
LC 64.3 12.2 23.5 0.63 0.200 0.725 0.978 0.983 0.1392 0.0183
ZG 47.8 8.5 43.8 0.38 0.200 0.580 0.993 0.999 0.1151 0.0151
BK 81.6 6.8 11.7 0.44 0.200 0.741 0.991 0.718 0.1054 0.0138

Xk calcic xerosols, X XEROSOLS, Lc chromic luvisols, Zg gleyic solonchaks, Bk calcic cambisols

LS‑factor As = flow accumulation × resolution.

The topographical factor is one of the main factors of water The surface card (­ As) is determined by multiplying the
erosion. The degree of slope Play an important role in the pro- map of the runoff accumulation (Flow accumulation) by the
cess of detachment, transport and deposition of soil particles. pixel size (resolution).
The LS factor refers to the topography of the region,
which manifests itself in two sub-factors: The length of the Factor (LS) according to RUSLE
slope (L) and the degree of slope (S), however, it is advis-
able to use both variables as an only topographic value (LS ( )
resolution m
LS = flow accumulation ×
to evaluate overall the influence of the slope on the speed 22.1 (15)
( )
of erosion). × 0.065 + 0.045 × S + 0.0065 × S2 .
The sub-factors were determined in a particular way using
the digital terrain model DTM obtained from the DEM With ‘S’ the angle of the slope in (%) and ‘m’ is a param-
ALOS PALSAR (Alaska Satellite Facility) under GIS tool eter related to the slope classes (Wischmeier and Smith
in the WGS_1984 coordinate system. 1978) (Table 4).
In our study, we used three calculation formulas accord-
ing to the models: C‑factor
Factor (LS) according to USLE The vegetation plays an important slowdown role of the
water erosion process, indeed the vegetation cover factor is
( ) the second most important factor controlling the risk of soil
resolution m
LS = flow accumulation × erosion (Kalman 1967).
( 22.1 ) (11)
2
× 65.41sin 𝜃 + 4.56 sin 𝜃 + 0.065 , Many studies have been proposed to estimate the Cover
Management factor (C-factor) using the Normalized Dif-
[ ] ference Vegetation Index (NDVI) for assessing soil loss
m = 0.6 × 1 − e(−35.835×S) , (12) (Lin et al. 2002; Wang et al. 2002). These approaches use a
( ) regression model to perform a correlation analysis between
𝜃 = tan−1
S
, (13) field-measured C-factor values or obtained from guide
100 boards and NDVI values derived from remote images. The
where S is the field slope (%) and θ is the field slope steep- unknown values of the C-factors of the land cover classes
ness in degrees. can be estimated using the equation obtained from a linear
regression analysis.
Factor (LS) according to MUSLE
( )0,4
As sin(𝜃 × 0.01745)1,4 Table 4  Value of ‘m’ relative to
LS = 1.4 × × . (14) Slope (%) M
22.1 0.09 each class of slope (Wischmeier
and Smith 1978) > 5 0.5
With 𝜃 : the angle of the slope in degrees, ­As: The particu- 3–5 0.4
lar surface composition under a GIS in the following method 1–3 0.3
(Moore and Burch 1986): < 1 0.2

13
Modeling Earth Systems and Environment (2019) 5:725–743 731

The Normalized Difference Vegetation Index (NDVI) val- In order to estimate the values of the C-factor, some
ues range from − 1 to + 1, the negative values are attributed authors (Toumi et al. 2013) have used the regression between
for surfaces without plant cover, such as snow, water, or two extreme values of (NDVI), the obtained regression line
clouds, for which red reflectance is greater than near-infrared is written as
(Souidi et al. 2014). For bare soils, the reflectance being of
about the same order of size in the red and the near infrared,
c = 0.9167 − NDVI × 1.1667. (16)
the (NDVI) has value close to zero. The vegetal formations P factor
as for them have a positive values of (NDVI), generally
between 0.1 and 0.7. The highest values correspond to the The P factor represents the effect of the conservation prac-
densest cutlery. tices on the water erosion processes. It varies according to
In this study, the (NDVI) data (period 2018) are generated the conservation technics practiced in the watershed from
from a Satellite Landsat 8 photo with a spatial resolution of 0 in the zones well protected to 1 without any conservation
30 m (Fig. 2), was used to estimate the C-factor and explain practices.
the effect of different vegetation cover on the Loss of soil, In our study area, no significant anti-erosif technic is
the (NDVI) was calculated from a combination of red and practiced; the value of 1 was assigned to the P factor in the
infrared bands. entire watershed area.

