Mapping and Quantification of Soil Erosion and Sediment Delivery in Poorly Developed Urban Areas: A Case Study

Abstract

Poorly developed regions in the Gaza Strip, Palestine, face significant risks to public safety, the environment, and stormwater infrastructure due to soil erosion and deposition. This study is the first of its kind to investigate soil erosion in this area. A revised universal soil loss equation (RUSLE) model was utilized and validated using field measurements of accumulated sediments at three major stormwater detention basins from 2014 to 2020. High-resolution maps were created to capture the urbanization effect and to further improve the future prediction of urbanization. The findings revealed that the highest potential for sediment generation in the Gaza governorate occurred over the slopes of the eastern ridge, which drain toward the city center. Sediment generation ranged from 1784 to 4281 ton/ha for the years of 2018 and 2020, respectively. The average sediment delivery ratio (SDR) was calculated to be 0.00134. The estimations for sediment export ranged from 0 to 135.3 ton/ha for the year 2020, with an average of 0.0737 ton/ha. The urban areas exhibited the least sediment export rate; however, the model revealed abnormal behavior for a dataset of the field measurements which was ascribed to the impact of destruction/reconstruction activities in the corresponded watersheds that followed the war in 2014. This conducted research stands as a pioneering effort in quantifying and cartographically representing sediment erosion potential within the Gaza Strip. Thus, it serves as an indispensable point of reference for future researchers in terms of the employed parameterization and calibration methodology. Furthermore, it holds distinct significance as an unparalleled resource for experts and stakeholders who are invested in comprehending the ramifications of erosion on urban landscapes and drainage systems.

1. Introduction

Soil erosion is one of the most critical concerns to the environment and economy and it poses both direct and indirect risks [1]. The majority of research has been done with an eye toward protecting soil as a natural resource and preventing relevant natural disasters like landslides and soil fertility deterioration [2,3]. However, it has been noted that waterborne sediments in urban areas have additional effects that put the public health at risk because of the accumulation of biological and chemical contaminants in the deposited sediments [4,5]. The overflows of combined stormwater and sewer drainage systems, vehicular traffic, and improper waste disposal are potential sources of these contaminants [6,7,8]. Additional impacts on urban environments include reduced soil permeability, failed conduits and drainage infrastructure, limitations to the optimum use of stormwater, increased maintenance costs, and general environmental deterioration, including the air, soil, and water [4,9].

There are numerous approaches for quantifying and describing soil erosion in watersheds with natural or arable land cover. The revised universal soil loss equation (RUSLE), developed by Renard et al. [10], is the most widely used method among the approaches [1,11]. However, there are far fewer examples of this approach being used in densely populated or partially populated areas.

1.1. Background

Systematic reviews conducted by Borrelli et al. [11] and Pandey et al. [1] have shown that the revised universal soil loss equation (RUSLE), which was described by Renard et al. [10], is the most employed method by researchers in the field. The popularity of this method is ascribed to its simplicity and its demonstrated validity for various climate conditions and topography characteristics. It also offers better flexibility in terms of the temporal increments that can be applied [1,11]. This encouraged employing it for the current study, since the acquisition of high-quality data for the Gaza Strip is quite challenging.

Research utilizing the RUSLE for soil erosion quantification is abundant in the literature [12,13,14,15,16,17]. It is generally used for serving conservation practices by assessing the erosion risk of natural soil. However, there are fewer studies that have been concerned with watersheds that involve urban structures [18,19].

Furthermore, it is important to underscore that a literature analysis has revealed a significant challenge in the application of RUSLE within densely populated areas. This challenge pertains to the absence of well-founded justifications for estimating parameters that accurately represent the physical land cover and urban practices. Specifically, this pertains to the factors associated with cover management (C) and erosion control practices (P). The lack of consensus among researchers regarding standardized or unified parameterization for diverse land-use scenarios is apparent, and a multitude of issues must be addressed before such parameters can be consistently and reliably employed.

1.2. Study Area

The current study focused on sediment generation and deposition in the Gaza governorate in Palestine. The aforementioned problems are caused by soil erosion and sediment deposition that typically occur in the study area [20]. Gaza is one of the southern coastal governorates of Palestine which is located by the Mediterranean Sea. It is one of the largest and most densely populated governorates in Palestine with an area of 73 km² and a population of around 780,000 [21]. This population represents 15.0% of the whole population in Palestine [21]. The Gaza governorate includes four main municipalities: Al-Zahra, Al-Mughraga, and Wadi Gaza, and Johr Eddik, as shown in Figure 1. The governorate borders are outlined by the natural topographic terrain of Wadi Gaza from the south, and by the sea shoreline from the west. On the one hand, the land topography at the eastern borders of the Gaza governorate has been amended such that the natural hydrological overlapping beyond the borders became extremely limited. On the other hand, the urban infrastructure at the northern borders of the governorate has been designed such that they are explicitly drained by the governorate’s stormwater drainage infrastructure; and thus, a very limited area of the North governorate drains inward to the boundaries of the Gaza governorate. Accordingly, the study area extends over the expanse of the hydrological zone of the Gaza governorate, as shown in Figure 1. This zone contains all the sub-watersheds that fall within or intersect with the governorate administrative boundaries. The stormwater management plan for the described region involves drainage of stormwater to three major basins that involve infiltration facilities. These basins are Sheikh Radwan, Sadaqa, and Asqula, as shown in Figure 1.

1.3. Problem and Study Motive

Political instability in the Gaza Strip and the blockade that started two decades ago have caused severe environmental and socioeconomic issues. The municipalities have struggled to fulfill the critical services of water supply and sanitation, and investments in the stormwater management sector have been downgraded [20,22,23]. Thus, problems associated with waterborne sediments have been developing, while they are being insufficiently addressed in the Gaza Strip. The local municipalities in Gaza allocate significant resources for the clean-up of street sediments and for restoring the capacity of stormwater drainage and management infrastructure on an annual basis. The Municipality of Gaza reported that 4885 tons of deposited sediments were swept from streets and removed from 115 km of Gaza streets between March and May in 2019. In addition, a significant load is annually removed from stormwater detention and infiltration basins; these data are elaborated later in the present paper. Unfortunately, no records are available on the impact of the accumulation of the sediments on the drainage system; however, cases have been reported where the sediments have clogged 80% of the diameter of main carriers, as shown in Figure 2a.

