Geographical Information System-Based Site Selection in North Kordofan, Sudan, Using In Situ Rainwater Harvesting Techniques

Abstract

The systematic identification of appropriate sites for different rainwater harvesting (RWH) structures may contribute to better success of crop production in such areas. One approach to improving crop yields in North Kordofan, Sudan, that is mostly adaptable to the changing climate is in-field water harvesting. The main objective of this study is to employ a geographical information system (GIS) in order to identify the most suitable sites for setting in situ water harvesting structures, aiming to address climate change in this area. A GIS-based model was developed to generate suitability maps for in situ RWH using multi-criteria evaluation. Five suitability criteria (soil texture, runoff depth, rainfall surplus, land cover, and slope) were identified; then, five suitability levels were set for each criterion (excellent, good, moderate, poor, and unsuitable). Weights were assigned to the criteria based on their relative importance for RWH using the analytical hierarchy process (AHP). Using QGIS 2.6.1 and ArcGIS 10.2.2 software, all criterion maps and suitability maps were prepared. The obtained suitability map for the entire region showed that 40% of the region area fell within the “good” class, representing 7419.18 km², whereas 26% of the area was “excellent”, occupying 4863.75 km². However, only 8.9% and 15.6% of the entire region’s area were “poor” and “unsuitable” for RWH, respectively. The suitability map of the delineated pilot areas selected according to the attained FAO data revealed that one location, Wad_Albaga, was found to be in an excellent position, covering an area of 787.811 km², which represents 42.94% of the total area. In contrast, the Algabal location had 6.4% of its area classified as poor and the remaining portion classified as excellent. According to the findings from the validated trial, Wad_Albaga is located in a good site covering 844 km², representing 46.04%, while Algabal is classified as a moderate site, covering 341 km² or 18.6% of the area. This study concluded that the validation of the existing trial closely matched the suitability map derived using FAO data. However, ground data from field experiments provided more accurate results compared to the FAO suitability map. This study also concluded that using GIS is a time-saving and effective tool for identifying suitable sites and discovering the most appropriate locations for rainwater harvesting (RWH).

1. Introduction

Sudan is facing severe water shortages (particularly during the dry season) due to the impact of climate change and rainfall variability [1,2,3]. Water scarcity in such arid regions has always been a serious issue worldwide. This is related to limited water resources that affect all life aspects, including the agricultural, economic, livestock, and social sectors. Agricultural sustainability and food production are most affected by water shortage and climate change, and this threatens food security in these regions. Therefore, using such techniques for an agricultural land suitability assessment is an important step in enhancing and developing land use planning and keeping agricultural areas effective [4].

The North Kordofan region of Sudan is located in a rain-fed agricultural area that receives inadequate and uneven amounts of rainfall throughout the year. This region is vulnerable to drought and subjected to severe conditions such as low rainfall and land degradation, thus reducing crop yield [5]. Land farming is considered a main source of livelihood in this region, behind animal herding, where 90% of the total cultivated area is used for traditional rain-fed agriculture. Moreover, the average cultivated area is very small, which is attributed to surface runoff water limitation and rainfall uncertainty.

There are many techniques and strategies adopted worldwide to reduce and mitigate the impacts of water scarcity resulting from climate change. Among these strategies is rainwater harvesting (RWH) for agricultural, human, and animal production purposes [6]. RWH techniques are effective strategies and solutions in arid zones where average annual rainfall ranges from 100 to 300 mm [7]. During the period of 1990s, some RWH techniques were promoted and were poorly adopted by a few farmers [8]. In situ RWH techniques have been used worldwide and are considered one of the most efficient management solutions and successful options for improving water retention in arid lands [9]. The effective implementation and success of rainwater harvesting (RWH) systems depend on several key factors. These include the location and design of RWH structures, land elevation and slope, orientation (aspect), rainfall levels, temperature, land use/land cover patterns, the geology of the site, soil characteristics, and engagement of local communities [10]. Some of the RWH techniques require gentle rainwater movements and surface runoff such as bund terraces and semi-circular bunds, while other systems prefer flat areas such as tied ridges [11].

At present, integrated research using modeling and geographic information systems (GISs) has shown an important role in identifying suitable sites for RWH. For example, Mahmoud and Alazba [12] used the GIS and decision support system methods to identify the potentiality of in situ rainwater harvesting sites, where the GIS tool was used to assess the available RWH structures in the target area. Moreover, computer technology in conjunction with GIS packages offers cost-effective and time-saving methods for identifying suitable sites for RWH [13]. In the North Kordofan region, GIS has not been widely used for delineating water harvesting techniques for agricultural activities. Rather, they have been used for other purposes such as rangeland surveys. The study conducted by Buraihi and Sharif [14] stated that in the Kirkuk area of Iraq, a GIS is a vital tool for identifying suitable RWH sites.

The study conducted by Ahmed et al. [15], which applied in situ water harvesting techniques as a micro catchment (terraces), stated that sorghum seed yield (productivity) was increased in clay soil (Gardud) with low fertility in the North Kordofan region. Farmers have been advised to use these techniques, particularly in regions characterized by low rainfall amounts. The experiment conducted by Ezzat et al. [16] compared the impact of different water harvesting techniques (contour ridges, runoff strips, and flat) and planting methods (reseeding and natural regeneration) on plant forage production, plant density, and vegetation cover in the North Kordofan region. These micro catchment techniques successfully improved the forage biomass production and vegetation cover in rangeland areas, but still, vast potential areas need to be identified using the GIS tool. Nyamadzawo et al. [8] stated that improved in situ RWH reduces crop moisture stress and can improve crop yield and sustain food security. Furthermore, they reported that, in the semi-arid regions of Zimbabwe, in situ RWH techniques were adopted by many farmers, which enabled their crops to survive during mid-season droughts. According to Mohammedien et al. [17], in situ water harvesting works best when the soil is deep enough and the water-holding capacity is large enough to retain moisture during dry periods.

Water scarcity in Sudan, in general, and North Kordofan, in particular, has been one of the biggest problems in the region, creating a profound issue for agro-pastoral communities in the area [18,19]. Therefore, it is very important to create a land suitability assessment for the rain farming sector in this area. Also, it is necessary to put emphasis on assessing the possibilities and restrictions of this land for RWH, particularly in arid and semi-arid areas. Our study focuses on the North Kordofan region, which is experiencing water shortage due to inadequate rainfall and variability. Despite the region’s low rainfall, rainwater usage is important for the local population. Thus, our research findings can improve agricultural sustainability in the region by providing managers and farmers with suitable information regarding appropriate RWH sites. In addition to that, adopting in situ RWH techniques can help farmers increase productivity by enhancing water retention, reducing crop moisture stress, and improving crop yields, resulting in good food security in the region.

