Integrating GIS and AHP for Drought Sensitivity Assessment in the Middle Moulouya and Guercif Basins, Morocco

doi:10.21203/rs.3.rs-4966568/v1

Drought sensitivity assessments play a critical role in understanding and mitigating risks associated with this phenomenon. This study presents a novel approach for comprehensive drought vulnerability mapping in the Middle Moulouya and Guercif Basins, Morocco. We leverage a multi-criteria framework integrating Geographic Information Systems (GIS) and the Analytic Hierarchy Process (AHP), focusing on three key dimensions: climatic sensitivity, soil sensitivity, and socioeconomic sensitivity.

Beyond traditional factors: We expanded the analysis beyond basic climatic data by incorporating future climate projections to estimate drought risk under different scenarios, thereby allowing for more future-proof vulnerability assessments. Soil health matter: Soil moisture content and infiltration capacity were incorporated into the soil sensitivity assessment, provide a more comprehensive picture of soil susceptibility to drought.

Socioeconomic considerations: socioeconomic sensitivity analysis goes beyond population density to consider factors such as water resource dependence, agricultural practices, and livelihood diversification. This holistic approach provides a deeper understanding of societal vulnerability to droughts.

AHP for informed weighting: AHP facilitates the incorporation of expert knowledge and data analysis by assigning weights to each sub-criterion in three dimensions. This ensured a robust weighting scheme that reflected the relative importance of different factors in drought vulnerability.

The resulting vulnerability map not only delineated areas with varying degrees of drought vulnerability (normal, mild, moderate, severe, and extreme) but also highlighted the spatial distribution of these vulnerabilities. Notably, the analysis revealed that 38.88% and 28.80% of the Middle Moulouya and Guercif regions fall under the severe and extreme vulnerability categories, respectively. These areas, particularly the northern, central, and southwestern regions, warrant immediate attention in the development of targeted drought mitigation and adaptation strategies.

Drought Sensitivity

Analytic Hierarchy Process AHP

Geographic Information Systems GIS

vulnerability assessment

Climate change

Soil

Multi-criteria analysis

Middle Moulouya

Morocco, like many other developing countries, suffers from the deterioration of non-renewable or feebly renewable natural resources such as forests, soil, and water due to poverty, population growth, and climatic changes. In recent years, the latter have caused a significant decrease in the annual precipitation rate, changes in the seasonal system, and an increase in temperature rates. Wrongful human activities and excessive exploitation of natural resources have made arid regions more fragile [1], and ecosystems threatened by desertification and irreversible degradation have deteriorated further in recent years.

Drought is often referred to as the "silent killer" among hazards owing to its unpredictable, ever-changing, and dynamic spatial and temporal occurrence. Despite being categorized as a discrete recurrent hazard [2], [3], the risks associated with droughts are undeniable. Severe drought has the potential to inflict significant harm on both people and the environment, because it can disrupt natural cycles, particularly hydrological processes [4], leading to environmental disturbances and even disasters. Therefore, it is necessary to adopt appropriate strategies to predict the various risks that destroy or severely impact natural resources in fragile arid environments and to assess and monitor the vulnerability of ecosystems and their ability to resist natural hazards such as desertification and drought.

Managing and preserving the environment in arid and semi-arid areas is generally based on accurate scientific studies conducted by national interests and institutions, as well as scientific research centers within universities.

To study drought in the Middle Moulouya and Guercif Basins and its impact on various natural resources and ecosystems, it is essential to focus on geographic information systems and remote sensing. It is important to evaluate and monitor drought situations and create maps of areas prone to drought owing to increasing human pressure and climatic changes.

Effective drought hazard mapping accounts for the selection and zoning of all the factors that influence drought hazards [5]. Spatial analysis and satellite information are important for drought hazard mapping because attribute analysis requires both spatial and temporal components. Satellite information provides precise, timely, and continuous attributes for large areas.

Monitoring and assessing drought through remote sensing techniques depends on the causes and effects of the drought. In this regard, satellite images are used for early warnings during the early stages of drought. If an anomaly occurs, it suggests a particular weather series [6]. These images also provide a database through which past drought periods can be interpreted and combined with other data to obtain more accurate and realistic maps, which in turn can aid drought cartography and provide a database for relief operations.

Spatial analysis supports data integration and analysis [7]. AHP is a dimensional theory based on pairwise comparisons that relies on the judgment of experts and previous studies or literature reviews to obtain priority scales. Additionally, it is a powerful approach for dealing with multi-criteria decision-making [8], [9]. The AHP approach is more effective when combined with GIS [10].

The contribution of this study is to propose a methodology based on the models presented above and to develop a drought sensitivity map for the Middle Moulouya and Guercif Basins. The development of this map is based on a weighted analytical method which considers that the selected criteria applied to a natural phenomenon remain unequal. Accordingly, a hierarchy was introduced between the variables by assigning a weight that corresponded to the importance of each criterion compared to the other criteria [11].