In well urbanized areas, the drainage system facilitates sediment delivery to their sink; however, sediment deposition is common in many paved streets in Gaza City where no drainage system exists, or when the existing drainage system is surcharged (See Figure 2b). These sediments represent a critical risk to the public health because the wastewater drainage system is equally surcharged by the infiltrating stormwater [20]. This directly exposes residents to these contaminated sediments, especially when contaminated depositions turn into airborne dust after they dry out, as shown in Figure 2c.

Such sediments have several sources in poorly developed regions of the city. It is suspected that the major mechanism of the sediment erosion in the Gaza Strip is rill erosion. This happens to the natural topsoil of agricultural lands located within the urbanized watersheds, backyards, public landscapes, undeveloped regions, and, with high emphasis, the unpaved streets, as shown in Figure 2d.

Nonetheless, to the best of the author’s knowledge, no work has yet been published that extensively explores this phenomenon in the region. This motivated the current study, which attempts to quantify and map the creation and deposition of sediments in the Gaza governorate, Palestine. This finally aids in the selection of the most effective soil erosion control strategies, which improves the operability of the stormwater management infrastructure and reduces threats to the environment and public health.

1.4. Aims and Objectives

To the best of the authors’ knowledge, no prior studies have been published that have delved into the investigation of soil erosion in the Gaza Strip, particularly within urbanized zones. This unique context necessitates a meticulous parametrization process to curate the most suitable dataset and methodologies for the development of the RUSLE model in this region.

This study focuses on quantification of sediments that result from soil erosion of natural soil. This is described using the five parameters of the RUSLE. No specific concern is given to the additional sources of sediments in urban areas, such as construction sites and asphalt decay gravel-covered parks, and gardens, as described by Russell et al. [24]. However, it is assumed that such sources will be considered during the parametrization of the RUSLE factors; namely, this will be applied to the soil erodibility and cover management factors. Avoiding generalization of different land-use subclasses and distinguishing them according to their characteristics is suggested by the current study to overcome the lack of investigation means.

In this paper, we describe the research employed RUSLE model and sediment delivery ratio (SDR) to identify the behavior of sediment erosion and delivery in densely inhabited regions in the Gaza governorate that have a substantial proportion of natural soil cover. Due to a lack of data and measuring tools, the research was geographically limited to three main watersheds in Gaza City, since the delivered sediments from these watersheds can be quantified at storm water basins where runoff is collected.

Consequently, the following goals were established:

• To compile and filter field data to the highest possible quality and resolution to depict micro-scale geographical variations and to emphasize aspects associated with parameterization of a RUSLE model for the case study location;
• To develop a reasonable calibration model that accounts for stream connectivity in order to enhance the accuracy of the model;
• To describe the urbanization impact on the model results and to highlight any deviations related to the abnormal processes that affected the urbanization structure in the case study;
• To identify regions of high erosion potential and the SDR.

This finally aids in the selection of the most effective soil erosion control strategies, which improves the operability of the stormwater management infrastructure and reduces threats to the environment and public health.

2. Methodology

2.1. Field Measurements and Compilation of Essential Hydrological Data

2.1.1. Field Measurements of Accumulated Sediments

Field measurements of the accumulated sediments at the three stormwater detention basins were acquired from the Municipality of Gaza Water and Wastewater Management Unit. The field measurements were collected explicitly for the current study for the years 2019 and 2020. For the years before 2019, restoration and cleaning activities were not conducted on an annual basis; thus, the available data were provided as a cumulative weight that occurred over several years of no cleaning. The data were estimated based on the contractors’ records for the number of trucks used for sediment removal from the sites. The field measurements of sediments delivered to the stormwater basins are elaborated in

Table 1. Sediments removed from stormwater infiltration basins after the rain season.

No.	Stormwater Basin	Measurement Year(s)	Quantity (ton)
1	Sheikh Radwan	2014–2018	25,755
		2019	4148
		2020	4420
2	Asqula	2018	1581
		2019	1785
		2020	1700
3	Sadaqa	2017–2018	2465
		2019	1207
		2020	1513

.

The field measurements were used for calibration and validation of the RUSLE model developed for the current study.

2.1.2. Compilation of the RUSLE Model Data

The essential data required for establishing the RUSLE model were compiled and processed according to the requirements of the RUSLE model, as shown in

Table 2. Essential data compiled for the RUSLE.

Data	Format	Source	Used For
Rainfall data	Monthly records at meteorological stations for the years from 2014 to 2020, in addition to the coordinates of these stations	Periodic Publications of Ministry of Agriculture (MoA)	R factor
Land-use data	A digital thematic map converted to a grid of 1 m resolution	Ministry of Planning and Administrative Development (MoPAD)	C, and P factors
Building and street map	A digital thematic map converted to a grid of 1 m resolution	Open Street Map Contribution Planet [25]	C, and P factors
Agriculture map and crop types	A digital thematic map converted to a grid of 1 m resolution	Agriculture Atlas [26]	C factor
DEM	A digital grid of 1 m resolution and a float value for elevation	Interpolated surface from an extensive dataset of field surveyed elevations [22]	LS and connectivity index (IC)
Hydrological delineation	A digital thematic map showing the urban hydrological delineation of watersheds in the study area	AAH employed a participatory approach for delineation of the urban watersheds [20,27].	LS and IC
Topsoil classification	A grid digitized into 1 m resolution	[28]	K factor
Topsoil organic content	A grid digitized into 1 m resolution	[28]	K factor
Field measurements	Records from total sediments removed from stormwater detention basins	Municipality of Gaza Water and Wastewater Management Unit	Calibration and validation of the model

. The collected data involved digital maps for the environmental data, as well as field measurements of the sediments for the three major stormwater detention and infiltration basins. The data are elaborated in the following relevant sections.

The rainfall data were acquired on a monthly basis for the years between 2014 and 2020 for the major rainfall stations in the Gaza Strip. These records were used to interpolate a grid map for the total annual rainfall for each year. The spherical ordinary kriging method was used for interpolation. Figure 3a shows the locations of the stations in addition to the normal annual rainfall data that were acquired over the last 30 years. The normal annual rainfall depth in the Gaza governorate ranges from 350 to 432 mm.