Fig. 2  Spatial distribution of NDVI index values in the Wadi Ghazouana

13

732 Modeling Earth Systems and Environment (2019) 5:725–743

Figure 3 gives a general idea of the functioning of the to 95.92 (MJ mm/ha/h/year) affects 33% of the surface. The
model that presents a summary of the used methodology of rest of the watershed area, about 44% has a moderate ero-
a soil erosion map are illustrated. sivity with R-factor ranging from 95.92 to 109.38 (MJ mm/
ha/h/year) (Table 5).

Results and discussion
K‑factor
Erosivity factor (R)
The soil erodibility factor K in the Wadi Ghazaouna water-
The erosivity map generated from the rainfall data of the shed varies from 0.013 to 0.022 (MJ mm/ha/h/an). A com-
12 used stations, shows that the value of the R factor in the posite K factor map was generated to show the spatial distri-
Ghazouana watershed varies from 89 to 130 (MJ mm/ha/h/ bution of erodibility across five soil groups (Fig. 5). About
year). The high values are recorded at the northwest of the 85% of the basin area has an erodibility factor of less than
basin, while the lowest values are recorded in the Southeast. 0.015, the rest of the surface, almost 15% of the, presents
La carte d’érosivité réalisé à partir des données pluvio- a high erodibility factor K, ranging from 0.015 to 0.021
métrique des stations pluviométriques, montre que la valeur (Table 6).
du facteur R dans le bassin versant de Ghazaouet varie de 89 Using the spatial K-factor map, it is possible to manipu-
à 130 (MJ mm/ha/h/year). Les valeurs élevées sont enregis- late the entire watershed surface, pixel by pixel, we can also
trées du Nord-Ouest du bassin, alors que les valeurs les plus generate a resulting maps by applying various operations.
faibles sont enregistrées à Sud-Est (Fig. 4).
The R-factor is distributed over the watershed area in LS‑factor
many classes, 24% of the basin area is caracterized by a high
erosivity with R-values from 109.38 to 129.96 (MJ mm/ha/h/ From the three used models USLE, RUSLE, MUSLE,
year). The class of low erosivity with R-factor from 89.59 LS vary between 0 and 238.7, 0–216.8 and 0–187.52;

Fig. 3  Diagram of the methodology to map the water erosion risk

13
Modeling Earth Systems and Environment (2019) 5:725–743 733

Fig. 4  Map of the rainfall erosivity factor R

Table 5  Distribution of rainfall erosivity classes slopes (Fig. 6a–c). The LS values that considered as low
Classes R factor Area (ha) Area (%)
are observed in the plain, this corresponds to low elevation
areas, lowland areas and stream bed. Thus, the basin is sub-
89.59–95.92 9284.328 32.6 ject to a high risk of erosion from upstream to downstream.
95.92–102.09 5515.239 19.4 These results are consistent with those obtained by many
102.09–109.38 6854.832 24.1 authors (Abdo and Salloum 2017; da Cunha et al. 2017;
109.38–117.61 3577.646 12.6 Imamoglu and Dengiz 2017; Djoukbala et al. 2018).
117.61–129.96 3238.477 11.4 Erosion has been shown to increase exponentially
Total 28462.82 100 depending on the degree of inclination of slope (Pham et al.
2018). Similarly, it has been reported that as the degree
of inclination of slope increases, the kinetic energy of the
respectively. LS values were grouped into six classes rains remains constant while the transport accelerates down-
(Table 7). The length and slope of the slope are decisive in wards due to an increase in the kinetic energy of the runoff
the erosion process. (Thomas et al. 2018).
The highest values of LS are located in the upstream part
of the Wadi Ghazouana basin and on the hills with very high