2.1. Study Area

This study focuses on the Middle Moulouya and Guercif Basins situated in northeastern Morocco, which encompass an area of approximately 19,819 km². The basin exhibits significant elevation variability, ranging from 255 meters above sea level (masl) in the lowlands to 3,335 masl in the surrounding highlands (Fig. 1). The basin experiences a precipitation gradient and is classified as a semiarid region. The Western Middle Atlas and Eastern High Atlas in the south receive over 700 mm of annual precipitation on average, which decreases towards the Moulouya depression, reaching approximately 260 mm annually during the Outat El Haj station season [12]. Temperatures fluctuate between high temperatures exceeding 40°C in the summer and low temperatures below 0°C in the winter.

The natural vegetation cover is dominated by steppe ecosystems characterized by low-lying shrubs and herbaceous allies. However, these ecosystems are subject to continuous degradation due to successive drought years and anthropogenic pressures. Unsustainable practices such as intensive grazing by local populations have exacerbated vegetation loss and soil erosion. Furthermore, the geological composition of the basin, characterized by friable rocks and weakly developed soils, renders the landscape particularly vulnerable to wind erosion. This vulnerability is manifested by aeolian sand accumulation, particularly during dry months when strong winds are present. The region is highly susceptible to drought, characterized by delayed or erratic precipitation patterns during the rainy season and water scarcity during the dry season. These factors collectively contribute to the fragile environmental state of the basin.

2.2. Data Preparation

This study leverages a multi-criteria approach for drought hazard mapping in the Middle Moulouya and Guercif Basins, Morocco. Geographic Information System (GIS) techniques were employed to integrate various data layers collected from diverse sources (Table 1). The selection of these criteria layers was guided by a comprehensive review of the drought literature and data availability [13].

The analysis of drought literature and data availability formed a fundamental assumption underlying this methodology: three basic factors (climate, soil, and socioeconomic sensitivity) influence drought hazards. Each of these basic criteria is based on three sub-criteria:

The climatic criterion includes three sub-criteria: PET, Aridity Index, and SPI.
The soil sensitivity criterion also includes three sub-criteria: water holding capacity, soil texture, and land use.

• The socioeconomic criteria include human pressure, animal pressure, and intensity of agricultural activities.

Figure 1 Study area is located in the northeast of the kingdom of Morocco

Table 1

Types and Sources for Developing Drought Hazard Map in This Research
Basic Criteria	Sub-Criteria	Data Source
climate sensitivity	SPI Index	https://openlandmap.org
	aridity index Ai
	Evaporation-transpiration PET
soil sensitivity	Water Holding Capacity of Soils	https://openlandmap.org
	Soil Organic Matter OM	iSDA - iSDAsoil (isda-africa.com)
	land uses	https://livingatlas.arcgis.com/
Socio-economic sensitivity	population	HCP
	cattle heads	Agricultural Consultation Center of Outat El Haj and Guercif
	Cultivated area (ha)

2.3. Methods

Drought severity arises from the complex interplay of factors that affect societies through agricultural losses, wind erosion exacerbation, and desertification. This study addresses this challenge by developing an integrated modeling approach utilizing the Analytic Hierarchy Process (AHP) and Geographic Information Systems (GIS) [5], [8], [14]. This framework aims to guide the development of mitigation strategies to minimize drought impacts in the Middle Moulouya and Guercif Basins.

This methodology employs a multi-criteria approach that integrates nine sub-criteria thematically linked to climate, soil, and socioeconomic sensitivity. ArcGIS's Reclassify tool was used to standardize the sub-criteria values on a scale ranging from 1 (least suitable) to 7 (most suitable).

These sub-criteria layers were built within ArcGIS and IDRISI using raster grids and vector shapefiles, and then aggregated to form three main criteria layers (Climate, Soil, and Socioeconomic). These main criteria layers were incorporated into the final assessment matrix to generate a drought hazard map (Fig. 2). This map serves as a valuable tool for identifying vulnerable areas and for informing targeted mitigation strategies.

2.3.1. Analytic Hierarchy Process (AHP)

The Analytic Hierarchy Process (AHP) is a powerful approach for solving multi-criteria analysis problems [15]. To weigh the sub-criteria, we used the pairwise comparison approach developed by [16], [17] to support decision-making.

The AHP methodology employs pairwise comparisons to establish the importance of each sub-criterion relative to the others. This process was implemented using the WEIGHT tool in IDRISI GIS software (Fig. 2). Weights were determined according to a series of pairwise comparisons of secondary factors based on their importance, producing standard weights whose sum equals 1. These weights were used to build the main factors within the framework using linear weighting, performed with the MCE tool integrated into IDRISI software tools.

A critical step in AHP is constructing a pairwise comparison matrix, that compares each sub-criterion to determine their relative importance in influencing drought vulnerability. We used a scale ranging from 1 to 9 [15] to quantify these comparisons. A value of 1 indicates equal importance, whereas higher values indicate that the row criterion is more important than the column criterion in terms of its influence on drought vulnerability in the study area. For instance, a value of 3 in the "Climate / Soil" cell suggests that the climate factor is considered three times more important than soil characteristics in affecting drought vulnerability. Conversely, a value of 1 in the "Climate / Socio-economic pressure" implies that both factors are perceived as having equal importance.