The lack of high-resolution/high-precision digital elevation model (DEM) for the Gaza Strip forced the authors to interpolate it based on an extensive field surveyed dataset. The precision of the field measurements was 0.01 m for both elevation and the location coordinates. Thus, the interpolated DEM was of 1 m resolution, and had a vertical precision of 0.01 m. A non-regulated spline method was used to interpolate the surface. The outcome was evaluated using a cross-validation approach and it showed a quality sufficient for the analysis required for the current study. The land elevation was observed to range from the mean sea level (MSL) to 85.7 m in the study area, as shown in Figure 3b. The land topography of the Gaza Strip is characterized by two hilly ridges that extend in parallel to sea shoreline all along the study area. While drainage at the western areas is naturally directed towards the sea, the eastern parts have small slopes that descend towards the eastern borders. The remaining inner lands which are confined between the two parallel ridges behave as an endorheic catchment area that drains to specific local low points.

The delineation of the hydrological zone of the Gaza governorate was generalized according to their destination, as shown in Figure 3c. The map shows the seven destinations of stormwater drainage in the Gaza governorate. They include the three major stormwater infiltration basins, as well as, the eastern borders, Gaza Wadi, and the sea. Two endorheic catchment areas were identified where stormwater infiltrates into the natural soil or is drained by wastewater drainage systems. The delineation of these catchment areas was prepared by AAH [20], by employing an integrated participatory approach that is described by Dawoud and Mansour [27]. The delineation of the catchment areas was initially by spatial analysis. Thereafter, field visits and extensive meetings with the staff of the local municipalities were conducted. This approach helped to identify the accurate flow paths and delineation of the urban catchment areas, which are highly governed by the urban infrastructure. For the current study, three catchment areas are of a special concern which are the ones draining to the three major stormwater infiltration basins: Shiekh Radwan Basin, Asqula Infiltration Lagoon, and Sadaqa Infiltration Pond, as shown in Figure 3c. Information was collected on sediment delivery to these three basins.

2.2. Soil Loss Model

2.2.1. Revised Universal Soil Loss Equation (RUSLE)

The essential form of the RUSLE, which was developed by Renard and Friemund [3], was employed by the current study as follows:

A = R × K × L × S × C × P

(1)

where A is the soil loss (t ha⁻¹yr⁻¹), R is the rainfall erosivity factor (MJ mm ha⁻¹h⁻¹), K is the soil erodibility factor (ton ha h MJ⁻¹mm⁻¹), L is the slope length factor, S is the slope steepness factor, C is the cover and management factor, and P is the support and conservation practices factor. The selection of the method was enforced by the availability of the data since the study area can be described as a data-poor region [29].

Both of the systematic reviews conducted by [1] and [11] showed that the revised universal soil loss equation (RUSLE) is the method that is mostly used by researchers in the field. This can be ascribed to the few limitations the method has and its flexibility against low-resolution datasets. Moreover, the method can be easily automated and applied to spatial data [30,31]. However, the RUSLE is not a great help once sediments are eroded [1]. Thus, it should be annexed by some approach for modeling the sediment delivery ratio (SDR).

2.2.2. Compilation of Spatial Data and Parameters

• Rainfall Erosivity Factor (R Factor)

The rainfall erosivity factor represent the kinetic energy from the impact of raindrops on the sheet and rill erosion of soil [2,32]. The original forms of the equations, which were developed by Wischmeier et al. [2], are the most widely used although they were explicitly developed for the United States region [33]. These equations are based on the long-term record of storm kinetic energy (E) which is derived from the maximum 30 min intensity records (I₃₀). For the case of the Gaza Strip, the best rainfall records are logged on a daily basis since the cumulative rainfall depth is manually observed every 24 h. For the years between 2018 and 2020, rainfall records were acquired on an annual basis. Such lack of high temporal-resolution records is common for many regions around the world, which has encouraged researchers to discover alternative regression equations that utilize annual or monthly rainfall records [34]. However, the use of such equations should be carefully applied and be regionally restricted, since the R factor could significantly vary over a limited spatial expanse following climate and storm patterns. This can easily be inferred from the global maps produced by Naipal et al. [35] and Panagos et al. [32], two studies that conducted extensive analyses of a wide range of datasets from all over the globe. Panagos et al. [33] employed a Gaussian process regression (GPR) model to interpolate the erosivity factor based on data collected from 3625 stations that are scattered over the globe. The erosivity factor at these stations was calculated using the Brown and Foster equation [36]. A global erosivity map at 30 arc-seconds resolution was produced. Despite its significance for understanding the global patterns of erosivity factor, it clearly stated that such maps are not “intended to be a substitute for any local/regional R factor database with higher-resolution data” [35].

The approach which was employed by Naipal et al. [35] provided reasonable guidance for estimating the R factor on a regional scale. Regression equations were developed and validated on the basis of globally available data. The data were spatially collated and classified according to the Köppen–Geiger climate classification [37]. This helped users to employ the equations which applied to their regions with respect to the prevailing climate conditions [36,37]. According to the Köppen–Geiger climate classification, the Gaza Strip is characterized by a hot dry summer which is annotated as C_sa. The average temperature of the hottest month is 26.5 °C, while the normal monthly precipitation ranges from 0 to 127.1 mm for the driest and wettest months, respectively. Naipal et al. [35] recommended the use of the Renard and Friemund [3] method for regions of the climate class C_sa, as it performs as good as the regression method.

This equation is as follows [3]:

$$\mathrm{R}=0.0483{\mathrm{P}}_{\mathrm{n}}^{1.610}$$

(2)

For P_n greater than 850 mm:

$$\mathrm{R}=587.8-1.219{\mathrm{P}}_{\mathrm{n}}+0.004105{\mathrm{P}}_{\mathrm{n}}^2$$

(3)

This equation was applied to the grid maps of the annual rainfall P_n of the Gaza Strip for the years of concern. Different R-factor maps were produced for the years between 2014 and 2020.

The grid maps of the rainfall were produced by spline interpolation of the recorded data at the working rainfall gauge stations in the Gaza Strip. The surface produced by the spline interpolation method can be defined piecewise by polynomials which pass through all the available exact points. This is conducted in fulfillment of the minimum curvature condition all over the surface; however, local curvature can be adjusted based on the natural behavior of the modeled phenomenon [38,39]. This method is widely used in modeling of natural phenomena such as surface water table, groundwater pollution, and rainfall, where no singularities and discontinuities are foreseen [40].

Once needed, the R-factor maps were accumulated to account for the sediment measurements that were conducted over more than one year, which are presented in Table 1.

• Soil Erodibility Factor (K Factor)

Soil erodibility is a measure of soil susceptibility to sheet and rill erosion by surface runoff [34]. High K-factor magnitudes indicate high vulnerability of soil. The K factor is governed by soil texture and the organic matter content (OMC) [41].