This research aimed to identify suitable sites for in situ RWH for the whole Kordofan region and validate the existing in situ RWH structures in Sheikan province using the GIS tool to emphasize their suitability for agricultural activities considering several criteria such as soil texture, slope, rainfall, runoff depth, and land cover layers. This is the first attempt in the area to ensure that the identified sites are not only physically suitable but also environmentally and agriculturally viable.

2. Materials and Methods

2.1. Study Area Description

2.1.1. Geographical Location

The selected study area is located in North Kordofan State (Figure 1), which is one of the most vulnerable areas in the country concerning desertification processes. It is located in central Sudan between latitude 12°40′ N to 14°20′ N and longitude 28°10′ E to 31°40′. This region is a major contributor to Sudan’s economy, owning over 25 million livestock animals and 8.5 million acres of arable land. It accounts for roughly 30% of Sudan’s non-oil exports, including livestock, gum Arabic, hibiscus, groundnuts, and watermelon seeds [20].

2.1.2. Climate Conditions

North Kordofan is characterized by a variable climate, with arid conditions in the north and semi-arid conditions in the south. Annual rainfall fluctuates significantly, with averages of 196mm and 324mm, respectively. Rainfall is concentrated in the summer months (June to September), peaking in August. Despite the aridity, temperatures are relatively mild, averaging around 20 °C annually [20]. However, during the summer, daytime temperatures can reach extreme highs of around 45 °C.

2.1.3. Soil

The northern part of this region is dominated by sandy sheets and dunes and stabilized by vegetation, locally referred to as Qoz [20]. In the southern part, the soil transitions to an alluvial origin, characterized by a silty clay texture, and is locally known as Gardud.

2.1.4. Land Cover

The land cover pattern in the North Kordofan region can be divided into three classes: crop production, mainly groundnuts, millet, sorghum, and sesame; animal production, mainly goats, sheep, cattle, and camels; and forest production, with emphasis on the production of gum Arabic from Acacia Senegal. The low rainfall in this region results in sparse vegetation cover [20].

3. Data Collection and Processes

3.1. Criteria Selection and Assessment of the Suitability Level

To map potential sites for in situ rainwater harvesting (RWH) across the entire region and validate the suitability of existing semi-circular bunds and terraces, five key criteria were identified: soil texture, runoff depth, rainfall surplus, land cover, and slope. A nine-point continuous scale was employed to delineate suitable RWH sites. To identify suitable sites, a multi-criteria decision-making (MCDM) approach was adopted, considering factors such as soil texture, mean annual precipitation (50–300 mm/year), soil depth (50 cm), slope (4%), and vegetation cover [13]. The suitability map was reclassified into five comparable categories: excellent, good, moderate, poor, and unsuitable.

3.1.1. Soil Texture of the Region

Rainwater harvesting (RWH) potential is strongly impacted by surface runoff and infiltration rates, both of which are significantly influenced by soil texture [21]. Clay soils are typically thought to be the best for in situ RWH because of their high water retention capacity and low permeability [22]. On the other hand, sandy soils are less appropriate and show poor water retention. The suitability of medium-textured loamy soils for agricultural productivity was highlighted by Bulcock and Jewitt [23].

This study categorized soil textures into seven classes (clay, clay loam, loam, rock sand, sandy clay loam, and sandy loam) and assigned suitability rankings for in situ RWH on a scale of 1 to 5, with 5 representing the most suitable (

Table 1. Suitability rank for the applied parameters for both the region and Sheikan province.

Suitability Rank for Soil Texture
No	Texture Classes	Description	Ranking Classes
1	Clay		2
2	Clay loam		4
3	Loam		5
4	Rock		1
5	Sand		1
6	Sandy clay loam		5
7	Sandy loam		3
Suitability rank for runoff depth
	Runoff depth (mm)
1	21–25	Very deep	5
2	17–20	Deep	4
3	12–16	Moderately deep	3
4	8–11	Shallow	2
5	0–7	Very shallow	1
Suitability rank for rainfall surplus
	Rainfall surplus values (mm)
1	0–1	Very large deficit	1
2	2–22	Large deficit	2
3	23–40	Medium deficit	3
4	41–60	Small surplus	4
5	61–80	Large surplus	5
Suitability rank for land cover
	Land cover types
1	Bare land		3
2	Cultivated land		5
3	Forest and tree plantation		2
4	Shrub- and grasslands		4
5	Water bodies and artificial surfaces		(Restricted) 1
Suitability rank for slope of the study area
	Slope (%)
1	0–0.4	Flat	5
2	0.5–1	Slightly flat	4
3	2–7	Moderately sloping	3
4	8–15	Strongly sloping	2
5	16–32	Mountainous	1

). Sandy and rocky classes were deemed unsuitable for RWH. These soil texture data were rasterized using QGIS and served as the initial input layer for the AHP model. Sandy clay loam and loam were assigned the highest weights (4 and 5, respectively), while clay loam and sandy loam received intermediate weights (3 and 4) (Table 1). This classification highlights the importance of soil texture in determining the potential for in situ RWH and provides a foundation for subsequent spatial analysis and prioritization of suitable areas.

3.1.2. Runoff Depth of the Region

Runoff depth is a key factor in selecting suitable sites for various RWH structures. The curve number (CN) method, which consists of the hydrological soil group (HSG) (A, B, C, and D), which is defined by SCS soil scientists [24], land cover (LC), and antecedent moisture conditions (AMCs), which is an indicator of wetness and soil moisture availability prior to the storm, which has a significant effect on runoff volume, was applied in this study to estimate runoff depth. This parameter was subsequently integrated into the analytical hierarchy process (AHP) model as a key criterion for site selection. A hierarchical ranking system was established for runoff depth, with higher values assigned to areas with greater runoff potential. This approach enables the identification of optimal locations for implementing RWH structures based on their capacity to generate sufficient runoff.