13

734 Modeling Earth Systems and Environment (2019) 5:725–743

Fig. 5  Spatial distribution of the soil erodibility factor K

Table 6  Distribution of K factor class in the Wadi Ghazouana Water- C‑factor

shed
Classes K factor Area (ha) Area (%) The spatial distribution of vegetation cover management
factor C of Wadi Ghazouana watershed is generated from
0.01388 2862.392 10.05
the NDVI map, under ArcGIS software (Fig.  7). The
0.01516 1219.79 4.2
observed C-factor value over the whole study area is very
0.01804 5509.52 19.3
variable, from 0.26 and to 0.95, 89% of the basin area has a
0.01834 14738.39 51.8
very low vegetation cover and only 11% of the area is well
0.02185 4120.94 14.4
protected with C < 0.7. Low vegetation cover is found in
Total 28462.52 100
open forests, the degraded rangelands and cultivated lands

13
Modeling Earth Systems and Environment (2019) 5:725–743 735

Table 7  Distribution of LS factor class in the Wadi Ghazouana Watershed

USLE RUSLE MUSLE
Classes LS factor Area (ha) Area (%) Classes LS factor Area (ha) Area (%) Classes LS factor Area (ha) Area (%)

0–2.8 17826.23 62.6 0–2.55 17901.16 62.9 0–2.5 16833.43 59.2

2.8–8.4 7804.270 27.4 2.55–8.47 7608.68 26.7 2.5–8.4 8897.443 31.3
8.4–18.66 2283.037 8.0 8.47–17.8 2221.66 7.8 8.4–17.5 1798.350 6.3
18.66–36.38 440.1233 1.5 17.8–33.03 561.65 2.0 17.5–35 604.40 2.1
36.38–70.89 97.51 0.3 33.03–66.07 138.32 0.5 35–65 301.20 1.1
70.89–238.8 16.46 0.1 66.07–216.8 17.92 0.1 65–187 14.57 0.1
Total 28462.63 100 Total 28462.63 100 Total 28462.63 100

Fig. 6  a The LS factor map (MUSLE). b The LS factor map (RUSLE). c The LS factor map (USLE)

are considered highly susceptible to erosion. Values below by low density, sparse forests and light matorrals. Values
0.5 refer to dense forests, dense matorrals and arboriculture. tending towards 1 are related to bare soils and harvested
Values between 0.5 and 0.9 are attributed to areas covered cropland (Table 8).

13

736 Modeling Earth Systems and Environment (2019) 5:725–743

Fig. 6  (continued)

Erosion assessment and quantification 9.65 (t/ha/year). The examination of the spatial distribution
show a large dispersion, over the whole watershed surface,
The average annual soil loss in Wadi Ghazouana watershed the water erosion vary from 0 to 303 (t/ha/year) (Fig. 8). It
was calculated with USLE RUSLE and MUSLE models is also observed that, the areas with high erosive risk are
according the Eqs. 1, 2 and 3; respectively. The final maps located on hills and areas characterized by steep slopes and
are obtained by superimposing the grids of five factors in a favorable substrates; they represent 47.2% of the surface of
single result grid under ArcGIS environment with a uniform Wadi Ghazouana watershed. The rest of the watershed sur-
spatial resolution of 30 m. face, ie 52%, is characterized by a low and medium erosion-
sensitivity (Table 9).
Potential soil loss
RUSLE model
USLE model
The erosion risk map (Fig. 9) obtained by this revised model
The erosion risk map obtained bu the USLE shows that the (RUSLE) indicates a results close to those obtained by the
average total annual soil loss of Wadi Ghazouana water- previous model (USLE), the average total annual soil loss
shed is about 2655.22 (t/year) equivalent to a specific rate of of Wadi Ghazouana watershed is about 2783.52 (t/year)