Table 3. Soil erodibility K-factor (ton/hectare) [41].

Textural Class	Organic Matter
	<2%	>2%
Sand	0.07	0.02
Sandy loam	0.31	0.27
Sandy clay loam	0.45	0.45
Sandy clay 1	0.50	0.46
Loam	0.76	0.58
Clay loam	0.74	0.63
Clay	0.54	0.47
Buildings and paved roads 2	0	0

¹ Estimated as the average of K-factor magnitudes of clay and sandy clay loam classes. ² Proposed for the current study.

shows the typical K-factor magnitudes that were applied to the soil classes that are present in the Gaza governorate. These magnitudes were assigned to the classes emerging from the intersection of the two maps shown in Figure 3d,e, which were prepared by [28].

It was found that assigning zero magnitude to the built-up and paved surfaces is reasonable as it suggests that the surface is non-erodible. Therefore, the impervious map, which is shown in Figure 3f, was appended to the final K-factor map.

The map of topsoil texture shows that sandy and sandy loam soils are concentrated at the western part of the study area close to the coast. The clay content increases towards the eastern boarders of the study area. No texture class can be described as dominant, as can be inferred from Figure 1.

In order to account for paved surfaces in urban areas, previous studies have manipulated the magnitudes of the cover management factor (C) in the RUSLE equation [34]. This has usually been conducted by assigning low values starting from zero to the urban areas [42,43]. Although this practice results in negligible rates of erosion for such classes of land use, it results in erroneous results for subsequent calculations of sediment delivery ratio and connectivity index. Moreover, the physical logic beyond a zero K factor suggests that the surface is non-erodible, while a zero C factor indicates a sink or a trap for the sediments. Therefore, the map of non-erodible surfaces (namely buildings and paved streets), which is shown in Figure 3, was appended to the K-factor map while being assigned a zero value.

• Slope Length and Steepness Factor (LS Factor)

The slope-length and gradient factor was defined in its original form as the distance from the sediment generation surface to the point where the slope gradient decreases to the magnitude that causes deposition of sediments, or when it reaches a defined drainage channel [2].

Previous studies have employed a variety of equations for calculating the LS factor [2,44]. There is insufficient guidance in the literature for the best formula to apply in the study area. Therefore, the original form developed for the USLE was employed as follows:

$${\mathrm{LS}}_i=\left(\mathsf{\lambda }/22.13\right)\mathrm{m}\left(65.4{\sin }^2\mathsf{\beta }+4.5\sin \mathsf{\beta }+0.0654\right)$$

(4)

where β refers to the slope angle expressed in radians, and λ is the slope length in meter. The m factor is calculated as follows:

$$\mathrm{m} =\left\{\begin{array}{c}0.5 \mathrm{if} \mathsf{\beta }>0.05\\ 0.4 \mathrm{if} 0.03<\mathsf{\beta }<0.05\\ 0.3 \mathrm{if} 0.01<\mathsf{\beta }<0.03\\ 0.2 \mathrm{if} \mathsf{\beta }<0.01\end{array}\right.$$

(5)

For the current study, the slope length λ was calculated by the flow accumulation multiplied by the cell size [44,45].

• Cover Management Factor (C Factor)

The cover management factor (C factor) accounts for the mitigation of the impact of raindrops on the soil surface by the vegetation cover, which dissipates the energy before the raindrops reach the soil surface. Thus, the C factor was originally linked to the vegetation cover type in the guiding tables developed by the USDA [46], as well as, in the modified and extended version provided by Wischmeier and Smith [2].

Despite extensive research that has investigated the C factor, a limited number of studies can be found which have identified the C factor for urbanized land-use classes. The USDA [46] assigns a zero value to the “built-up” land-use classes. However, this magnitude seems to be generalizing a range of classes which have a differential potential for sediment erosion. A key problem with the zero assumption arises when calculating the sediment delivery ratio, because some formulas can result in erroneous magnitudes dividing by the zero value. From a practical perspective, a zero magnitude indicates that the soil is totally protected from erosion by another factor. It also identifies a sink or a sediment entrapment structure for sediments. However, this is not the case for many “built-up” regions, especially in poorly developed regions where the natural soil is bare in a considerable proportion of the total area. In fact, urban infrastructure along a drainage path actually facilitates the delivery of sediments rather than retarding or controlling them [24]. Therefore, the zero assumption was avoided when assigning the magnitudes of the subfactors of the cover management factor. Instead, the use of zero magnitude was limited to the K factor of the built-up and paved surfaces.

The land cover map was compiled for the current study by combining land-use, open-street [25], and agricultural [26] maps, as shown in Figure 3g. The classified trees mainly include olive and citrus trees, while the unclassified trees include additional fruit trees such as almond, guava, fig, and grape trees. The mixed trees class refers to the case where trees are planted with sufficient distances in between which allow for seasonal crops to be cultivated. The crop class describes the seasonal crops of vegetables and seeds. This study followed the description provided by Panagos et al. [33] for such classes in order to identify the C factor.

Theobald et al. [47] elaborated the C factor for a set of land-use classes based on the probable fractions of paved and bare-soil surfaces. The magnitudes provided in that study were employed wherever they matched the land-use characteristics in the Gaza Strip, as shown in

Table 4. Cover management factors assigned to the land-use/land-cover classes of the study area.

No.	Land-Use Class		Percentage of Land Cover (%)	C Factor	P Factor
1	Classified trees		12.3	0.2	0.35
2	Unclassified trees		10.5	0.275
3	Mixed trees		2.5	0.135
4	Crops		9.6	0.21
5	Houses		12.2	0.001	0.001
6	Paved streets		4.3	0.02	0.0001
7	Unpaved streets		3.3	0.7	1
8	Water surfaces		0.1	0.0	0
9	Unclassified (mixed use)	Built-up	15.6	0.2	0.04
10		cultivated/undeveloped	16.7	0.7	0.85
11		Urban development	12.9	0.5	0.45 *

* Taken as the average between “built-up” and “cultivated/undeveloped”.

.

Unpaved streets are common in the Gaza Strip, and they are covered by natural soil that can be found in the area. For some cases, sand or gravel is imported from other locations to pave these roads. However, such streets are prone to erosion and no measures are applied to mitigate the erosion. Therefore, they were considered to be bare land and a factor of 0.7 was proposed in this case.