SCS-CN Method

The Soil Conservation Service curve number (SCS-CN) model, developed by the USDA [25], is a widely used hydrological tool for estimating runoff, especially in rural and agricultural landscapes [26,27,28]. Its versatility, simplicity, and adaptability make the SCS-CN model one of the most commonly adopted methods for runoff estimation in various hydrological studies [29,30]. In this application, the model enables the quantification of spatial variability in runoff depth at a high-resolution, pixel-based scale. This allows for a precise analysis of runoff across different land areas.

An estimation of the direct runoff depth from storm rainfall is calculated using the following equation:

$${Q=}_{P-Ia+S}^{{(P-Ia)}^2}$$

(1)

where P = precipitation in millimeters (P ≥ Q); Q = runoff in millimeters; S = potential maximum retention in millimeters; and Ia = initial abstraction.

Curve Number CN

The curve number (CN) is an empirical parameter used to predict direct runoff. It ranges from 0 to 100, with lower values indicating low runoff potential and higher values indicating high runoff potential.

In this method, the curve number (CN) for each pixel is calculated by considering critical hydrological factors such as land use, slope gradient, soil texture, and antecedent soil moisture conditions. By integrating these variables, the model effectively captures the influence of both static and dynamic land attributes on runoff potential. The model utilizes the annual average rainfall depth in conjunction with the CN value for each pixel to compute localized runoff depth, resulting in a detailed runoff profile across the landscape. Watersheds characterized by high annual rainfall, fine-textured soils, exposed or sparsely vegetated land, and high antecedent moisture exhibit elevated CN values, which contribute to increased runoff levels [31]. This detailed approach enables accurate runoff estimation, which is essential for water resource management and flood prediction in various environmental contexts.

A conversion table created by Mishra [25] was used to adjust CN values for different antecedent moisture conditions. To compute the CN, which is weighted according to the area of each land use and soil type, we employ Equation (2) as follows:

$${CN=}_{\sum \mathit{A}\mathit{i}}^{\sum \mathit{A}\mathit{i}\ast \mathit{C}\mathit{N}\mathit{i}}$$

(2)

where CN is the composite curve number and Ai is the area of each curve number.

3.1.3. Rainfall Surplus of the Region

Rainfall surplus, defined as the excess precipitation that is not immediately utilized by ecosystems or human activities, is a fundamental parameter in identifying suitable sites for in situ rainwater harvesting (RWH). Regions with higher rainfall surplus values are more likely to support larger and more efficient RWH systems, as they possess a greater volume of water available for capture and storage [32,33].

To accurately assess the distribution of rainfall surplus across the study area, a rainfall surplus map was created. This involved calculating the difference between precipitation and evapotranspiration (P-ET) using meteorological data collected from 12 stations. The resulting data were then interpolated using a multilevel b-spline method in QGIS to create a continuous spatial representation of rainfall surplus. Consequently, the generated rainfall surplus map was classified into five distinct categories based on the magnitude of the surplus, ranging from very low to very high. This classification system facilitated the identification of areas with the highest potential for RWH implementation. The rainfall surplus parameter was integrated as a critical criterion within the analytical hierarchy process (AHP) model, significantly contributing to the overall evaluation of site suitability for in situ RWH structures. By assigning higher weights to areas with greater rainfall surplus, the AHP model was able to prioritize regions that offer optimal conditions for rainwater harvesting.

3.1.4. Land Cover (LC) of the Region

According to the literature, land cover composition significantly impacts the suitability of sites for in situ rainwater harvesting (RWH) techniques. Vegetated land covers, such as forests, grasslands, and agricultural areas, generally exhibit higher infiltration rates and lower runoff coefficients compared to impervious surfaces like built-up and industrial zones [34]. These characteristics directly affect the potential for rainwater harvesting and storage.

This study employed Landsat 8 imagery to create a detailed land cover (LC) map, which was classified into five distinct categories based on their hydrological properties and potential for RWH. Cultivated lands, shrub- and grasslands, and forests received higher ranks (5, 4, and 2, respectively) due to their capacity for water infiltration and retention. In contrast, bare land and water bodies were assigned lower ranks (3 and 1, respectively) and were deemed less suitable for in situ RWH. This LC classification served as a fundamental layer within the geographic information system (GIS) for subsequent spatial analysis and prioritization of suitable RWH sites. By incorporating land cover data into the decision-making process, it becomes possible to identify areas with optimal conditions for rainwater harvesting, particularly those characterized by high vegetation cover and low impervious surface coverage. This information is essential for optimizing the design and placement of RWH structures, maximizing water capture and utilization, and minimizing potential negative impacts on the environment.

3.1.5. Slope of the Region

Topographic slope is a fundamental parameter in assessing rainwater harvesting (RWH) potential. Flatter terrain, which is characterized by lower slope gradients, facilitates increased infiltration rates, reduces runoff, and enhances water retention capacity. As a result, it significantly improves the effectiveness of in situ RWH systems [35,36]. Conversely, steeper slopes promote rapid surface runoff, which limits the potential for capturing and storing rainwater.

To quantify the influence of slope on RWH suitability, this study utilized a 30-m-resolution Aster-Derived Digital Elevation Model (DEM) to create a slope map. The resulting slope data were classified into five distinct categories based on the FAO classification: 0–0.4% is flat; 0.5–1% is slightly flat; 2–7% is moderately sloping; 8–15 is strongly sloping; and 16–32% is mountainous. Each category was assigned to a suitability rank for in situ RWH (Table 1). According to the literature review, flat areas with a slope of 0–0.4% were assigned a higher suitability rank. The category of 0.5–1% received a weight of 4 followed by 2–7%, which was assigned a weight of 3. The lowest classes, 8–15 and 16–32%, were assigned weights of 2 and 1, respectively (Table 1).

4. Validation of the Existent In Situ RWH Structures

4.1. Analysis of the Criteria Maps of Sheikan Province

In order to validate the appropriate locations for both RWH structures, new thematic layers were created using the data collected from the sites where the two types of in situ RWH structures (semi-circular and terrace bunds) were implemented. The soil texture layer, land cover layer, and runoff depth layer were developed based on these newly collected data. Meanwhile, the rainfall surplus and slope layers were clipped from previously prepared maps of the entire region.

For the soil texture layer of the existent in situ RWH, soil texture data were gathered with the assistance of soil experts to identify the soil texture classes throughout the province, where some farmers are implementing these in situ rainwater structures. A GPS tool was utilized to pinpoint all identified soil texture locations. The soil texture map of Sheikan province was created using the identified soil types as shown in Figure 2a. The map was then reclassified into five categories: sand (rank 1), sandy clay (rank 4), sandy clay loam (rank 5), and sandy loam (rank 3). The Wad_albaga site was identified as the highest rank, classified as sandy clay loam, while the Algabal site was classified as sandy clay, holding the fourth rank which is also favorable for in situ RWH.