13
Modeling Earth Systems and Environment (2019) 5:725–743 737

Fig. 6  (continued)

equivalent to a specific rate of 9.90 (t/ha/year). The distribu- RUSLE models. This indicates that erosion is very active
tion of erosion risk by class is summarized in the Table 10. downstream of Wadi Ghazouana watershed.
The results of this study in Wadi Ghazouana watershed,
together with other previous studies in this area, show that
MUSLE model the water erosion rates are much higher than tolerance lev-
els ­[A(USLE)(MUSLE)(RUSLE) > 7 (t/ha/year)]. This explains the
The erosion risk map (Fig. 10) obtained by this MUSEL severity of soil degradation in Wadi Ghazouana watershed
model shows slightly different results compared to USLE due to the unfavorable land use, lithology and the aggres-
and RUSLE maps, the average total annual soil loss of Wadi siveness of climate.
Ghazouana watershed is about 3231.34 (t/year) equivalent to The USLE, RUSLE and MUSLE soil erosion models
a specific rate of 11.33 (t/ha/year). The calculated soil loss produced relatively similar results, however, the MUSLE
by MUSLE model varies from 0 to 395 (t/ha/year), it is also model showed a higher spatial disputation of the erosion
classified from very low to very high (Table 11). risk compared to the others. R factor was more effective in
The high rates of erosion are observed at hills and lands the MUSLE model; which explain the higher erosion rates
characterized by steep slopes of Wadi Ghazouana watershed. obtained by this model. The comparison of the obtained
This finding is consistent with those obtained by USLE and results by the three models show a very close value of

13

738 Modeling Earth Systems and Environment (2019) 5:725–743

Fig. 7  The C-factor map of Wadi Gazouana watershed

Table 8  Distribution of C factor class in the Wadi Ghazouana Water- We can notice that, relatively the RUSLE results are the
shed most compatible to the nature of the data used. However, this
Classes c factor Area (ha) Area (%) may not always be the case, as it depends on the region and
state in which the model was developed. The best response
0.26–0.66 285.53 1.0 of RUSLE model in our region allows us to consider it as the
0.66–0.75 2334.25 8.2 most suitable model for the semi-arid Mediterranean. The
0.75–0.80 6399.03 22.5
observed differences in model results may also be due to the
0.80–0.84 12303.21 43.2
dominant erosion processes at different spatial and temporal
0.84–0.95 7140.81 25.1
scales, models that have been designed for different regions
Total 28462.82 100
such as RUSLE, and that have been designed for agricul-
tural areas in the United States, differ in mixing erosional
processes with dryland ecosystems. The advantages of such
the averages rates. The coefficient of determination ­R 2 model are their simplicity in terms of data requirements and
between the results ranges from 93 to 98%, which con- computation (Harmon and Doe 2001). However, the differ-
firms the strong correlation between the studied models ent modeled processes can not be disaggregated or modi-
(Fig. 11). fied, which is problematic if the model has been designed

13
Modeling Earth Systems and Environment (2019) 5:725–743 739

Fig. 8  Soil erosion map in the Wadi Ghazouana (USLE)

Table 9  Percentage of area under different soil erosion classes (Ghadiri and Rose 1993; Hudson 1993). The ideal approach
(USLE) is to develop new models or to calibrate existing ones to the
Soil loss (t/ha/year) USLE Area (ha) Area (%) regional context, such, (SLEMSA) the Soil Loss Estimation
Model for Southern Africa (Elwell 1978) and (EUROSEM)
0–3 15029.01 52.8 the European Soil Erosion Model (Hudson 1993).
3–6 5153.807 18.1 The current results are also compatible with other works
6–12 4960.176 17.4
on the evaluation of water erosion carried out in other Med-
12–24 2469.052 8.7
iterranean watersheds having climatic and environmental
24–303 850.7745 3.0
characteristics that are similar. such as in Wadi Mina water-
Total 28462.82 100
shed at 11.2 (t/ha/year) (Benchettouh et al. 2017), in Wadi
Boumahdane of 11.18 (t/ha/year) (Bouguerra et al. 2017)
and in Wadi Sahouat basin average soil losses between 12
for a different location or spatial scale. Besides, empirical and 16 (t/ha/year) (Toubal et al. 2018).
relationships are often calibrated for a particular set of data The values calculated by the empirical soil loss mod-
that is valid only for the dataset from which they are derived els are subject to discussion, but the method is among the