The remaining unclassified areas were spatially differentiated based on the master plan of the Gaza Strip, as shown in Figure 3h. These areas represent the open spaces and undeveloped lands in “built-up” zones. However, on the one hand, sediment generation in these areas is usually controlled by human-built structures or by green cover. Therefore, they are best described as green urban landscapes. On the other hand, less control is applied to prevent sediment erosion from an “urban development” zone. Fewer built-up areas and more unpaved streets appear in this zone, while the unclassified areas in the cultivated/undeveloped zone are prone to high erosion since no measures are applied, and the natural topsoil is exposed to the erosion mechanisms.

• Erosion Control Practice Factor (P Factor)

The P factor describes the management practices which modify the flow pattern which alters soil erosion potential, such as contouring, strip cropping, or terracing [10]. Previous studies have proposed P-factor magnitudes for a range of management practices. However, due to the lack of information on such practices, some researchers have tended to ignore the P factor, assigning it a magnitude of 1.0 [48,49].

The Gaza governorate has a mild topography as no steep slopes appear in large scales. Agricultural lands are concentrated in the eastern and southern parts of the governorate, which are characterized by small land slopes. Therefore, erosion control practices are not common in these agricultural lands. In urban areas, it is suggested that the human-built barriers (such as walls, fences, curb stones, street bumps, earth ramps, etc.) play a role in retarding sediment delivery by physical mechanisms like the ones utilized by the practice of soil conservation for farming purposes. Such conditions would not be well simulated if the P factor is ignored by assigning it a magnitude of 1.0. Therefore, the open spaces that fall within the zones classified as “built-up” and “cultivated/undeveloped” were assigned magnitudes of 0.04 and 0.85, respectively [47]. For the “urban development” land class, the P factor was calculated to be 0.45, which is the average of the previous two magnitudes.

Table 4 shows the magnitudes that were assigned to the remaining land-use classes, which include trees, crops, houses, and paved streets. The magnitudes were proposed after assessing previous studies for similar cases as described by Theobald et al. [47].

Unpaved streets were assigned a magnitude of 1.0 since they are the most vulnerable class for soil erosion. They are exposed to traffic dynamic loads that pulverize the soil and increase the potential of rill erosion of the topsoil. Such lost soil, once needed, is usually subsidized from external sources during the annual maintenance conducted by the municipality. However, no interventions are applied to mitigate the erosion. Thus, soil erosion potential is at its maximum in the unpaved streets in the Gaza governorate.

2.3. Sediment Delivery Ratio and Calibration of the Model

The sediment delivery ratio (SDR) describes the efficiency at which the eroded sediments are transported from upstream hillslopes down to the sinks and along the corresponding streams. The transport of generated sediments from the source to the sink is subject to geomorphological, hydrological, and ecological influential factors, which can facilitate or retard the lumped delivery of sediments. These influential factors are considered by the connectivity index (IC), which describes the linkage between runoff and sediment sources in upper parts of catchments and the corresponding sinks [50].

The connectivity index, which was introduced by Borselli et al. [51], is one of the successful models that has been employed by many RUSLE-based studies [5,52,53,54].

The model formula is [51]:

$$\mathrm{IC}={\log }_{10}\frac{ \overline{\mathrm{C}} \overline{\mathrm{S}}\sqrt{\mathrm{A}}}{{{\displaystyle \sum }}_{\mathrm{i}}\frac{{\mathrm{d}}_{\mathrm{i}}}{{\mathrm{C}}_{\mathrm{i}}{\mathrm{S}}_{\mathrm{i}}}}$$

(6)

where $\overline{\mathrm{C}}$ and $\overline{\mathrm{S}}$ are the average C factor and slope gradient (m/m) for the upslope contributing area, respectively; d_i is the is the length of the flow path along the ith cell according to the steepest downslope direction (m). The connectivity index is calculated for the current study on the basis of the cell scale.

The IC index is used to estimate the sediments delivery ratio as follows [52]:

$$\mathrm{SDR}=\frac{0.8}{1+\exp \left(\frac{{\mathrm{IC}}_0-{\mathrm{IC}}_{\mathrm{i}}}{\mathrm{k}}\right)}$$

(7)

where IC₀ and k are calibration factors, and IC₀ is the IC value at the ith cell. The SDR for any cell represents the sediments that reach the stream as a proportion of the total sediments generated at that cell. This proportion of sediments is referred to as the sediment export from that E_i (units: ton/ha/yr); and thus, it is calculated as follows:

$${\mathrm{E}}_{\mathrm{i}}=\mathrm{SDR}.\mathrm{A}$$

(8)

Therefore, the approach described by Vigiak et al. [52] can be used to calculate the lump sediments at any point along the stream as follows:

$$\mathrm{E}=\sum {\mathrm{E}}_{\mathrm{i}}$$

(9)

This method does not provide information on the spatial distribution of the sediment deposition. However, it fits the purpose of the current study since the model was calibrated using the field measurements of the accumulated sediments at the endpoint of the drainage line of streams, as listed in Table 1. Equation (7) was iterated for IC₀ and k seeking the optimal magnitudes. These calibration parameters were identified based on the residual of the lumped estimates of sediment exports from the three catchment areas.

A spatial analysis model was developed according to the flow-chart shown in Figure 4. The analysis was limited to the delineated catchment areas which are drained to the three stormwater basins [27].

Both IC₀ and k were iterated, producing SDR maps that were linked to a grid of the two calibration parameters. The SDR map was used to calculate the sediment export for each one of the three basins at different observed time increments. This was used to produce solution functions that correlate IC₀ and k. The use of any combinations of the calibration factors that satisfy any function results in zero-residual estimations. However, each function is an explicit solution for the basin and the time increment it is developed for. For an ideal RUSLE model, these functions intersect at a single point that represent a solution for any catchment area in the study area. Of course, this is unlikely to happen due to the uncertainty involved in the collected and measured data, as well as the assumptions conducted for the RUSLE parameters. Therefore, the following objective function was employed to identify the optimal calibration factors that result in minimal residual:

$$\min \mathrm{z}=\mathrm{SD}\left\{{\mathrm{f}}_{\mathrm{i},\mathrm{j}}\left({\mathrm{IC}}_{\mathrm{o}},\mathrm{k}\right)\right\}$$

(10)

where SD is the standard deviation, f_i,j(IC₀, k) is the solution function for a basin i, at a time increment j.