2.3.2. Consistency Check

In practice, it is incorrect to expect decision-makers to present pairwise comparison matrices of criteria in an integrated and consistent manner, especially in cases where there are many criteria. The consistency index (CR) was calculated to verify the consistency of the matrix results as follows:

$$\:\varvec{C}\varvec{R}=\frac{\varvec{C}\varvec{I}}{\varvec{R}\varvec{I}}$$

RI is the stochastic stability index (mean CI of 500 randomly filled arrays) such that if the CR is less than 0.1 or 10%, the matrix is said to have acceptable consistency. The RI of this research was 0.58 (Table 2).

Table 2

**RI (Random Index) Values for Matrices of Different Sizes** [16].
N	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
RI	0	0	0,58	0,9	1,12	1,24	1,32	1,41	1,45	1,49	1,51	1,48	1,56	1,57	1,58

And CI is the consistency index, which is calculated by:

$$\:\varvec{C}\varvec{I}=\frac{\varvec{\lambda\:}\mathbf{m}\mathbf{a}\mathbf{x}-\varvec{n}}{\mathbf{n}-1}$$

Where, λmax is the largest number of eigenvector matrices, and n is the number of drought hazard parameters

In this matrix, the value of CR = 0.03 indicates that the comparison of the main parameters of the surface sensitivity map to drought was highly stable and that the weights were appropriately selected in the suitability assessment matrix. This method was applied to determine all the criteria, whether basic or secondary.

3.1. Sub-Criteria of Climate Sensitivity

The sub-criteria for assessing climate sensitivity were selected on the basis of their specific contributions. The Standardized Precipitation Index (SPI) stands out as the primary factor because of its detailed insights into climatic drought and its variability, significantly influencing drought severity [18]. Consequently, SPI is assigned a higher weight in the pairwise comparison matrix. Given its moderate impact on climate sensitivity, the De Martonne Aridity Index follows. Finally, the evapotranspiration criterion was ranked lower, owing to its comparatively weaker effect (Table 3).

Table 3

Pairwise Comparison Matrix of Sub-Criteria of Climate Sensitivity
	SPI	AI	PET
SPI	1	3	5
AI	1/3	1	3
PET	1/5	1/3	1

Table 4

Normalized Pairwise Comparison Matrix
	SPI	AI	PET	Weight
SPI	0.65	0.70	0.56	0.64
AI	0.22	0.23	0.33	0.26
PET	0.13	0.07	0.11	0.10
SUM	1	1	1	1
CR = 0.03

Table 5: Climate Sub-Criteria Reclassification and Weights Determined by the Analytic Hierarchy Process

Basic Criteria	Sub-Criteria	Classes	Reclassify	Weight
Climate sensitivity	SPI Index	1 - 0.50	1	0.64
		0.50 - 0	2
		0 - -0.49	3
		-0.50 - -0.99	4
		-1 - -1.49	5
	aridity index Ai	> 55	7	0.26
		35-55	6
		28-35	5
		24-28	4
		20-24	3
		10-20	2
		< 10	1
	Evaporation-transpiration PET	1255 – 1904	1	0.10
		1905 – 1961	2
		1962 -2043	3
		2044 - 2241	4

This map goes beyond merely depicting the spatial distribution of SPI drought index values; it serves as a critical visual tool to uncover trends in precipitation decline across the Middle Moulouya and Guercif Basins from 2000 to 2021. By analyzing the SPI values, we can identify areas that have experienced statistically significant decreases in precipitation. This information is crucial for identifying the region’s highest risk of future droughts, enabling proactive planning and resource allocation to mitigate potential impacts. This map is also valuable for researchers investigating the link between the observed precipitation trends and climate change drivers in the region.

From the map, it is evident that the southwestern and northeastern parts experienced a severe decrease in precipitation, with SPI values ranging from − 1.49 to -1. The southern and southeastern regions, along with a narrow part in the northeast, also witnessed significant rainfall deficiencies with index values between − 0.99 and − 0.55. The northern and northwestern parts experienced a moderate decrease in precipitation, as indicated by SPI values ranging from − 0.49 to 0. Conversely, the central part of the study area did not experience any decrease in precipitation, with SPI values between 0 and 1 indicating no drought during the study period according to available statistics.

The aridity index is crucial for determining the degree of Drought in the Middle Moulouya and Guercif Basins from 2000 to 2021. The study area is predominantly arid, with an aridity index ranging between 11 and 20, covering approximately 51% of the total area. According to the De Martonne Aridity Index, approximately 7% of the area is located in mountainous regions and is characterized by a semi-humid climate.

We adopted the evaporation-transpiration criterion as a significant factor in identifying areas that have experienced substantial water loss, leading to a deficit and decline in water reserves. The evaporation-transpiration distribution map revealed that the southern regions of the study area recorded the highest values, ranging between 2045 and 2240 mm, covering approximately 463,313.70 hectares, or 23.54% of the study area. The next highest values, ranging between 2040 mm and 1960 mm, are concentrated in the southern and northern parts of the study area, particularly in the Guercif Basin, covering approximately 626,313.58 hectares, or 31.82%.