13

740 Modeling Earth Systems and Environment (2019) 5:725–743

Fig. 9  Soil erosion map in the Wadi Ghazouana (RUSLE)

Table 10  Percentage of area under different soil erosion classes Conclusion

(RUSLE)
Soil loss (t/ha/year) RUSLE Area (ha) Area (%) The availability and the quality of required data to estimate
the water erosion risk is a real challenge in the large and the
0–3 16775.61 58.9
ungauged basins, empirical models are ideal tools to over-
3–6 4423.125 15.5
come this handicap, especially, with the quick development
6–12 3937.541 13.8
of the GIS tools. This study aimed to use and compare the
12–24 2434.64 8.6
results of the water erosion rates obtained from three models
24–275 891.9087 3.1
USLE, MUSLE and RUSLE, the spatial distribution of the
Total 28462.82 100
phenomenon is also verified in the coastal watershed Wadi
Ghazouana in the Mediterranean region.
The three models were successfully applied in the
decision support tools used by planners, as it simulates soil study area, the obtained results are relatively close. The
protection and conservation scenarios to plan erosion control estimated specific erosion in the entire Wadi Ghazouana
interventions, especially on slopes where water erosion is watershed surface is 9.65, (t/ha/year), 9.90 (t/ha/year) and
predominant. 11.33 (t/ha/year) by USLE, RUSLE and MUSLE models,

13
Modeling Earth Systems and Environment (2019) 5:725–743 741

Fig. 10  Soil erosion map in the Wadi Ghazouana (MUSLE)

Table 11  Percentage of area under different soil erosion classes The MUSLE model results showed that more than
(MUSLE) 30% of watershed surface, present a higher erosion rates,
Soil loss (t/ha/year) MUSLE Area (ha) Area (%) greater than 6 (t/ha/year), observed at hills and lands char-
acterized by steep slopes of the watershed. This explains
0–3 16971.2 59.6 the severity of soil degradation in Wadi Ghazouana water-
3–6 3627.829 12.7 shed due to the unfavorable land use, lithology and the
6–12 3840.49 13.5
aggressiveness of climate.
12–24 2860.651 10.1
The erosion maps, obtained from the three models, can
24–395 1162.653 4.1
be useful to select the suitable zones for soil conservation
Total 28462.82 100
planning and revegetation efforts based on assessed ero-
sion rates or composition factors mapped by each input
factor independent. The development of GIS tools and
respectively. The MUSLE model showed a higher spa- remote sensing technics can significantly improve the
tial distribution of erosion due a better efficiency of rain results of this kind of models.
factor.