3. Results

3.1. RUSLE Parameters

The produced maps for the rainfall factor (R factor) showed a temporally and spatially differential margin of magnitudes. These variations follow the annual rainfall that has been recorded at the rain gauge stations in the Gaza Strip. Figure 5a shows the R-factor map for 2019, as an example. The R factor was at its minimum in 2018 with a range from 320 to 590 MJ mm ha⁻¹ h⁻¹. While greater magnitudes can be noticed for 2019, a wider range that peaked at 1242 MJ mm ha⁻¹ h⁻¹ was noticed for the R factor in 2020. The averages of the R Factor were identified as 425, 625, 861 MJ mm ha⁻¹ h⁻¹, for 2018, 2019, and 2020, respectively; and the standard deviations were 65, 47, and 130. The spatial distribution of the R factor shows that the highest magnitudes are located at the northern parts of the governorate. This follows the general behavior of the spatial distribution of the annual rainfall which ascends on the way from south to north.

The final map for the LS factor was developed as shown in Figure 5b. The LS factor ranged from 0.06 to 42, with 0.71, and 0.76 as mean and standard deviation, respectively. As the statistical analysis indicates, high magnitudes of LS appear at small scales. As can be inferred from the LS-factor map, these areas are concentrated along the shore and some internal areas along the two parallel ridges that characterize the region’s terrain. Despite the fact that the delineation of the urban areas is governed by the human infrastructure, the conditioning of the map to reflect the amended streamlines was avoided since it resulted in unrealistic slopes.

The soil erodibility map resulted in a categorical map, as shown in Figure 5c. The K factor ranged from 0.02 by the coast to 0.78 at the eastern and southern parts of the study area. The highly urbanized areas were found to be located at the low erodibility regions. It is noteworthy that the K-factor map which was included in the analysis was appended by the impervious map shown in Figure 3f to account for the zero erodibility. This was found to provide a more reasonable representation of the soil erosion potential.

The C factor was mapped in Figure 5d. Apparently, this map is a parametrization of the land-use classes shown in Figure 3g; and thus, the map shows that the highly developed areas exhibit the lowest magnitudes of the C factor. The C factor ranged from 0.0001 to 0.7.

The P-factor map, which is shown in Figure 5e, reflects the general classes of the master plan of the Gaza governorate. A low P factor prevailed in “built-up” zones with magnitudes within the range 0.0001–0.35. Higher magnitudes appeared in “urban development” zones that ranged mainly between 0.35 and 0.5. Greater magnitudes appeared in “cultivated/undeveloped” regions. In fact, the P-factor map reflects the spatial distribution of population density and the associated human activities that might mitigate the soil erosion activities. As mentioned before, the P factor was assigned to reflect the human activities that hinder soil erosion even if they are not designed for that purpose.

3.2. Sediment Generation Potential

The sediment erosion potential (A) using the RUSLE was calculated as the product of the maps shown in the subfigures of Figure 5. An example of the resulting maps is shown in Figure 6 for the year 2019. The differential annual rainfall rates caused significant changes in the outcomes for different years even if they are difficult to visually capture. However, these differences can be effectively realized by the statistical analysis shown in

Table 5. Descriptive statistical analysis of the soil erosion potential (ton/ha/year).

Year	Min	Max	Mean	St. Dev.
2020	0	4281	62	108
2019	0	2940	45	78
2018	0	1784	29	50

, which shows a descriptive statistical analysis of the soil erosion potential (ton/ha/year). The maximum soil erosion ranged from 1784 to 4281 ton/ha/year between 2018 and 2020.

Apparently, the densely urbanized areas of the Gaza municipality exhibit the least soil erosion potential, while the greatest soil erosion is likely to happen in the eastern parts of the governorate. This is ascribed to high soil erodibility, high slopes, and high cover management factors that characterize the eastern ridge of the terrain. The expanding urbanization at the eastern region of the Gaza municipality has had a clear effect on mitigation of soil erosion in the region compared to the undeveloped regions to the south of it. In addition, natural factors that include mild slopes and low-erodibility topsoil have had a clear effect on the undeveloped area in the middle region close to the shore. This area has maintained low soil erosion potential over the years.

The sum of soil erosion potential for the three basins was plotted against the measured accumulated sediments at these basins during different time increments, as shown in Figure 7. The results show that the soil erosion potential is greater than the measured magnitudes by approximately one order of magnitude. This indicates a low sediment delivery ratio for the study area. The graph also shows noteworthy behavior over the basin scale, where the estimated magnitudes correlated directly with the measured magnitudes. Apparently, this is a direct response to the changes in the rainfall erosivity which is the only changing factor for each set of data. However, this response, which is represented by the slope of the trend lines, varied substantially from one basin to another. These variations are ascribed to the differences in other components of the RUSLE.

3.3. Sediment Delivery Ratio and Model Calibration and Validation

The calibration process involved an extensive analysis that iterated k over the margin from 0.001 to 4, and IC₀ from −5 to 15. The analysis was conducted in increments of 0.1, and 0.5 for k, and IC₀, respectively. The resulting maps were analyzed for the k, and IC₀ combinations that achieved the solution for each individual basin and time increment. These possible solutions were plotted in Figure 8, and they were called the solution functions in the current study. Thus, each curve represents a solution function for a certain basin and time increment. Using any k and IC₀ from the values on the curve results in a zero residual in the estimated sediment export for the corresponding basin and time increment. Ideally, these curves should intersect at a single point that represents the global solution. However, the uncertainties involved in the parametrization process and the collected data deviate the solution functions from the global solution. Thus, the standard deviation of the curves was calculated with respect to IC₀, as shown in Figure 9. The least sum of residuals was achieved at IC₀ = 2.374 and k = 0.839. These two magnitudes were fixed for the subsequent analysis.

The SDR map was calculated using Equation (7) using the optimal values calculated for the calibration parameters, as shown in Figure 9. The SDR ranged between nearly zero and 0.4038. The map shows that the northern parts have the greatest magnitudes of the SDR. These areas are part of the watersheds that drain towards the sea, as illustrated in Figure 3c. The high SDR can be ascribed to the short flow length and the relatively high slopes in these areas. Most of the study areas have an SDR within the range from 0.0001 to 0.001.

The sediment export was calculated as the product of the two maps of the SDR and soil erosion potential (i.e., Figure 9 and Figure 10, respectively). Figure 11 shows the outcomes for the year 2019, as an example. As mentioned before, this map shows the quantities of eroded sediments that reach the drainage point.

