Introduction

The arid regions of Morocco such as the study area, where rainfall is not regular, are suffering from water scarcity problems. The difficulties are exacerbated by the over-exploitation of water resources for agriculture, mining, or daily activities. However, access to water resources, especially potable water, is a crucial factor for sustainable development (Baki et al. 2017). The aim of this study is to assess the potential for groundwater storage in the region that is underlain by the Isly basin, to contribute to the search for suitable areas for groundwater withdrawal for productive uses, and to enable a good coverage of the water needs of the region.

In the present study, we have combined remote sensing and Geographic Information System (GIS) techniques with a hydrogeological conceptual model to determine the groundwater prospectivity of the region. We did this as these are the most effective and economical methods of assessing the likely availability of groundwater while covering extensive zones. In addition, satellite imaging remains the only method for obtaining information from areas that are difficult to access on the ground. The application of these procedures to estimate groundwater potentiality has recently attracted the attention of many researchers in different regions of the world (Ahmed et al. 1984; Teeuw 1995; Sander et al. 1996; Edet et al. 1998; Nouayti et al. 2017).

The process that was used in this study to assess the likely availability of groundwater was based on the exploitation of data extracted from topographic maps, satellite images, digital terrain models, and geological map, to develop a thematic mapping of the parameters that were used to develop a groundwater availability model. These factors included the annual rainfall of the area, the lithology of underlying rocks, the degree to which these rocks were fractured, the slope of the land, and the characteristics of surface drainage. The validation procedure was based on a comparison of the results got from the developed model with flow rates at known groundwater withdrawal locations.

Geological and hydrogeological review

The Isly basin is in the northeast of Morocco (Fig. 1), geographically between X = 790,000, X = 850,000 and Y = 415,000, Y = 460,000 (Lambert_Conformal_Conic). Representing one of the principal watersheds of the Horst belt, Isly valley is surrounded by several mountain ranges (Metsila, Touissit, and Tiouli mountains) that extend from 597 to 1654 m in altitude.

Fig. 1
figure 1

Location of the study area and simplified geological map of the Isly basin (from the geological map of Morocco at 1/1,000,000, Lambert coordinates in m)

Full size image

Geologically, rocks within the Isly basin range from the Paleozoic to the Plio-Quaternary in age. The oldest outcrops are intruded by granitic bodies and have been deformed by several tectonic orogenesis (Pique et al. 1989; Torbi and Gélard 2000). The Mesozoic cover comprises detrital deposits, dolerite intrusions, calcareous, and marls (Salvan 1974; Rakus 1979; Oujidi 1994). The Plio-Quaternary deposits are represented by calcareous crusts, conglomerate deposits, lacustrine carbonates and basalt flows (Michard 1976; Laville 1981; Cattaneo 1987).

The geomorphology of the Isly basin covers a variety of mountain ranges, from which several streams flow into the Isly valley over an area of about 920 km2. There is a dense hydrographic network in the hydrological framework of the study area, constituting the headwaters of a majority of the rivers that run along the Isly watershed (Fig. 2).

Fig. 2
figure 2

The Isly watershed and location of springs

Full size image

The groundwater depths in the study area vary from a few meters to over 76 m. Yields from individual wells can exceed 40 l/s in the fractured Liassic formations in the region and the electric conductivity values of groundwater vary between 890 and 1190 μS/cm. The main springs documented in the Isly watershed have been identified by Mortier et al. (1967).

The Isly basin is in an arid bioclimatic region (El Hafid et al. 2017). There are significant seasonal variations in temperature with a hot summer and a cold winter; however, the annual rainfall regime is characterised by the existence of two rainy seasons in autumn and spring, with an average annual rainfall reaching 245 mm in the high basin and 200 mm downstream of the basin (Fig. 3).

Fig. 3
figure 3

Correlation of rainfall with temperature from Guenfouda meteorological station

Full size image

Data and processing

It involves various factors in determining potential areas for groundwater storage. Mashael (2010) summarized several studies by indicating the quality of the results got according to the parameters used. He found that the variables that were considered in developing a groundwater availability model would have an enormous influence on the outcomes and their reliability. The elements that are often used to develop such models are usually rainfall, rock type, drainage, lineaments, and slope. In this study, Sentinel 2 satellite images and a DEM obtained from the Earth explorer site web allowed us to identify and digitize the last two determinants.

Integrating these factors systematically into a GIS tool as thematic layers will cause a map showing potential zones for groundwater storage, with several categories (Fig. 4). In this study, this was done using the following weights for each parameter: rainfall (10%), rock fractures (lineaments) (30%), lithology (30%), drainage (15%) and slope (15%).

Fig. 4
figure 4

Organizational methodology adopted for empirical Model of the Isly basin

Full size image

Precipitation

The frequency and magnitude of rainfall events and their reliability have a significant impact on the amount of recharge that groundwater systems will receive. In this study, we classified the rainfall map that was compiled for the survey area into five major classes: > 250, 250–200, 200–150, 150–100 and < 100 mm/year. As we can see from Fig. 5, the annual average precipitation gradually decreases from the east to the western part of the study area.

Fig. 5
figure 5

Rainfall map classes of the Isly watershed

Full size image

Lithology

Lithology has a significant impact on groundwater availability as this factor influences the amount of water that can be stored in the sub-surface (El-Baz and Himida 1995) and that can be transported in groundwater flow (Oikonomidis et al. 2015). In the study area, we got Lithological data from the available 1/100,000 geological map of Morocco (Fig. 6).