Runoff depth layer of the existent in situ RWH: This layer was created by combining the new soil and land cover layers using ArcGIS with a CN extension, as shown in Figure 2b. The map layer was then rasterized and reclassified into the following categories: 0–19 mm (rank 1), 20–24 mm (rank 2), 25–26 mm (rank 3), 27–29 mm (rank 4), and 30–33 mm (rank 5). Both locations fell into the highest category (rank 5) with a runoff depth of 30–33 mm.

Rainfall surplus layer of the existent in situ RWH: This layer was clipped from the region layer, which was interpolated using data from 12 rainfall stations over 13 years, as depicted in Figure 2c. The layer was reclassified into categories similar to the region, namely 0–9 mm, 10–20 mm, 21–31 mm, 32–41 mm, and 42–53 mm, with corresponding weights of 1, 2, 3, 4, and 5, respectively. The Wad_albaga location fell within a significant surplus, ranking as the highest (rank 5) with 42–53 mm, whereas the Algabal location ranked as 4, representing 32–41 mm.

Land cover layer of the existent in situ RWH: The new land cover data were derived from Landsat 8 imagery. This layer was digitized using the digitalization tools in QGIS software as shown in Figure 2d. The vector layer was also classified into five categories: cultivated land, bare land, forest and tree plantation, shrub- and grassland, and water bodies and artificial surfaces. Subsequently, the map was rasterized and reclassified as follows: cultivated land with a weight of 5, shrub- and grassland with a weight of 4, bare land with a weight of 3, forest and tree plantation with a weight of 2, and water bodies and artificial surfaces with a weight of 1, as they are restricted. As a result, both locations were assigned within the cultivated land category, which was regarded as the highest class in the context of in situ RWH.

Slope layer of the existing in situ RWH: This was also clipped from the slope of the region derived from Aster-DEM with a pixel of 30 as shown in Figure 2e. Filtering for undefined areas was performed with a majority filter. The slope map was reclassified into five classes as follows: 0–0.4% weight 5, 0.5–1% weight 4, 2–7% weight 3, 8–15% weight 2, and 16–32% weight 1. In situ RWH is not recommended for areas with slopes greater than 5% because runoff is distributed unevenly and large amounts of earthwork are required, which are often costly [37].

Therefore, flat areas with a slope of less than 2% were assigned a higher degree of suitability for in situ RWH. The two sites with a slope of less than 2% are the best sites for in situ RWH.

4.2. Weighted Linear Combination (WLC) Procedures

Weighted linear combination (WLC) is a powerful tool in geographic information systems (GISs) that can be used effectively for stormwater management planning and management. By combining multiple factors or criteria, WLC can help identify suitable locations for stormwater management facilities, evaluate the potential benefits of different techniques, and optimize resource allocation. The WLC process includes the following steps.

4.2.1. Reclassification

This is often used to simplify or change the interpretation of raster data by replacing a single value with a new value. Each factor was divided into classes and each class was assigned a weight from 1 to n (n = number of classes). For example, the class considered less suitable was assigned a weight of 1 [38].

4.2.2. Normalization

To ensure the comparability of the five reclassified thematic maps, the data were normalized. In this process, the values were scaled to a common range. Using the QGIS raster calculator, a standardized transformation was applied to the raster datasets by applying the following formula:

$$Data Normalization={\displaystyle \frac{Rastervalue-MinValue}{MaxValue-MinValue}}$$

(3)

Five maps were generated using these procedures and reported on a common scale of values, from 0 to 1 [38].

4.2.3. Weighted Linear Combination (WLC)

To assess the relative importance of the different factors influencing the suitability of sites for in situ RWH, two methodological approaches were considered: ranking and pairwise comparison. Ranking assigns ordinal values to factors based on their perceived importance, while pairwise comparison allows for a more nuanced assessment of the importance of factors. Given the subjective nature of ranking, the AHP method was used to perform pairwise comparisons of factors to provide a quantitative assessment of their relative weights [39]. This approach mitigated potential biases associated with direct ranking and provided a more solid basis for decision-making.

5. Analytical Hierarchy Process (AHP)

The analytical hierarchy process (AHP) is a multi-criteria decision-making method that prioritizes and evaluates factors in complex decisions by organizing them in a hierarchical structure from the main objective down to criteria, sub-criteria, and alternatives. At each level, pairwise comparisons are made using a nine-point scale [40] to evaluate the relative importance of each factor. For example, a factor rated “7” for higher importance has an inverse value of “1/7”, meaning that it is less important in comparison. Saaty [41] reports that the process involves the structuring factors being selected in a hierarchy ranging from the overall objective to the criteria, sub-criteria, and alternatives in successive levels. According to Saaty [42], the AHP includes the defining problem, setting an objective, creating a pairwise comparison matrix, and calculating the priority weights.

To integrate the most important criteria, GIS tools were used in addition to the AHP model to prioritize and rank each criterion based on its relative importance. Pairwise comparisons, as shown in (

Table 2. Pairwise comparison matrix for in situ RWH of both region and Sheikan.

	Texture	Runoff Depth	Rainfall Surplus	Land Cover	Slope
Texture	1	2	3	7	4
Runoff Depth	½	1	4	5	3
Rainfall Surplus	1/3	¼	1	4	3
Land Cover	1/7	1/5	¼	1	½
Slope	¼	1/3	1/3	2	1

), provided weights for each criterion that were guided by methods from studies in Ethiopia [37] and Saudi Arabia [43] to contextualize the relative importance of each factor. These assessments were based on literature reviews, field surveys, and input from experienced rainwater harvesting (RWH) practitioners. The results are presented in (Table 2) for the suitability ratings of the criteria weighting.

Assignment of the Criteria Weights

The criteria were weighted using pairwise ranking and the rank sum method. For the final calculation of the weights, the principal eigenvector of the pairwise comparison matrix was calculated to obtain the most appropriate weights. This calculation was performed using the WEIGHT module based on the analytical hierarchy process (AHP). The relative importance of each pairwise factor combination was assessed using a 9-point rating scale (Figure 3). The expected value method was used to calculate the weight, W_k, for criterion K based on Equation (4) as follows [43]:

$$W_k^{ev}={\sum }_{i=1}^{n+1-k}\left({\displaystyle \frac{1}{n(n+1-i)}}\right)$$

(4)

where n = the number of criteria and k = criterion.