The sum of sediment exports was calculated for the watersheds of the three basins (Asqula, Sadaqa, and Shiekh Radwan) and the results are plotted in Figure 12. The model provided better estimations using this approach. The slope of the regression line between the calculated and measured data is very close to unity, and R² is 0.977. The regression standard error, t statistic, and p-value were calculated as 0.0582, 17.1, 5.80 × 10⁻⁷, respectively. This demonstrates a good quality of data fitting to the linear regression described herein. Therefore, the model accuracy was verified, and the study proceeded for further analysis of the data.

4. Discussion

This study successfully developed a valid RUSLE model that helped to quantify and produce the corresponding maps for soil erosion potential, which are the first of their type for the Gaza governorate. The spatial data were carefully processed to maintain a high resolution that was sufficient to capture the impact of urban structures and allow for modeling of partially and poorly developed regions in the Gaza governorate. For this target, in this study, we considered parametrization of the RUSLE for urban land classes, which has not been a very common practice in previous studies. The RUSLE was initially developed for rural and arable lands; however, the demonstrated validity of the model suggests its applicability for similar urban areas. No previous studies can be found that have investigated or described the RUSLE parameters for the study area; therefore, the aspects encountered during the parametrization of the RUSLE are discussed herein.

The RUSLE parameters were selected by identifying the reported cases in the literature which best matched the field conditions in the Gaza governorate. Notwithstanding that Wischmeier and Smith [2] expressed caution against applying the USLE for watersheds that have mixed land use, the development of GIS and spatial analysis capabilities have made it possible to identify environmental and urban effects at the cell scale. Thus, generalization of the different land-use classes was avoided; and the subclasses were defined for the sake of enhancing the accuracy of results. However, uncertainties were faced in developing the RUSLE model for the Gaza Strip. Some of these uncertainties were associated with the compiled data and the accuracy of the employed approaches. It is feasible to recommend further investigation of C and P parameterization for urban land classes such as industrial, commercial, green landscapes, etc.

A scarcity of comprehensive research that has focused on the influence of spatial resolution on RUSLE models is evident. Consequently, it is strongly advised that forthcoming investigations delve into this facet, utilizing well-validated models grounded in extensive field measurements. Such investigations would contribute substantially to the understanding of this dimension.

The model was found to be sensitive to the R factor, which is the annual variable for the current study. However, the available infrastructure for meteorological observation in the Gaza Strip is outdated and has limited resolution. In fact, the major driving factor for using Renard and Freimund’s equation [3] was the lack of 30-minute rainfall records. This highlights the necessity of developing a network of meteorological stations in the Gaza Strip. Such a network would serve a wide range of environmental and hydrological purposes.

Notably, it should be highlighted that this study exclusively considered the yearly fluctuations of the parameter R, attributing this decision to the constrained temporal resolution inherent in the accessible data and the reliance on field measurements for model validation. Nonetheless, it is advisable that forthcoming investigations incorporate these seasonal fluctuations. This could be accomplished through the establishment of a specialized observation system that encompassed the areal sediment yield, as opposed to merely aggregating sediment accumulation at the endpoint of the drainage pathway.

The equation employed for calculating the LS factor was chosen as one of the commonly used equations for similar studies. However, other methods have been reported to perform better in different regions; therefore, further investigations to identify the best LS equation for the Gaza Strip would be useful in future studies.

The K factor was calculated based on the intersection of two maps of soil characteristics, which were interpolated from a set of field measurements. The current K-factor map served the model requirements very well; however, a more detailed map can be achieved by additional field surveying that intensifies the sampled locations.

The model performed well when non-erodible surfaces (such as buildings and paved streets) were assigned a zero magnitude to the K factor instead of the C factor. The effect was not noticed when calculating soil erosion potential. However, the cells which were assigned a zero C factor resulted in unrealistic magnitudes when the soil delivery ratio was calculated. Thus, they appeared as no data cells in the final outcomes. In addition, assigning zero magnitude to the K factor for the previously described conditions provided a better description of the physical behavior of the soil. The zero C factor represents a sink or a sand trap, which is not the case for non-erodible surfaces.

The C and P factors were identified based on guiding tables found in the literature. However, there is insufficient information in the literature that describes these two factors for urban areas. Some of the urbanization activities amend the surface in a manner that they retard or hinder sediment erosion even if they are not designed for that purpose. This includes the land-use classes which were identified as “mixed use”. They cover a significant proportion of the total area. C and P values were assigned to these areas based on the best knowledge of their characteristics. Therefore, further investigation of the best values that describe the C and P parameters for the land-use classes in urban areas is recommended for future studies. Currently, no specific practices are applied by farmers to mitigate topsoil erosion in the Gaza Strip. The best practices for soil conservation in arable lands in the Gaza Strip can be elucidated by further participatory studies that involve collecting farmers’ feedback and conducting training and shared pilot studies.

The current study avoided arbitrary assignment of the value of 1.0 to any C or P factor. Ignoring any of the factors’ effects resulted in erroneous spatial distribution of the results. Only the cover management factor of unpaved streets was assigned 0.7 for the C factor and 1.0 for the P factor, because they are characterized by disturbed topsoil and lack any measures for erosion control.

The model estimation for potential soil erosion was greater than the actual measurements, as shown in Figure 7. This is actually very common since the RUSLE is used to estimate the potential erosion of sediments and does not help the quantification of sediment delivery or deposition. The sediment delivery ratio is still governed by the network of streams and their connectivity, in addition to other factors such as soil cover and land slope. The graph shows a linear behavior of change for sediment erosion potential over the basin scale. The slopes of the trend lines reflect the sensitivity of the model to changes in the R factor which varied according to the basin. This behavior is ascribed to the variations in other parameters of the RUSLE. Thus, the steepest slope can be observed in the case of the Asqula basin, which is likely due to the relatively large agricultural and undeveloped surfaces that exist in the watershed. Also, the land slopes in the watershed are markedly higher than its magnitudes in the other two basins.

The equation formulated by Vigiak et al. [52] for the SDR, which utilizes the connectivity index developed by Borselli et al. [51], provided high flexibility for calibration. This helped with the optimization of the calibration parameters to attain the minimal residual and to develop a valid model for the study area, as shown in Figure 12.

Based on the behavior of the solution functions (shown in Figure 8), it is envisioned that better accuracy could be achieved by involving the C and P factors of mixed land-use classes in the calibration process. The current study employed an iterative process, which was adopted due to the complexity of the system. However, increasing the number of calibration parameters in the optimization process requires the development of a dedicated algorithm in order to minimize the time and resources required.