This variation in the distribution of potential evapotranspiration values significantly affected soil structure and texture. The larger the soil pores, the greater the evaporation intensity and soil moisture loss, which are exacerbated by dry southern winds.

3.2. Sub-Criteria of Soil Sensitivity

The construction of the soil sensitivity criterion map required reliance on three secondary criteria:

1. The layer of organic matter percentage in the soil
2. The layer of land uses for the study area
3. The layer of soil water holding capacity

The importance of each criterion varies according to its effect on the sensitivity of soil to drought hazards. Generally, most studies consider the criterion of soil moisture and its ability to retain water to be of greater importance than the percentage of organic matter in the soil and land use standards.

Table 6

Pairwise Comparison Matrix of Sub-Criteria of Soil Sensitivity
	SH	OM	LU
SH	1	2	3
OM	1/2	1	3
LU	1/3	1/3	1

Table 7

Normalized Pairwise Comparison Matrix
	SH	OM	LU	Weight
SH	0.55	0.60	0.43	0.55
OM	0.27	0.30	0.43	0.27
LU	0.18	0.10	0.14	0.18
SUM	1	1	1	1
CR = 0.05

Table 8

Soil sensitivity sub-criteria reclassify and weights determined by the analytic hierarchy process
basic criteria	sub-criteria	CLASSES	RECLASSE	WEIGHT
Soil sensitivity	Organic matters in soil	1.3–4	1	0.27
		0.45–1.3	2
		0.25–0.45	3
		0.20 - 0.25	4
		0 - 0.20	5
	Land use	forest	1	0.18
		irrigated land	2
		matorral	3
		Bare rocks	4
		Bult-up	5
		steppes	6
		bare land	7
	soil humidity	5 - 0	1	0.55
		8 - 5	2
		10 - 8	3
		12 - 10	4
		12 - 22	5

Land use maps of the study area is used to understand the current landscape, revealing that most activities in this region are concentrated in the southwestern part of the watershed. Seven classes were developed for land use (Fig. 4), with bare ground ranking the highest because of its susceptibility to drought hazard.

Soil moisture and its ability to retain water are among the most critical factors responsible for stabilizing soil and increasing soil resistance by providing suitable conditions for vegetation cover growth. The higher the water-holding capacity of the soil, the more drought-tolerant it was. This process is governed primarily by the structure and texture of the soil.

The soil moisture coefficient map [14] revealed significant variations in moisture content across the arid and drought-prone Middle Moulouya and Guercif Basins. The soil moisture values ranged from 0–22%, with distinct categories affecting drought vulnerability.

- Dry Soils: Occupying the largest area (26.7%), the dry soil category (0–5% moisture) stretches across the central eastern and northern regions. This includes areas around the Moulouya River (estimated 22.3% of dry soils).

- Limited Moisture Soils: Slopes in mountainous areas are dominated by weakly moist soils (8–10% moisture), encompassing approximately 19.1% of the total area.

- Moderate to High Moisture Soils: Areas with better drought resilience are those with medium (11–12% moisture) and high moisture (13–22% moisture) moisture content. These categories represent 11.7% and 20.2% of the total area, respectively.

The presence of organic matter in the soil was directly tied to the vegetation cover. Denser vegetation promoted a higher organic matter content. Because vegetation also thrives with soil moisture, areas lacking sufficient organic matter are more susceptible to drought. This highlights the crucial role of land management practices that promote vegetation growth in improving soil moisture retention and reducing drought vulnerability.

The map demonstrating the distribution of organic matter in the soil across the study area revealed a marked disparity. The percentage of organic matter decreases as we move from the highlands and slopes towards the depressions, with the highest percentages recorded in the highlands and irrigated agricultural lands.

The proportion of organic matter in the studied field was divided into five categories:

Constitutes: This accounted for 2% of the study area, with organic matter values ranging from 1.3–4%. These areas are concentrated in mountainous highlands and irrigated cultivation zones, where vegetation cover is relatively dense.
Covers: This accounted for 10.8% of the study area, with organic matter values ranging from 0.45–1.3%. This category is widespread on some slopes of the eastern Middle Atlas Mountains to the west, the High Atlas to the south, and the Debdou Heights in the northeast.
Occupies: Approximately 37% of the study area, with values ranging from 0.25–0.45%. This category is widespread in the eastern and western parts, where steppe plants are found on the slopes of the Middle Atlas to the west, the Eastern High Atlas to the south, and the edge of the upper plateaus and the Debdou Heights to the east.
Represents: About 5.6% of the study area, with organic matter percentages ranging between 0.20% and 0.25%. It is concentrated in the northeast and south of the study field.
The most widespread category: Covering approximately 43.5% of the total study area. It contains the least organic matter, with percentages ranging from 0–0.20%.

The degree of susceptibility of the surface to drying varies with the extent of its protection. Land uses were divided according to their sensitivity to drought as shown in Table 8. Generally, areas with dense vegetation are considered less sensitive than bare and open lands, while populated areas are more sensitive than sparsely populated or unpopulated areas.