The rank sum method was applied to calculate the weight, W_k, for criterion K following Equation (5):

$$K_k^{rs}={\displaystyle \frac{n+1-k}{{\sum }_{i=1}^n(n+1-i)}}$$

(5)

where n = the number of criteria and k = criterion.

The accuracy of the pairwise comparisons was assessed by calculating the consistency index (CI). This index determines the inconsistencies in the pairwise judgments and is a measure of departure from consistency based on the comparison matrices. It is expressed as follows:

$$CI={\displaystyle \frac{\mathsf{\lambda }-\mathrm{n}}{n-1}}$$

(6)

where CI is the consistency index, and λ is the average value of the consistency vector and n is the number of columns in the matrix [43].

To validate the results of the AHP, the consistency ratio of the matrix, which represents the degree of consistency achieved when comparing the criteria of the matrix rating, was randomly generated using Equation (7) below:

$$\mathit{C}\mathit{R}=\mathit{C}\mathit{I}/\mathit{R}\mathit{I}$$

(7)

where CI is the consistency index calculated by the model.

RI is the random index which is 1.12 (a constant value given by Saaty).

6. Geographic Information System (GIS)

Geographic information systems (GISs) provide a robust platform for integrating and analyzing spatial data to identify optimal locations for in situ rainwater harvesting (RWH). By overlaying multiple thematic layers representing rainfall surplus, land cover, soil texture, slope, and runoff depth [44], GISs enabled the identification of areas with the highest potential for RWH implementation.

In this study, both QGIS and ArcGIS software were used for spatial analysis and map generation. The weighted overlay process (WOP) was used to combine the different thematic layers, assessing each layer as a weight reflecting its relative importance in determining the suitability of RWH (Figure 4). The final suitability map was created by calculating a weighted average of the standardized thematic layers. The final RWH suitability map was created through a weighted linear combination (WLC) process. Prior to this integration, the continuous criteria were normalized to a common scale to ensure their comparability. Each of the five criteria was assigned a weighting reflecting its relative importance in determining overall RWH suitability. These weighted criteria were then summed to create a composite suitability index, which resulted in the final RWH suitability map using Formula (8) as follows:

$$\mathbf{S}=\sum \mathbf{W}\mathbf{i}·\mathbf{X}\mathbf{i}$$

(8)

where S = suitability;

wi = weight of factor i;

xi = criterion score of factor i.

Figure 4. Flow chart for identifying the suitable sites for in situ RWH.

Figure 4. Flow chart for identifying the suitable sites for in situ RWH.

7. Results

7.1. Analysis of the Criteria Maps of the Region

7.1.1. Soil Texture Map

Table 1 below shows the suitability ranks of soil texture. The classes for the study were categorized based on their suitability for specific applications, with a focus on the potential for in situ rainwater harvesting. The highest-ranked textures, sandy clay loam and loam, were weighted at a 5, indicating an optimal balance between permeability and water retention, making them ideal for water harvesting. These soil textures have favorable physical properties that allow for sufficient infiltration while retaining enough moisture for practical agricultural use.

As a result, clay loam and sandy loam were weighted 4 and 3, respectively. Clay loam, with a higher clay content, can retain more water than sandy loam soil but can cause problems with infiltration rates. Sandy loam is more permeable but retains less water, making it less suitable for water retention than clay loam, although it is still suitable for certain types of land management.

At the lower end of the scale, sand and rock received the lowest weighting of 1, reflecting their minimal water retention capacity and lack of suitability for in situ water harvesting. Clay, while having high water retention properties, was given a score of 2 as its low permeability may limit its usefulness in systems requiring efficient infiltration.

The map of the soil texture of the study area (Figure 5a) shows the presence of seven different soil classes. The northern region is dominated by clay loam, which makes up a large part of the northern landscape. In contrast, the central area is predominantly covered by sandy loam, which is the largest soil class in the study and extends in a wide strip from east to west. In the southern part of the region, sand is predominant, while the southernmost area has a more complex mix of minor clay loam, sandy clay loam, clay, and loam soils, reflecting a diverse soil composition. In particular, the Algabal and Wad_albaga pilot study areas are characterized by sandy clay loam, which completely covers Wad_albaga and partially covers Algabal. Moreover, the sand class covers about half of the area of Algabal, emphasizing the variable soil texture. This distribution provides important insights into the soil characteristics of the region, which could influence land use and agricultural potential.

7.1.2. Runoff Depth Map

The study area was divided into five different runoff depth classes (Table 1), each of which was assigned a weight based on its hydrological significance. The highest runoff depth, which is between 21 and 25 mm, is assigned a maximum weight of 5, indicating its critical importance for surface water flow and potential water resource management. Thereafter, runoff depths of 17–20 mm and 12–16 mm are weights of 4 and 3, respectively, reflecting moderate to high runoff conditions that still contribute significantly to surface water movement. The shallower runoff depths of 8–11 mm and 0–7 mm are classified with weights of 2 and 1, respectively, representing minimal runoff that may have less impact on water collection or erosion processes. These classifications provide a clear hierarchy of runoff potential across the study area and help to develop strategies for water management, land use planning, and soil conservation.

The runoff depth map of the study area (Figure 5b) shows considerable spatial variability in water flow patterns. The northern region is predominantly characterized by the lowest runoff depth (0–7 mm) and occupies almost 50% of the entire study area, making it the largest section in terms of area. In contrast, the central part of the study area shows strips with the highest runoff depth (21–25 mm), which extend in a continuous band from east to north, indicating a concentrated surface water flow in this zone. The near southern region has a mixture of 8–11 mm and 12–16 mm runoff depths, indicating moderate runoff in this part. Further south, the southern region is characterized by a more diverse runoff pattern, with patches of deep runoff (21–25 mm) and moderate runoff (17–20 mm), but occupying a relatively small portion of the overall landscape. This spatial distribution highlights the need for targeted water management strategies, especially in areas with high runoff potential.