13

742 Modeling Earth Systems and Environment (2019) 5:725–743

Fig. 11  Variation of the erosion

calculated by the three models

References Djoukbala O, Mazour M, Hasbaia M, Benselama O (2018) Estimat-

ing of water erosion in semiarid regions using RUSLE equation
under GIS environment. Environ Earth Sci 77:345. https​://doi.
Abdo H, Salloum J (2017) Spatial assessment of soil erosion in Alqer-
org/10.1007/s1266​5-018-7532-1
daha basin (Syria). Model Earth Syst Environ 3:26. https​://doi.
Elaloui A, Marrakchi C, Fekri A et al (2017) USLE-based assess-
org/10.1007/s4080​8-017-0294-z
ment of soil erosion by water in the watershed upstream Tes-
Bangash RF, Passuello A, Sanchez-Canales M et al (2013) Ecosystem
saoute (Central High Atlas, Morocco). Model Earth Syst Environ
services in Mediterranean river basin: climate change impact on
3:873–885. https​://doi.org/10.1007/s4080​8-017-0340-x
water provisioning and erosion control. Sci Total Environ 458–
Elwell HA (1978) (1978) Modelling soil losses in Southern Africa. J
460:246–255. https​://doi.org/10.1016/j.scito​tenv.2013.04.025
Agric Eng Res 23:117–127
Beasley DB, Huggins LF, Monke EJ (1980) ANSWERS: a model
FAO and ISRIC (2012) Harmonized world soil database. FAO, Rome,
for watershed planning. Trans ASAE 23:0938–0944. https​://doi.
Italy and IIASA, Laxenburg, Austria
org/10.13031​/2013.34692​
Fernández C, Vega JA (2016) Evaluation of RUSLE and PESERA
Benchettouh A, Kouri L, Jebari S (2017) Spatial estimation of soil
models for predicting soil erosion losses in the first year after
erosion risk using RUSLE/GIS techniques and practices conserva-
wildfire in NW Spain. Geoderma 273:64–72. https​ : //doi.
tion suggested for reducing soil erosion in Wadi Mina watershed
org/10.1016/j.geode​rma.2016.03.016
(northwest, Algeria). Arab J Geosci. https:​ //doi.org/10.1007/s1251​
Ghadiri H, Rose CW (1993) Water erosion processes and the enrich-
7-017-2875-6
ment of sorbed pesticides. Part 2. Enrichment under rainfall domi-
Bera A (2017) Assessment of soil loss by universal soil loss equation
nated erosion process. J Environ Manag 37:37–50
(USLE) model using GIS techniques: a case study of Gumti River
Harmon RS, Doe WW (2001) Landscape erosion and evolution mod-
Basin, Tripura, India. Model Earth Syst Environ 3:29. https​://doi.
eling. Springer Science & Business Media, New York
org/10.1007/s4080​8-017-0289-9
Hudson N (1993) Field measurement of soil erosion and runoff. Food
Bouguerra H, Bouanani A, Khanchoul K et al (2017) Mapping erosion
& Agriculture Org, Rome
prone areas in the Bouhamdane watershed (Algeria) using the