The map of land uses for the study area shows that bare lands dominate, covering about 62% of the total area. This is followed by steppes, which are used extensively and intensively as pastures, losing much of their vegetation cover and accounting for about 21%. Forests and natural vegetation cover about 6% and 5% of the area, respectively. Agricultural lands, residential areas, and bare rocks collectively represent 6% of the study area.

3.3. Sub-Criteria of Socioeconomic Sensitivity

Socioeconomic data were obtained from the departments and institutions whose field of intervention covers the study area. Population data were based on the General Population and Housing Census for 2014, available on the website of the High Commissioner for Planning. Agricultural data related to the number of livestock and agricultural areas were obtained from the Agricultural Consultation Center in Outat El Haj city and Guercif city.

The numerical data collected from these various departments were processed and converted into geographic databases using ArcGIS software. These databases were then transformed into a grid layer (Raster) using the Inverse Distance Weighting (IDW) tool.

Table 9

Pairwise Comparison Matrix of Socioeconomic Sensitivity
	pastoral pressure	agricultural exploitation	population distribution
pastoral pressure	1	1/3	1/5
agricultural exploitation	3	1	1/3
population distribution	5	3	1

Table 10

Normalized Pairwise Comparison Matrix
	pastoral pressure	agricultural exploitation	population distribution	Weight
pastoral pressure	0.11	0.07	0.13	0.10
agricultural exploitation	0.33	0.23	0.22	0.26
population distribution	0.56	0.70	0.65	0.64
SUM	1	1	1	1
CR = 0.05

Table 11

Socioeconomic Sensitivity Sub-Criteria Reclassification and Weights Determined by the Analytic Hierarchy Process
basic criteria	sub-criteria	CLASSES	RECLASSE	WEIGHT
Socio-economic sensitivity	pastoral pressure	1.5–35.5	1	0.10
		35.5–70	2
		70–105	3
		105–140	4
		140 175	5
	agricultural exploitation	> 100	1	0.26
		100–200	2
		200–350	3
		350–550	4
		550 >	5
	population distribution	100–270	1	0.64
		280–380	2
		390–550	3
		560–1800	4

The human presence in the field can have two contradictory roles. The first is negative and predominant, as it can contribute to exacerbating the sensitivity of the surface to drought through the nature and intensity of exploitation and the suitability of the latter with the environment in which humans live [19]. The second is positive in some cases, represented by human contributions to environmental balance and helping it adapt to phenomena that threaten its balance, such as drought, thereby reducing losses. However, in the Middle Moulouya and Guercif basins, humans live in this environment with a fragile structure. It has been associated with the intensive and excessive exploitation of resources, so their presence exacerbates the sensitivity of the surface to drought (Fig. 5).

The animal pressure map reveals a higher concentration of livestock in mountainous areas, decreasing towards the Moulouya River. This distribution reflects the dominance of pastoralism in the highlands, where suitable pastures exist compared to agriculture-focused flatlands. Further contributing to this pattern is the traditional nomadic herding practice, where shepherds constantly move their herds, potentially hindering plant regeneration. The influx of outside herds adds to the pressure on vegetation, highlighting the need for sustainable grazing practices to maintain rangeland health.

The Middle Moulouya and Guercif Basins are characterized as arid and dry areas with low to medium population density [12]. Agricultural exploitation is increasing and exceeding it carrying capacity for this type of activity. This over-exploitation has led to the expansion of agricultural areas at the expense of pastoral areas, especially following agricultural reclamation and subsequent developments in agricultural policy, while disregarding the issue of reclaiming deserted lands or those threatened with sand encroachment and desertification. This increases vulnerability to natural factors represented by the lack of precipitation and the succession of dry periods.

The intensity of agricultural exploitation in arid, dry, or semi-arid regions increases the severity of drought impacts on these fragile environments, emphasizing the need to consider this criterion in determining the socioeconomic sensitivity criterion. Limited cultivated land increases the vulnerability of smallholders to drought [20], [21], [22]. Smaller landholdings are often associated with lower production due to limited capitalization and low technological input [23].

Analyzing the map of agricultural exploitation intensity in the Middle Moulouya and Guercif basins, it is clear that there is significant agricultural pressure on the lands, especially in the northern and central parts of the study area and along the Moulouya Valley. The concentration of agricultural activities in these areas is due to the presence of water resources, represented by the Moulouya Valley and its tributaries, "Mlalou River" and "Masoun River," in addition to an intensively exploited shallow aquifer (Fig. 5).

The multi-criteria decision-making method is a branch of effective applied research models for solving complex problems characterized by a high degree of uncertainty and conflicting objectives. It depends on various data, information, interests, and points of view, as well as monitoring changes in environmental and socio-economic systems [24].

As previously mentioned, the first step is to form a series of pairwise comparisons within the pairwise comparison matrix by combining criteria according to their relative importance and impact on susceptibility to dryness. We used a number scale indicating the number of times one criterion is less or more important or exerts greater dominance than the others, with values ranging from 1 to 9 [17]. For example, entering a value of 3 in the field of "Climate / Soil" means that the climatic factor is three times more important and affects susceptibility to drought in the study area than the soil factor. Entering a value of 1 in the field of "Climate / Socio-economic stress" means that both factors are equally important (Table 12).