7.1.3. Rainfall Surplus Map

The study area is classified into five distinct classes of rainfall surplus, ranging from a large surplus to a large deficit. Each class is assigned a weight based on its relevance for the selection of suitable sites for in situ rainwater harvesting (RWH) systems. The large surplus category, with a precipitation surplus between 61 and 80 mm, is assigned the highest weight of 5, indicating its critical importance for water availability and potential RWH capacity. This is followed by the low surplus range of 41–60 mm, which is weighted as 4, reflecting a significant surplus of water suitable for water harvesting. The moderate surplus class with values between 22 and 40 mm is weighted at a weight of 3, indicating moderate potential for water harvesting.

Lower rainfall surplus levels of 2–21 mm and 0–1 mm, representing a small deficit and a large deficit, respectively, are weighted 2 and 1. These lower ranks reflect the limited water availability in these areas, making them less suitable for in situ RWH structures. This classification scheme enables the prioritization of areas with the highest rainfall surplus for effective water harvesting and management strategies and ensures that the regions with the greatest potential for rainwater harvesting and storage are optimally utilized.

The rainfall surplus map (Figure 5c) in the study area shows clear spatial patterns of water availability. The northwestern region is predominantly characterized by a low rainfall surplus (2–21 mm) and thus occupies the largest position in the study area. In contrast, vertical strips with relatively high rainfall surplus (41–60 mm and 61–80 mm) are concentrated in the eastern part of the study area, indicating localized zones with greater water availability. However, the central and western regions have the lowest rainfall surplus (0–1 mm), indicating significant water scarcity in these areas. In the southern region, near the south, the precipitation surplus is mostly low (2–21 mm), indicating similar conditions to the northwest. These patterns indicate areas where rainwater harvesting (RWH) may be most effective, particularly in the east where the surplus is highest.

7.1.4. Land Cover Map

The land cover of the study area is categorized into five different classes (Table 1), reflecting the varied landscape and its potential use for environmental and land management practices. These classes include bare land, cultivated land, forests and tree plantations, shrub- and grasslands, and water bodies and artificial surfaces. Based on extensive research and reclassification criteria from cited studies, land cover types were ranked according to their ecological and functional importance. Cultivated land was given the highest rank of 5, underlining its crucial value for agricultural productivity and human use. Following this, shrub- and grasslands were ranked 4 to recognize their role in promoting biodiversity and providing grazing.

Bare land was rated 3. There are areas with limited vegetation cover that are often prone to erosion but could offer potential for future development or restoration projects. Forests and tree plantations were ranked 2, reflecting their importance for carbon sequestration biodiversity, and ecosystem services, although they may be less important for certain land uses such as agriculture. Finally, water bodies and artificial surfaces were given a limited rank of 1, indicating a low suitability for land use applications due to their limited availability for cultivation or modification. This ranking system provides a structured approach for land use planning and management, focusing on the most valuable and usable land types.

The land cover map of the study area shows distinct spatial patterns in land use (Figure 6a). The northern parts of the study area are completely covered with bare soils, indicating areas with minimal vegetation or development. The central region is dominated by a mixture of cultivated land and shrub- and grasslands, which make up most of the study area. This mixture indicates a combination of agricultural activity and natural vegetation that likely supports both farming and grazing. In addition, there are scattered forests and tree plantations which are mainly found in the western and southern parts of the study area, although these areas make up a smaller proportion of the overall landscape. This distribution reflects a balance between natural and managed land cover types in the region.

7.1.5. Slope Map

The slope of the study area was classified into five categories according to the FAO classification system, ranging from flat to mountainous terrain. The classification includes flat areas (0–0.4%), slightly flat (0.5–1%), moderately sloping (2–7%), steeply sloping (8–15%), and mountainous areas (16–32%) (Table 1). These slope classes were categorized based on their suitability for different land uses, with a focus on agricultural and building potential.

The flat areas (0–0.4%) were considered the most suitable and were therefore given the highest suitability rank. This flat terrain is ideal for agriculture, infrastructure development, and other land uses that require minimal slope. The slightly flat areas (0.5–1%) were weighted 4, reflecting their moderate suitability as they offer favorable conditions for land use, but they still have a slight slope. The moderately sloping areas (2–7%) were weighted at 3, indicating less suitability but still suitable for certain types of land management, such as terracing or selective development.

The steeply sloping (8–15%) and mountainous areas (16–32%) were given the lowest weighting of 2 and 1, respectively. These steep slopes pose a challenge for cultivation and infrastructure due to the risk of erosion, difficulty to access, and potential instability and are therefore less suitable for most forms of development. This slope classification and ranking system enables targeted planning for land use, optimizing the flat and slightly sloping areas for agriculture and development while considering the limitations of steeper terrains.

The map of the slope of the land surface of the study area (Figure 6b) shows that most of the terrain has a gentle slope, with gradients between 0.5 and 1 degree dominating most of the landscape. There are some isolated terrain elevations in the northern parts of the area, but these are minimal and do not significantly affect the overall flatness of the region. There are no abrupt slopes or steepness (2–7 degrees) throughout the study area, with the exception of a few small patches in the central region and the extreme south, where slopes of 8–15 degrees are observed. These areas with steeper slopes are limited and do not significantly alter the overall gently sloping character of the study area.

7.1.6. AHP Model Outputs

In the pairwise comparison matrix for in situ water harvesting systems (Table 2), criteria such as texture, runoff depth, rainfall surplus, land cover, and slope are compared to determine their relative importance using the AHP model. Each value in the matrix represents the relative importance of one criterion compared to another.

Texture is considered the most influential criterion in this matrix. It has the highest comparative values across multiple criteria, indicating its great importance in determining suitable sites for in situ water harvesting. For example, texture is rated “7” compared to land cover and “4” compared to slope, indicating that texture is considered much more important than these factors.

Runoff depth is the next most important criterion, with high scores compared to other criteria, such as a “4” compared to rainfall surplus and a “5” compared to land cover. This indicates that areas with a higher runoff depth are favored due to their potential to effectively capture and store water.

Rainfall surplus has moderate importance, with values such as “4” compared to land cover and “3” compared to slope. This means that although rainfall surplus is important, it has a lower priority compared to texture and runoff depth in influencing site suitability.

Land cover is given less importance compared to other criteria, which is reflected in the low comparative values (e.g., “1/7” compared to texture and “1/5” compared to runoff depth). This indicates that land cover is less critical when identifying suitable in situ structures.

Slope has a moderate influence, especially when compared to land cover (rated “2”), but is less important than texture and runoff depth. A moderate slope is beneficial, but not as crucial as the other higher-rated factors.