Imamoglu A, Dengiz O (2017) Determination of soil erosion risk using
Revised Universal Soil Loss Equation through GIS. J Water Land
RUSLE model and soil organic carbon loss in Alaca catchment
Dev 32:13–23. https​://doi.org/10.1515/jwld-2017-0002
(Central Black Sea region, Turkey). Rend Lincei 28:11–23. https​
Bryan K, Albritton CC (1943) Soil phenomena as evidence of climatic
://doi.org/10.1007/s1221​0-016-0556-0
changes. Am J Sci 241:469–490
Jiang L, Yao Z, Liu Z et al (2015) Estimation of soil erosion in some
Chiew FHS, Whetton PH, McMahon TA, Pittock AB (1995) Simula-
sections of Lower Jinsha River based on RUSLE. Nat Hazards
tion of the impacts of climate change on runoff and soil mois-
76:1831–1847. https​://doi.org/10.1007/s1106​9-014-1569-6
ture in Australian catchments. J Hydrol 167:121–147. https​://doi.
Kalman R (1967) Essai d’évaluation pour le pré-Rif du facteur couver-
org/10.1016/0022-1694(94)02649​-V
ture végétale de la formule de Wischmeier de calcul de l’érosion.
Cormary Y, Masson J (1964) Etude de conservation des eaux et du
Rapport Rabat 1–12
sol au Centre de Recherches du Génie Rural de Tunisie: appli-
Kinnell P (2001) Slope length factor for applying the USLE-M to
cation à un projet-type de la formule de perte de sols de Wis-
erosion in grid cells. Soil Tillage Res 58:11–17. https​://doi.
chmeier. Cahiers ORSTOM, série pédologie 2.3:3–26
org/10.1016/S0167​-1987(00)00179​-3
da Cunha ER, Bacani VM, Panachuki E (2017) Modeling soil erosion
Knisel WG (1980) CREAMS: a field scale model for Chemicals, Run-
using RUSLE and GIS in a watershed occupied by rural settlement
off, and Erosion from Agricultural Management Systems (USA).
in the Brazilian Cerrado. Nat Hazards 85:851–868. https​://doi.
United States Dept Agric Conserv Res Rep
org/10.1007/s1106​9-016-2607-3
Lal R, Ahmadi M, Bajracharya RM (2000) Erosional impacts on
Davis FM, Leonard RA, Knisel WG (1990) Groundwater loading
soil properties and and corn yield on alfisols in central Ohio.
effects of agricultural management systems (GLEAMS) user
Land Degrad Dev 11:575–585. https​://doi.org/10.1002/1099-
manual. USDA-ARS Southeast Watershed Res Lab Univ Georg
145X(20001​1/12)11:6%3C575​::AID-LDR41​0%3E3.0.CO;2-N
Tifton, Tifton
Leopold LB (1951) Rainfall frequency: an aspect of climatic variation.
Demmak A (1982) Contribution à l’étude de l’érosion et des transports
Eos Trans Am Geophys Union 32:347–357
solides en Algérie septentrionale [Contribution to the study of
erosion and sediment transport in northern Algeria]. PhD Thesis,
Manuscript, Université de Pierre et Marie Curie, Paris