The next step is to calculate the principal eigenvector, determined by calculating the geometric mean for each criterion and then deriving the weighting factor (weight) for each criterion by dividing each vector by the sum of the vectors, which must equal 1.

The results obtained by the AHP method must be verified to ensure that the matrix weights are coherent, and this verification is evaluated by the coherence/consistency index CI. The concept of coherence in the pairwise comparison of (SAATY) is based on the respect and appropriateness of the provisions made in the comparison matrix, even when this index is calculated in the following form:

𝐶𝐼= (𝜆𝑚𝑎𝑥−N)/(N−1)

Where: N is the number of items compared.

𝜆max is a value calculated based on the SAATY matrix, eigenvectors, and N.

Table 12

Pairwise Comparison Matrix of Socioeconomic Sensitivity
	Climat	Socio-Economic	Soil
Climat	1	3	5
Soci-Economic	1/3	1	3
Soil	1/5	1/3	1

Table 13

Normalized Pairwise Comparison Matrix
	Climat	Socio-Economic	Soil	Weight
Climat	0.65	0.70	0.56	0.64
Soci-Economic	0.22	0.23	0.33	0.26
Soil	0.13	0.07	0.11	0.10
∑₌	1	1	1	1
CR = 0.03

In this matrix, the value of CR = 0.03, which indicates that the comparison of the main parameters of the sensitivity drought map had a high degree of stability and that the weights were appropriately chosen in the suitability evaluation matrix.

4.1. Soil Sensitivity Criterion Map

The distribution map of the categories of soil sensitivity to drought reveals a variation in the area occupied by each category and a difference in its geographical distribution. Soil sensitivity increases as we move from the highlands towards the Moulouya depression, approaching the valley course except for the irrigation areas on its banks. The vulnerable areas constitute 8.55% or 168,495.94 hectares of the total area. The vegetation cover of this region is of medium to high density, consisting of irrigated agricultural lands, open forests, matorral, and highlands with a semi-humid climate, which have relatively cohesive soils. The slightly and moderately sensitive areas constitute 14.17% or 279,308.50 hectares and 11.62% or 229,092.12 hectares, respectively, distributed over pastures, deserted fields, and allied steppes. Highly sensitive areas constitute 28.17% of the surface or 555,345.98 hectares, where the vegetation cover is sparse and consists of saline plants and degraded pastures. The areas of very high sensitivity represent 37.49% or 739,143.21 hectares, mainly located in barren and sandy soils where moisture and water retention are very low, and they are the same areas that have been intensively exploited.

4.2. Climate Sensitivity Criterion Map

An analysis of drought vulnerability across the study area reveals significant spatial variations. Non-sensitive areas, encompassing 16.49% (or 64,135.53 hectares) of the land, are primarily concentrated in the central region. Areas with weak and moderate sensitivity levels are more extensive, covering 23.69% (or 66,265.95 hectares) and 18.40% (or 72,434 hectares), respectively. These are predominantly located in the northern part. However, the situation becomes more concerning in the southern and southwestern portions of the Middle Moulouya basin, where areas classified as highly and very highly sensitive to drought collectively represent 41.42% of the total land area. This includes 44,495.17 hectares (20.34%) designated as highly sensitive and 42,200.22 hectares (21.08%) classified as very highly sensitive. These findings highlight the urgent need for targeted drought mitigation and adaptation strategies, particularly in the most vulnerable southern and southwestern regions.

4.3. Socioeconomic Sensitivity Criterion Map

Human presence and its associated facilities are crucial criteria for measuring the degree of sensitivity of a surface to desiccation [11]. The distribution map of socioeconomic sensitivity shows significant variation in sensitivity categories both spatially and in terms of area occupied. Sensitivity increases in populated areas and regions with intensive agricultural exploitation and decreases as we move away from them.

Non-sensitive areas constitute 10.43% of the study area or 206,735.59 hectares and are located in the heights of the eastern Middle Atlas and the northwestern slopes of the eastern High Atlas. These areas are situated in the western and eastern parts of the study area. Medium and high-sensitivity areas are located in the southeastern part of the study area, covering 18.59% (368,454.37 hectares) and 21.41% (424,267.74 hectares), respectively. These categories are spread near roadways and low-density population centers.

Very high-sensitivity areas represent about 4.12% or 81,605 hectares and are concentrated in regions with high population density and intensive agricultural activities, particularly around small urban centers like Missour and Outat El-Haj.

4.4. Surface Sensitivity Drought Map

The above analysis enabled the development of a decision map to assess areas according to their vulnerability to drought. This map classifies the study area into five categories of vulnerability based on pixel values. The drought-prone areas are distributed differently in terms of vulnerability and the area occupied by each category.