The results of the AHP model from the pairwise comparison presented in the matrix table were determined by a systematic evaluation of the criteria based on a comprehensive literature review. Each criterion was evaluated based on its relative importance for in situ rainwater harvesting, with relative weights and suitability ranks assigned according to multiple studies (Table 2). Notably, the soil texture criterion was 7 times more important than the land cover criterion, meaning that land cover is 1/7 as important as soil texture in determining suitability for RWH. This significant weighting difference emphasizes the crucial role that soil properties play in water retention and infiltration potential, compared to land cover, which is less influential in this context.

The consistency ratio (CR) of the comparative matrix in this study was calculated using Equations (4) and (5). The calculated ratio was 0.026, which is less than 0.1 and indicates that the judgments are trustworthy and appropriate.

This is well within the acceptable threshold, as CR values less than or equal to 0.1 indicate a reliable level of consistency in the decision-making process [40]. This low CR value confirms that the matrix scores were not randomly generated but reflect a logical and consistent approach to weighting the criteria based on their importance to RWH suitability.

The results of the AHP model indicate that a large portion of the study area is well suited for in situ rainwater harvesting (RWH), as shown in

Table 3. Suitable areas of the entire region.

Suitability Levels	Area (km2)	Percentage (%)
Excellent	4863.75	26.25
Good	7419.183	40.05
Moderate	1705.481	9.2
Poor	1645.227	8.9
Unsuitable	2888.126	15.6
Total	18,521.767	100.00

and Figure 7. Approximately 40% of the total area falls into the good suitability class, which covers an area of 7419.183 km², while 26% of the area, or 4863.75 km², falls into the excellent suitability class. These classifications make it clear that two-thirds of the region is favorable for RWH, with a larger proportion falling into the good class, which implies a high potential for efficient water harvesting.

The dominance of the good class reflects the integration of high-ranking criteria, such as sandy clay loam soil texture, rainfall surplus (>60 mm), runoff depth (>20 mm), cultivated land cover, and a moderate slope of 2%. These factors are crucial for the suitability of RWH as together they improve the water retention and infiltration capacity. The sandy clay loam texture supports both water infiltration and water retention, while the rainfall surplus and runoff depth indicate a high availability of water for collection. In addition, the presence of cultivated land reinforces the area’s suitability for water management as it reflects agricultural activities where RWH structures are most beneficial. The gentle slope of 2% facilitates water movement without causing excessive erosion, making it ideal for in situ harvesting.

Overall, the AHP model is successful in identifying areas or regions with the highest potential for RWH. The combined influence of favorable environmental and land use conditions is reflected in the majority of the study area being categorized as good or exceptional. Targeting these high-potential zones for RWH implementation is crucial for optimizing the efficacy and efficiency of water resource management plans, as this distribution shows.

According to the results of the AHP model, the Wad_albaga location, which occupies 46.04% of the entire area, is classified as having good suitability, while the Algabal location, which occupies 18.6% of the area, is classified as having moderate suitability (

Table 4. Suitable sites of Sheikan province.

Suitability Levels	Area (km2)	Percentage (%)
Excellent	198.182	10.81
Good	844	46.04
Moderate	341	18.60
Poor	328.998	17.95
Unsuitable	121	6.60
Total	1833.18	100.00

). These findings demonstrate the Wad_albaga site’s feasibility for in situ rainwater harvesting (RWH), primarily because important factors including soil texture, land cover, and runoff depth were well integrated and carefully gathered from field observations.

The quality and dependability of the newly gathered data, which represent actual field conditions, are closely related to Wad_albaga’s higher appropriateness grade. A more accurate assessment of the site’s ability to retain and collect water was made possible by the field assessment of variables such as soil type, land cover, and runoff depth. Wad_albaga’s better adaptability and potential for effective RWH methods were influenced by its soils, particularly its sandy clay loam, cultivated land cover, and ideal runoff depth.

However, even if Algabal is still viable for RWH, it ranks lower, in the moderately good class, which can be attributed to less favorable conditions in terms of these same criteria. The region’s potential is reflected in its moderate ranking, which also identifies areas that could require improvement, perhaps through further holistic land management practices. The accuracy of the assessment was significantly increased by the direct field data collection and integration into the AHP model, which offered a practical and useful evaluation of both locations for in situ RWH practices.

The AHP model results provide valuable insights into the spatial variability of RWH suitability, as illustrated by the analysis of Wad_albaga and Algabal sites (Figure 8). Wad_albaga is entirely classified as excellent, indicating favorable conditions for siting RWH structures. Its superior environmental and land characteristics make it an optimal location for efficient water collection and management.

Algabal, on the other hand, exhibits a more diverse suitability profile. Half of the area is rated “excellent” while the other half is poor. This variation highlights the site’s heterogeneous environmental features, with some areas highly conducive to RWH, and those less so are likely due to factors such as soil, runoff, or land cover. Overall, the study reveals that 42.94% of the total area is classified as excellent, offering significant potential for effective RWH implementation. An additional 5.90% is categorized as good, further expanding the feasible zones for RWH. However, 24.5% of the area is moderately suitable, suggesting a need for careful planning and potential adjustments to water management practices. On the less favorable end, 20.25% of the area was deemed “poor”, and 6.4% was classified as “unsuitable”, indicating the limited potential for efficient RWH systems (

Table 5. Suitable areas of the study area developed using FAO data.

Suitability Levels	Area (km2)	Percentage (%)
Excellent	787.811	42.94
Good	42	5.90
Moderate	649.4	24.51
Poor	371.2	20.25
Unsuitable	117	6.41
Total	1833.18	100.00

). These findings provide crucial information for spatial planning and targeted RWH interventions, enabling the optimization of land use in the most suitable areas while acknowledging the constraints of less favorable sites.

A comparative analysis of the two suitability maps generated for the study area, using FAO data (Figure 8b) and the newly collected field data (Figure 8a), provides valuable insights into the precision of in situ rainwater harvesting (RWH) site selection. Initially, the FAO-derived map indicated that Algabal had an equal split between “excellent” and “poor” suitability areas, while Wad_albaga was consistently classified as “excellent.” However, when incorporating the new field data, the results shifted: Algabal was reclassified as “moderate” suitability, and Wad_albaga was downgraded to “good”.

Despite this classification, Wad_albaga remains the superior site for RWH compared to Algabal, as shown in

Table 6. Comparison between the two suitability sites of both structures in Sheikan province.