The current study revealed remarkable sediment generation behavior. The high risk of soil erosion was identified to be at the western slopes of the ridge that extends parallel to the eastern borders. This zone intersects with the endorheic catchment area that drains to the three stormwater basins, as shown in Figure 13. The map shows how these zones contribute to sediment accumulation in the three basins. This explains the potential sources of sediment that reach the stormwater infrastructure and highlights the potential zones where sediment control measures can be applied by the municipality.

Maintaining the durability and operability of drainage infrastructure is conditioned by the adoption of effective measures for mitigation of sediment erosion and delivery. Field practices could involve gully pots, gross pollutant traps, sediment ponds, grassed filter strips, and swales [24]. In addition, pavement or treatment of the topsoil on unpaved streets to control annual erosion would substantially mitigate sediment accumulation in the stormwater infrastructure, since unpaved streets are suspected to have the highest potential for soil erosion.

The impact of urbanization on the mitigation of sediment export can be seen in the map shown in Figure 13. The zone of high erosion risk at the east of Gaza City, which is highlighted in yellow and orange, is fringed by the green color that highlights the urban structures. For further characterization of this effect, the outcomes of the current study were put in the context of the studies provided by Wolman [55] and Russell et al. [24], which provided margins for sediment yield at different phases of development based on a set of field observations, as shown in Figure 14. The curves shown in the figure were acquired from Russell et al. [24], and the upper and lower colored lines represent the confidence levels. The average sediment yields for the urban and agricultural areas in the Gaza governorate were calculated as 166 and 215 ton/km²/year, respectively. Plotting these magnitudes on Figure 14 shows that the agricultural lands touch the 75% upper confidence level, while the sediment yield of the urban areas is below the median. On the one hand, the high yield of agricultural lands in the Gaza governorate could be ascribed to the lack of cover management and erosion control practices that has been applied to the model. On the other hand, the low yield of the urban areas comes from the fact that Russell et al. [24] observed sediment yield at the outlet natural streams and suggested that the increase in the observed yield in their study was likely caused by the erosion of natural soil along these streams. This is not the case for the Gaza governorate, since natural drainage systems do not exist in the urban areas. However, the urban yield of the study area is still considerably higher than the one reported by Wolman [55]. A reasonable cause for this is a lack of development and erosion control measures for the open spaces between the built-in areas. Moreover, the unpaved streets are suspected to significantly contribute to the overall yield of urban areas.

The compliance of the urban and agricultural yield in the Gaza governorate with the general trends with respect to the development status brings questions regarding the events that could cause shocks of sediment loads to the stormwater infrastructure. For example, the Gaza Strip underwent a war in 2014 that caused destruction of vast urban areas in the eastern parts of the Gaza governorate. Reconstruction started in 2015 and lasted for several years. There are areas of destruction in the basins of Asqula, and Sadaqa, as shown in Figure 15. Such large-scale destruction/construction events resulted in clearly witnessed erosion and transport of sediments from the affected regions. Unfortunately, apart from the citizens’ statements, no reported evidence exists on the impact of the destruction/reconstruction activities on sediment generation. However, looking at Figure 8, it can be observed that the solution function of Asqula basin for the year 2018 has an abnormal behavior, which indicates that the model is underestimating the actual accumulated soils. In fact, the same abnormality can be noticed in the solution function of Sadaqa 2017–2018, which can be observed to zoom in at the intersection point of the solution functions. Consequently, the model finally resulted in a negative residual of 27%, and 32% for Asqula and Sadaqa basins, respectively. This calls for additional investigations to assess the impacts of the events that have caused shock loading of sediment in order to identify the best post-disaster interventions for preventing damage to the stormwater infrastructure.

5. Conclusions

This study is the first to model and map the generation of waterborne sediments in the Gaza governorate. The model was calibrated using field measurements of the accumulated sediments at three major stormwater detention basins. The available data were acquired for different time increments. The model obtained good prediction accuracy with a coefficient of determination of 0.976 for the correlation between measured data and model estimations. The regression line has a slope that is close to unity.

Sediment generation potential in the Gaza governorate reached its maximum, which ranged from 1784 to 4281 ton/ha for 2018 and 2020, respectively, over the slope of the eastern ridge. The urbanized areas showed minimal soil erosion potential. A calibration model was used to calculate the optimal sediment delivery ratio. The average SDR was 0.00134. The estimations for actual sediment export ranged from 0 to 135.3 ton/ha, with an average of 0.0737 ton/ha.

This study showed that the eastern areas of Gaza City are highly prone to sediment transport and deposition along the flow path of streams and inside the drainage infrastructure. This is because urban watersheds intersect the zones of the highest soil erosion risk in the upstream, and the eroded sediments travel all along the watershed streams down to the final drainage point.

Generalization of land-use classes was avoided, and a wide range of agricultural and urban land-use classes were considered. The thorough investigation conducted in developing every map of the RUSLE parameters, and the approach that was employed for each one of them, provides good guidance for future studies in the Gaza Strip. Furthermore, the results of this study highlight a set of aspects for further investigation in the future.

The high resolution of the employed maps facilitated the identification of the urbanization effect on the sediment generation potential. The results showed abnormal behavior for some zones where the urban infrastructure and facilities were exposed to severe damage during the war in 2014. It is suggested that the high erosion quantities observed in these basins were the result of destruction rubble and the following reconstruction activities that lasted for several years.

This study elucidated the behavior of sediment erosion and delivery in the Gaza governorate, which has always been ambiguous. The information provided can be utilized both by the operational and planning units in the municipality. This study provided high resolution maps that can be utilized in the planning and design of measures for erosion control.

Agricultural lands were highlighted by significant erosion potential and relatively high SDR. However, the available data for the current study allowed limited capability to verify the behavior in these agricultural lands. Therefore, a study that investigates the erosion potential and its impact on the fertility of the topsoil in the Gaza Strip is highly recommended.

Finally, this study recommends further application of the developed model to cover the whole region of the Gaza Strip. Also, the municipalities are urged to involve sediment erosion in their planning and design activities. Sediment control measures should be considered to avoid the high risks of soil erosion. The current study could guide a detailed technical study on the direct impact of sediment erosion and accumulation on the operability of the stormwater management facilities and the costs of maintenance and operation. Future studies should consider the risks of accumulated sediments on public health and the environment. Also, the impact on the fertility of arable lands in the Gaza Strip and possible mitigation practices should be identified and elucidated.