Very Low Sensitivity Areas: These areas, which are not at risk of drought, represent 4.78% of the study area or 93,784.13 hectares. They are mainly located in mountainous heights, particularly on slopes oriented to the north and west, where the soil is relatively moist and precipitation exceeds 600 mm. Additionally, agricultural lands with dense vegetation and extensive irrigation also fall into this category. These regions enjoy environmental privileges compared to the rest of the study area due to the density of vegetation and isolation from human pressure, making them a haven for many animal and plant species despite their small size.
Low Sensitivity Areas: Constituting 11.23% or 220,280.90 hectares of the study area, these areas are spread in the central part of the study area and surrounding regions that are not at risk of drought, particularly on the outskirts of agricultural lands. These areas require a permanent water supply, leading to increased pressure and over-exploitation.
Moderate Sensitivity Areas: Covering approximately 320,194.73 hectares or 16.32% of the total area, these areas are located on uncultivated fallow lands and steppes.
High Sensitivity Areas: These areas represent 38.88% of the study area, covering 762,804.65 hectares, and are distributed across barren and sandy soils, especially in the northern and southern parts of the study area.
Very High Sensitivity Areas: Constituting 28.80% or 565,079.13 hectares of the total area, these regions are also located in barren and sandy soils, particularly in the northern and southern parts of the study area.

Figure 7

Spatial Distribution of Surface Sensitivity to Drought Hazard over the Middle Moulouya and Guercif Basins

This study applied a detailed multi-criteria assessment method to evaluate drought sensitivity in the Middle Moulouya and Guercif basins in Morocco. By combining Geographic Information Systems (GIS) with the Analytic Hierarchy Process (AHP), we were able to explore the complex factors contributing to drought vulnerability, such as climate, soil, and socioeconomic conditions.

Our findings highlight that the northern and southern parts of the study area are particularly vulnerable to drought. These regions often experience high and very high sensitivity, mainly due to intensive agricultural activities and the presence of barren, sandy soils with low moisture retention. The combination of high population density and intensive land use further increases their vulnerability, underscoring the urgent need for targeted drought mitigation efforts.

On the other hand, areas with dense vegetation, higher soil moisture, and sustainable land management practices showed lower sensitivity to drought. This points to the importance of encouraging sustainable agricultural practices, such as crop diversification and improved irrigation techniques, to boost resilience against drought.

The spatial distribution maps generated in this study are crucial for decision-makers, as they pinpoint priority areas that need immediate attention for drought mitigation and adaptation. Implementing sustainable water management solutions, like rainwater harvesting and efficient irrigation systems, could significantly lessen the impact of future droughts in these regions.

Moreover, community-based initiatives can play a vital role in empowering local residents to adopt sustainable land use practices and manage grazing pressures effectively. Educating communities about the risks of overexploitation and the benefits of environmental stewardship can foster a shared sense of responsibility for protecting the area's natural resources.

Overall, the methodology and insights from this study offer a valuable framework for assessing drought vulnerability in other arid and semi-arid regions. By integrating diverse data sources and expert input, this approach supports informed decision-making and proactive resource management, ultimately enhancing ecosystem resilience and promoting long-term sustainability.

Author Contribution

-ELYAGOUBI Said wrote the main manuscript.-All authors prepared the figures and reviewed the manuscript.