Locations	Suitability Map of FAO Data	Suitability Map of New Data
Algabal	1/2 Excellent 1/2 Poor	Moderate
Wad_albaga	Excellent	Good

. This comparison highlights a key point, while both maps show general agreement on overall RWH suitability. The newly collected field data, which includes more precise measurements of soil texture, land cover, runoff depth, and other crucial factors, provides a more accurate and realistic assessment of site potential. The new data led to more refined classifications that better reflect the actual ground conditions. For example, the reclassification of Algabal from a mix of “excellent” and “poor” to “moderate” suitability underscores the importance of field data in correcting over- or underestimations in suitability that may arise from generalized datasets like the FAO data.

8. Discussion

8.1. Criteria Used to Delineate Suitable Sites for In Situ RWH

This study considered several key environmental factors, including soil texture, runoff depth, rainfall surplus, land cover, and slope. These findings revealed that most of the study area is dominated by sand and sandy loam soils. Most of the published literature revealed that loamy to sandy loam soils are widely recognized as suitable for in situ RWH. This agrees with the observations of [21,31,37], who emphasized that soils having small and medium pores, loamy sand to sandy loam, and sandy clay loam are particularly favorable for RWH systems due to their balance of infiltration and water retention capacity.

Regarding runoff depth, the study’s findings align with Rajbanshi [45] and Gebremedhn [31], who highlighted the importance of estimating runoff depth for determining RWH site suitability. Areas with greater runoff depth, especially in flatter regions, offer more opportunities to control and trap harvestable water using structures like bund terraces, semi-circular bunds, and U/V shapes, as supported by [21,35,46]. Slowing down and storing runoff is crucial for enhancing water harvesting effectiveness, especially in arid regions where water scarcity is a significant challenge. Mahmoud et al. [47] further confirmed the significant role of runoff depth in shaping suitability maps for RWH, emphasizing its importance in sustainable water management practices.

The rainfall surplus layer emerged as the most influential factor in developing the RWH suitability map, aligning with findings from [21,37,48]. Areas with significant rainfall surplus were ranked highest in suitability, as the abundance of surplus water ensures an ample supply for harvesting and soil retention. This study’s results echo [43], which found that areas with higher rainfall surplus support more complex and efficient RWH systems, improving the overall water capture efficiency. As noted by [21], such areas also have greater potential for soil moisture retention, making them ideal for water harvesting techniques.

Land cover criteria is one of the crucial factors influencing geo-hydrological processes like runoff control, infiltration, and evapotranspiration reduction [21]. Similar to [31], this study ranked the land cover (LC) layer as moderate or less important, as water harvesting techniques can be implemented within various land cover types with minimal investment. Consistent with [43], the findings confirmed that land cover is a vital criterion in the AHP model, as areas with cultivated land and vegetation enhance the potential for RWH systems.

Finally, the slope map revealed that the majority of the study area is flat, the most suitable terrain for in situ RWH techniques. Studies by [37] agree that flat or gently sloping areas have a greater capacity to retain water within the soil profile compared to steeper areas, where runoff is lost more rapidly. This aligns with [21,31,35], who stated that gentle slopes (0–2%) are highly suitable for rainwater harvesting development. They allow runoff to slow down and infiltrate the soil more effectively. Additionally, ref. [46] concluded that the most successful and sustainable RWH systems are typically found in areas with slopes of less than 2%, further emphasizing the suitability of flat terrain for water harvesting efforts.

8.2. Identifying Optimal Locations for In Situ Rainwater Harvesting (RWH)

Developing suitability maps for RWH structures in the study area is a crucial step towards enhancing agricultural sustainability, especially in regions with limited water resources. By systematically identifying the most suitable locations, this study aims to support the implementation of effective RWH practices and improve crop production. The generated suitability maps revealed that a significant position of the study area, approximately 40%, falls into the “good” category, while 26% falls into the “excellent” category. These results align with previous studies [31,37,43,47], which demonstrated that areas with higher rankings in soil texture, rainfall surplus, runoff depth/coefficient, land cover/use, and slope are generally more suitable for in situ RWH. Given the challenges posed by climate change and extended droughts in the region, implementing in situ RWH practices in areas with “good” and “excellent” suitability ratings can significantly benefit crop productivity. Even areas with “moderate” suitability may be considered for RWH interventions, especially if the potential benefits outweigh the challenges.

To validate the generated suitability maps, a comparison was conducted between the results obtained using FAO data and newly collected field data. While the FAO-based maps suggested a mix of excellent and poor suitability for the Algabal and Wad_albaga locations, the field data revealed that Wad_albaga consistently demonstrated higher suitability for in situ RWH.

The findings of this study highlight the importance of using a combination of geographic information systems and field observations to accurately assess the suitability of sites for in situ RWH. The generated suitability maps provide valuable guidance for policymakers, farmers, and other stakeholders in identifying and prioritizing areas for RWH interventions. By implementing RWH practices in these suitable locations, it is possible to enhance agricultural productivity, reduce water scarcity, and improve resilience to climate change. This study’s findings are similar to those of [31,37,43,47] who used similar thematic layers including runoff depth, land cover/land use, slope, and soil texture as inputs to develop in situ RWH techniques in Pakistan, Ethiopia, and Saudi Arabia using GIS and RS tools. By employing these criteria, this study successfully identified the most suitable sites in the area for rainwater harvesting through the installation of various techniques.

9. Conclusions

This study demonstrates the effectiveness of GISs and AHPs in identifying optimal locations for in situ rainwater harvesting (RWH) in arid and semi-arid regions. By integrating factors such as soil texture, rainfall surplus, runoff depth, land cover, and slope, the study developed a comprehensive suitability map that can inform sustainable water management decisions. The results indicate that a significant portion of the study area is well suited for RWH implementation, with 40% falling into the “good” category and 26% in the “excellent” category. This suggests that RWH can be a valuable tool for enhancing agricultural productivity and addressing water scarcity in the region. Validation of the suitability map using both existing RWH structures and field data further confirmed the accuracy and reliability of the model. The findings highlight the importance of considering multiple factors and utilizing advanced spatial analysis techniques to identify the most promising sites for RWH interventions. The generated suitability map can be used by water resource managers and planners to prioritize RWH initiatives, allocate resources effectively, and develop targeted strategies for improving water security in the region. By implementing RWH in suitable areas, it is possible to increase water availability, enhance agricultural productivity, and mitigate the impacts of climate change.