Fig. 6
figure 6

Map of lithology classes in the Isly watershed

Full size image

Further modifications of that map were performed after the existing remote sensing data for the study area were reviewed. In particular, the use of Sentinel-2 satellite imagery (pixel size 30 m) was helpful and allowed us to correct many rock formations boundaries. Afterwards, an empirical classification of the lithological formations was carried out. A chart with five classes revealed rock layers with similar hydrologic properties.

Lineaments

Fracturing may enhance the porosity and permeability of rocks and increase the ability of surface water to infiltrate into the subsurface and seep into groundwater (Dhekra et al. 2018). The pattern of bedrock fracturing can also influence the surface drainage model. The distribution and density of fractures were included in the groundwater prospectivity model to identify potential areas for groundwater storage (Fig. 7). Fracture systems can be successfully recognized from remote sensing imagery (Toutin 1996). In this study, we identified the locations of lineaments from Sentinel 2 imagery.

Fig. 7
figure 7

(A) Lineaments map of the Isly watershed and the directional rosettes (B: total length, C: frequencies)

Full size image

We superimposed the resulting map on the geological map by using a GIS for correlation and interpretation. Lineaments were expressed and then counted by classifying the area into specific frames. The main directions of these fractures are NE–SW to E–W; NW–SE trending segments are relatively less abundant.

We divided the resulting lineaments density map into five classes (Fig. 8). A high lineament length density shows the likelihood of good permeability and porosity in rocks and provides better targets for groundwater recharge than geologic joints, with potentially high levels of deep percolation (Yeh Hsin-fu et al. 2016).

Fig. 8
figure 8

Map of lineament density classes Km/Km2

Full size image

Slope

Topography plays a role in influencing the amount of surface water that can infiltrate into the groundwater. In general, less infiltration of rainfall into soils takes place on steep slopes than on gently sloping land. The slope parameter controls water flow and its energy and thus affecting groundwater recharge processes. This factor can help with the identification of water potential areas (Fig. 9).

Fig. 9
figure 9

Slope map classes

Full size image

The greater slope, the faster flow of surface water and the less recharge will result. In this study, we got the generation of a slope map from a digital elevation model (DEM), which gives a three-dimensional view of the field surface. The classification adopted is the SOTER (SOils and TERain) model (European Commission 1995).

Drainage

Drainage density is a sum of the lengths of all talweg axes per unit area (Pouliot et al.1994). Drainage characteristics have a contribution in distributing storm-water and provide a sign of the infiltration regime. In the Isly basin, we digitized the hydrographic networks from the sheets of the 1: 50,000 scale topographic map using GIS functions. The rate of surface-water recharge into groundwater is governed also by the characteristics of the drainage system (Huang Chih-Chao et al. 2013).

The resulting drainage density map (Fig. 10) yielded five drainage classes by following the same approach used in representing the lineament density map. Each class, therefore, represents a certain number of drainage segments per unit area. These classes were < 0.3, 0.3–0.6, 0.6–0.9, 0.9–1.2, > 1.2 km/km2. The area with high drainage density has a high level of groundwater recharge.

Fig. 10
figure 10

The map of drainage density classes Km/Km2

Full size image

Empirical model and data validation

The empirical model results from the integration and superposition of the different parameters, which influence groundwater storage (rainfall, lithology, lineaments, slope, and drainage) as layers in the GIS tool. For each parameter, we assigned a coefficient with a weight according to its importance. The map of potential groundwater storage zones is got by adding the effect of the five factors (Fig. 11).

Fig. 11
figure 11

Empirical Model showing groundwater potential zones of the study area

Full size image

This model presents 10% of the total area (920 km2) corresponds to the zone of “very high” recharge potential, 20% “high” potential, 30% “moderately” potential, 25% and 15% “low” to “very low” recharge potential of groundwater in the study area. The zone of “medium” to “very high” potential surfaces represents approximately 60% of the total area, suggesting that the Isly basin is relatively rich in groundwater.

We validated the model using the distribution and density of water supply wells in the study area. It should be noted that these parameters are geographically dependent. By including these factors in the model, it is possible to produce an empirical model that shows areas of potential groundwater storage (Fig. 12).

Fig. 12
figure 12

Location of springs and boreholes and their productivity in the potential model of groundwater storages

Full size image

To validate the empirical model, the positions and yields of wells and springs in the study area were verified and compared with the potential groundwater availability for these zones that was got from the model. From the water point flows, we found that there was a good agreement between the predicted productivity of different areas and the returns of the actual water supply points (Fig. 13), except for a few water springs and boreholes with high flows in parts with “medium” to “very low” groundwater storage potential.

Fig. 13
figure 13

Diagram showing the situation of water points in the model

Full size image

Accordingly, the observed groundwater yields from water springs and boreholes generally correlate well with the mapped zones of groundwater storage potential. Consequently, the empirical model would allow developers to target areas that are the most prospective for proving an adequate water supply without wasting time and resources in investigating groundwater availability in unprospective area.

Conclusions

We developed an empirical model in this study to predict groundwater availability in the Isly basin in Morocco. The model assumes that the spatial variability of groundwater in the study area depends on annual rainfall, lithology, land slope, and the density of surface drainage in the Isly watershed. We integrated existing data for these parameters within a GIS environment to map the predicted variation of groundwater storage capacity in the survey area.

This new empirical model of groundwater availability was applied in the Isly basin in Morocco. The prototype shows that about 60% of the Isly watershed has a “moderate” to a “very high” potential for groundwater storage, indicating that the survey area is potentially rich in groundwater, especially in areas with intense fracture systems. We validated this model using the existing distribution of wells and springs in the study area. The proposed model was proved very satisfactory and could be used to select a favorable area to dig water wells. We present the resulting model as a decision-making aid and also a pre-prospecting which can avoid heavy, slow, and expensive research phases.