D. Saini, O. Singh, T. Sharma, and P. Bhardwaj, “Geoinformatics and analytic hierarchy process-based drought vulnerability assessment over a dryland ecosystem of north-western India,” Nat. Hazards, vol. 114, no. 2, pp. 1427–1454, Nov. 2022, doi: 10.1007/s11069-022-05431-7.
N. Brooks, W. Neil Adger, and P. Mick Kelly, “The determinants of vulnerability and adaptive capacity at the national level and the implications for adaptation,” Glob. Environ. Change, vol. 15, no. 2, pp. 151–163, Jul. 2005, doi: 10.1016/j.gloenvcha.2004.12.006.
D. Wilhite and M. Glantz, “Understanding the Drought Phenomenon: The Role of Definitions,” Drought Mitig. Cent. Fac. Publ., Jan. 1985, [Online]. Available: https://digitalcommons.unl.edu/droughtfacpub/20
T. R. Ault, J. S. Mankin, B. I. Cook, and J. E. Smerdon, “Relative impacts of mitigation, temperature, and precipitation on 21st-century megadrought risk in the American Southwest,” Sci. Adv., vol. 2, no. 10, p. e1600873, Oct. 2016, doi: 10.1126/sciadv.1600873.
C. Koem, K. Nusit, and S. Tantanee, “Spatial Distribution of Drought Hazard Mapping Based on AHP and GIS in Kampong Speu Province,” vol. 16, pp. 442–450, Jan. 2022.
V. Choudhary, P. Jain, and H. S. G. Vishwavidylaya, “Screening of alkaline protease production by fungal isolates from different habitats of Sagar and Jabalpur district (M.P),” 2012. Accessed: Feb. 10, 2023. [Online]. Available: https://www.semanticscholar.org/paper/Screening-of-alkaline-protease-production-by-fungal-Choudhary-Jain/9e7a9ebd3326f2a4057dfc15549d2ea861f513e4
M. A.-A. Hoque, S. Tasfia, N. Ahmed, and B. Pradhan, “Assessing Spatial Flood Vulnerability at Kalapara Upazila in Bangladesh Using an Analytic Hierarchy Process,” Sensors, vol. 19, no. 6, Art. no. 6, Jan. 2019, doi: 10.3390/s19061302.
M. Palchaudhuri and S. Biswas, “Application of AHP with GIS in drought risk assessment for Puruliya district, India,” Nat. Hazards, vol. 84, Dec. 2016, doi: 10.1007/s11069-016-2526-3.
R. Pandey, V. Jain, and M. Jain, “Spatio-temporal assessment of vulnerability to drought,” Nat. Hazards, vol. 76, Nov. 2014, doi: 10.1007/s11069-014-1502-z.
M. Hoque, B. Pradhan, and N. Ahmed, “Assessing drought vulnerability using geospatial techniques in northwestern part of Bangladesh,” Sci. Total Environ., vol. 705, Feb. 2020, doi: 10.1016/j.scitotenv.2019.135957.
A. Bensaïd, “SIG AND rtemote sensig FOR the STUDY OF The STRANDING IN An ARID REGION: <br />A case study: Wilaya of Naâma (Algéria),” phdthesis, Université Joseph-Fourier - Grenoble I, 2006. Accessed: Feb. 10, 2023. [Online]. Available: https://theses.hal.science/tel-00169433
S. Elyagoubi and A. Mezrhab, “Using GIS and remote sensing for mapping land sensitivity to wind erosion hazard in the middle Moulouya Basin (North-Eastern Morocco),” J. Arid Environ., vol. 202, p. 104753, Jul. 2022, doi: 10.1016/j.jaridenv.2022.104753.
S. Pandey, A. C. Pandey, M. S. Nathawat, M. Kumar, and N. C. Mahanti, “Drought hazard assessment using geoinformatics over parts of Chotanagpur plateau region, Jharkhand, India,” Nat. Hazards, vol. 63, no. 2, pp. 279–303, Sep. 2012, doi: 10.1007/s11069-012-0093-9.
N. M. Singh and T. T. Devi, “Assessment and Identification of drought prone zone in a Low Laying Area by AHP and MIF method: A GIS based study,” IOP Conf. Ser. Earth Environ. Sci., vol. 1084, no. 1, p. 012047, Oct. 2022, doi: 10.1088/1755-1315/1084/1/012047.
T. L. Saaty, “How to make a decision: The analytic hierarchy process,” Eur. J. Oper. Res., vol. 48, no. 1, pp. 9–26, Sep. 1990, doi: 10.1016/0377-2217(90)90057-I.
T. Saaty, “The Analytic Network Process – Dependence and Feedback in Decision-Making: Theory and Validation Examples,” 2005, pp. 360–386. doi: 10.4018/978-1-59140-702-7.ch018.
T. L. Saaty, “A scaling method for priorities in hierarchical structures,” J. Math. Psychol., vol. 15, no. 3, pp. 234–281, Jun. 1977, doi: 10.1016/0022-2496(77)90033-5.
L. Zhai, “Spatial and temporal pattern of precipitation and drought in Gansu Province, Northwest China,” Nat. Hazards, vol. 49, pp. 1–24, Apr. 2009, doi: 10.1007/s11069-008-9274-y.
M. Boultif, “SIG et Modélisation pour la Cartographie des Zones Vulnérables à la Sécheresse et à La Désertification dans un Ecosystème Méditerranéen Semi-Aride : Application dans la Zone d’El Hodna,” doctoral, Université de Batna 2, 2018. Accessed: Feb. 11, 2023. [Online]. Available: http://eprints.univ-batna2.dz/1654/
C. A. Harvey et al., “Extreme vulnerability of smallholder farmers to agricultural risks and climate change in Madagascar,” Philos. Trans. R. Soc. B Biol. Sci., vol. 369, no. 1639, p. 20130089, Apr. 2014, doi: 10.1098/rstb.2013.0089.
O. Jamshidi, A. Asadi, K. Kalantari, H. Azadi, and J. Scheffran, “Vulnerability to climate change of smallholder farmers in the Hamadan province, Iran,” Clim. Risk Manag., vol. 23, pp. 146–159, Jan. 2019, doi: 10.1016/j.crm.2018.06.002.
A. Keil, N. Teufel, D. Gunawan, and C. Leemhuis, “Vulnerability of smallholder farmers to ENSO-related drought in Indonesia,” Clim. Res., vol. 38, no. 2, pp. 155–169, Feb. 2009, doi: 10.3354/cr00778.
J. F. Morton, “The impact of climate change on smallholder and subsistence agriculture,” Proc. Natl. Acad. Sci., vol. 104, no. 50, pp. 19680–19685, Dec. 2007, doi: 10.1073/pnas.0701855104.
J.-J. Wang, Y.-Y. Jing, C.-F. Zhang, and J.-H. Zhao, “Review on multi-criteria decision analysis aid in sustainable energy decision-making,” Renew. Sustain. Energy Rev., vol. 13, no. 9, pp. 2263–2278, 2009.

No competing interests reported.

Integrating GIS and AHP for Drought Sensitivity Assessment in the Middle Moulouya and Guercif Basins, Morocco

Status:

Version 1

Abstract

Figures

1. Introduction

2. Material and Method