13
Modeling Earth Systems and Environment (2019) 5:725–743 743

Lin CY, Lin WT, Chou WC (2002) Soil erosion prediction and sedi- the Loess Plateau, China. Environ Earth Sci 73:1715–1724. https​
ment yield estimation: the Taiwan experience. Soil Tillage Res ://doi.org/10.1007/s1266​5-014-3523-z
68:143–152. https​://doi.org/10.1016/S0167​-1987(02)00114​-9 Tetford PE, Desloges JR, Nakassis D (2017) Modelling surface geo-
Maeda EE, Pellikka PKE, Siljander M, Clark BJF (2010) Potential morphic processes using the RUSLE and specific stream power
impacts of agricultural expansion and climate change on soil ero- in a GIS framework, NE Peloponnese, Greece. Model Earth Syst
sion in the Eastern Arc Mountains of Kenya. Geomorphology Environ. https​://doi.org/10.1007/s4080​8-017-0391-z
123:279–289. https​://doi.org/10.1016/j.geomo​rph.2010.07.019 Thomas J, Joseph S, Thrivikramji KP (2018) Estimation of soil erosion
Markose VJ, Jayappa KS (2016) Soil loss estimation and prioritization in a rain shadow river basin in the southern Western Ghats, India
of sub-watersheds of Kali River basin, Karnataka, India, using using RUSLE and transport limited sediment delivery function.
RUSLE and GIS. Environ Monit Assess. https​://doi.org/10.1007/ Int Soil Water Conserv Res 6:111–122. https​://doi.org/10.1016/j.
s1066​1-016-5218-2 iswcr​.2017.12.001
Moore ID, Burch GJ (1986) Modelling erosion and deposition: topo- Toubal AK, Achite M, Ouillon S, Dehni A (2018) Soil erodibility
graphic effects. Trans ASAE 29:1624 – 1630 mapping using the RUSLE model to prioritize erosion control in
Nearing M, Foster G, Lane L, Finkner S (1989) A process-based soil the Wadi Sahouat basin, North-West of Algeria. Environ Monit
erosion model for USDA-Water Erosion Prediction Project tech- Assess. https​://doi.org/10.1007/s1066​1-018-6580-z
nology. Trans ASAE 32:1587–1593 Toumi S, Meddi M, Mahé G, Brou YT (2013) Cartographie de
Neitsch S, Arnold J, Kiniry J, Williams J (2011) Soil & water assess- l’érosion dans le bassin versant de l’Oued Mina en Algérie par
ment tool theoretical documentation version 2009. Texas Water télédétection et SIG. Hydrol Sci J 58:1542–1558. https​://doi.
Resources Institute, pp 1–647 org/10.1080/02626​667.2013.82408​8
Pham TG, Degener J, Kappas M (2018) Integrated universal soil loss Wang G, Wente S, Gertner GZ, Anderson A (2002) Improvement
equation (USLE) and geographical information system (GIS) for in mapping vegetation cover factor for the universal soil loss
soil erosion estimation in A Sap basin: Central Vietnam. Int Soil equation by geostatistical methods with Landsat Thematic
Water Conserv Res 6:99–110. https​://doi.org/10.1016/j.iswcr​ Mapper images. Int J Remote Sens 23:3649–3667. https​://doi.
.2018.01.001 org/10.1080/01431​16011​01145​38
Planning Ministry of the Environment and Spatial (2000) National Wijesundara NC, Abeysingha NS, Dissanayake DMSLB (2018) GIS-
report on the state of the environment based soil loss estimation using RUSLE model: a case of Kirindi
Remini B (2000) L’envasement des barrages. Bull Réseau Eros Oya river basin, Sri Lanka. Model Earth Syst Environ. https:​ //doi.
20:165–171 org/10.1007/s4080​8-018-0419-z
Renard K, Foster G, Weesies G et al (1997) Predicting soil erosion by Williams JR (1975) Sediment routing for agricultural watersheds.
water: a guide to conservation planning with the Revised Univer- JAWRA J Am Water Resour Assoc 11:965–974
sal Soil Loss Equation (RUSLE). In: Agriculture handbook, No. Williams JR, Berndt HD (1977) Sediment yield prediction based on
70. US Government Printing Office, Washington, pp 404 watershed hydrology. Trans ASAE 20:1100–1104
Roose É, Noni G De (2004) Recherches sur l ‘érosion hydrique en Williams JR, Nicks AD, Arnold JG (1985) Simulator for water
Afrique: revue et perspectives. Sci Chang planétaires/Sécheresse resources in rural basins. J Hydraul Eng 111:970–986
15:121–129 Wischmeier WH, Smith DD (1965) Predicting rainfall erosion losses
Ruhe RV, Scholtes WH (1956) Ages and development of soil land- from cropland east of the Rocky Mountains [online]. In: Agricul-
scapes in relation to climatic and vegetational changes in Iowa 1. tural Handbook, No. 282. US Department of Agriculture - Agri-
Soil Sci Soc Am J 20:264–273 cultural Research Service, Brooksville, pp 47
Souidi Z, Hamimed A, Donze F (2014) Cartographie du risque de Wischmeier WH, Smith DD (1978) Predicting rainfall erosion losses
dégradation des terres en région semi-aride: Cas des Monts de - a guide to conservation planning. In: Agriculture Handbook No
Beni Chougrane dans le Tell Occidental Algérien. Geo Eco Trop 537. U.S. Department of Agriculture, Washington, DC 
38:85–102 Zhang X-C (2012) Cropping and tillage systems effects on soil erosion
Tang J, Cheng XQ, Zhu B et al (2015a) Rainfall and tillage impacts under climate change in Oklahoma. Soil Sci Soc Am J 76:1789.
on soil erosion of sloping cropland with subtropical monsoon https​://doi.org/10.2136/sssaj​2012.0085
climate—a case study in hilly purple soil area, China. J Mt Sci
12:134–144. https​://doi.org/10.1007/s1162​9-014-3241-8 Publisher’s Note Springer Nature remains neutral with regard to
Tang Q, Xu Y, Bennett SJ, Li Y (2015b) Assessment of soil erosion jurisdictional claims in published maps and institutional affiliations.
using RUSLE and GIS: a case study of the Yangou watershed in

13