Keywords

1 Introduction

Urban sprawls and densely populated cities have created an immense urge on higher water consumption to fulfill several needs. The primary hurdle faced by any developing nation is inadvertent urbanization, deficient water resources and unproductive management of water supply and distribution. Developing nations specifically in arid and semi-arid climatic regions are still gathering reforms and resources to combat such issues [1]. Countries closer to sea or ocean focus on desalination and reservoir storage as water preservation practices. Since the 1990s, another technique that has gained attention is artificial groundwater recharge (AGR)[2]. It has paved the way for arid and semi-arid nations as one of the most sustainable approaches to water conservation due to higher evaporation rates [3].

The United Arab Emirates (UAE) is one of the most brisk developing nations. Its population leaped sharply from 531,265 in 1975 to 10 million in 2022 [4]. The niche lifestyle of the UAE population demands water consumption of about 550 L per capita per day in 2020 compared to the global average of 170–300 L per capita per day which is 82% higher than any countries [5]. Note that the UAE is located in an arid climate region with limited freshwater resources and scarce rainfall [6]. Approximately, 51% of the UAE’s water supply comes from freshwater resources [7]. Thus the UAE depends on alternative water preservation practices such as desalination, water storage units, artificial groundwater recharge to meet demands during lean periods. As the country’s temperature rises approximately 50 degrees Celsius in summers, any surface water storage takes a heavy toll on the economy due to high evaporation rates. Therefore the country is indulging into more sustainable practices of water conservation by adopting AGR. AGR is a scientific water storage technique of available water into aquifers and use it in dry periods for agriculture, industrial and potable purposes [8]. Many studies [8,9,10,11,12,13,14] have been implemented through Remote Sensing (RS) and Geographic Information System (GIS) for AGR demarcation and implementation. Primary step is to identify potential zones where RS & GIS comes to a rescue. Humble decision making approach is adapted for AGR demarcation and then tools of machine learning like artificial neural network, support vector machine, random forest and so on are further integrated to validate the data if we analyse previous researches for AGR across the globe [10, 11, 14,15,16,17,18]. Less AGR studies have been carried out in UAE and specially central northern UAE covering Emirates of Sharjah, Umm al Quwain, Ras al Khaimah, Fujairah and Ajman [19]. Therefore this research made an attempt to demarcate potential AGR sites for the previously mentioned emirates.

The research utilized multicriteria decision analysis by Saaty 1990 paired with weighted overlay analysis to identify potential zones for AGR. The study was carried out at the Central northern Emirates including emirates of Sharjah, Fujairah, ras al Khaimah, Umm al Quwain, Ajman, and portions of Oman. This study utlized more previous studies for analysing ranking and weighting of Analytical Hierarchical Process (AHP) compared to previous research [19] and also calculated the percentage of the area suitable and not suitable for AGR. Nine thematic layers were prepared: precipitation, drainage stream density (DSD), geomorphology, geology, groundwater level, Total Dissolved Solids (TDS), elevation, lineament density (LD), distance from residential areas. The research aims to delineate suitable locations for implementing AGR by employing RS, GIS, AHP, and the weighted overlay technique. The main objectives of this study is summarized within the following:

  • Investigating suitable zones for AGR in Sharjah by utilizing RS and GIS.

  • Identifying and mapping spatial thematic layers for AGR zonation: precipitation, drainage stream density (DSD), geomorphology, geology, groundwater level, Total Dissolved Solids (TDS), elevation, lineament density (LD), Euclidean distance from residential areas.

  • Employing AHP and weighted overlay techniques to obtain AGR map.

  • Calculating percentage of area suitable for AGR and the primary factors governing it.

2 Study Area

The United Arab Emirates is a Mediterranean country located in the western part of Asia and is surrounded by the Arabian Gulf in the west and north, Gulf of Oman in the northeast, Oman in the east, and Saudi Arabia in the south (Fig. 1) [6, 20,21,22]. The country’s water demands are met by 43% of its available groundwater resources. However, due to its arid climatic condition, the country has reserves of only 640 billion cubic meters (BCM) of groundwater, and 20 BCM is fresh [21, 23]. The country receives approximately 102 mm of mean annual rainfall within cities and 130 mm in mountains with temperatures rising to 48 degrees during summer months in desert [21, 23, 24]. Majority of the land is covered with sand and so forth, the sand dune aquifer system dominates the geographic region. Other classifications of the aquifer system are northern limestone, ophiolite, eastern gravel, western gravel, and coastal marshes [25]. Five major cities of northern UAE have been covered in this study which includes Sharjah, Umm al Quwain, Ras al Khaimah, Ajman, and Fujairah. This region comprises equal proportions of all the aquifers classes.

Fig. 1.
figure 1

Study Area.

Full size image

3 Methodology and Data Processing

Figure 2 demonstrates the methodology developed for identifying potential zones for AGR. Suitable remote sensing imageries such as digital elevation model (DEM), Landsat 8, enhanced thematic mapper plus, historical data of precipitation, and salinity were used to prepare desired thematic layers contributing to AGR potential zones. The previously discussed factors pprecipitation, DSD, geomorphology, geology, groundwater level, TDS, elevation, lineament density, euclidean distance from residential areas were developed as spatial thematic layers in ArcGIS Pro. Five classes ranging from 1 to 9 derived from Natural Breaks technique, have been assigned to each layer, with 1 representing least suitability and 9 as most suitable. AHP technique helped to identify weights of each criteria. To obtain the final potential map, weighted overlay analysis was performed.

Fig. 2.
figure 2

Methodology Framework.

Full size image

3.1 Thematic Layers Preparation

This section demonstrates the data collection, image processing aspects and its correlation to the potentiality of AGR, of each parameter primarily considered for the objective of the study. Delineating suitable AGR zones were achieved through nine spatial parameters: Precipitation, DSD, geomorphology, geology, groundwater level, TDS, elevation, lineament density, euclidean distance from residential areas [26,27,28]. The detailed discussion about the layers are mentioned in the following paragraphs.

Precipitation.

National Centre for Meteorology, UAE, was referred to deduce annual total rainfall historical data from 2003–2017 refer, Fig. 3(a). Data was collected from the rain gauge stations and coordinates of each station were utilized to develop the point map. Inverse distance weighting interpolation technique further obtained the spatial layer of precipitation. The data helped to understand that the least average of annual total rainfall was recorded as 75mm for the study and ranged maximum up to 103mm. From multiple research articles it is profound that the regions receiving more rainfall is directly proportional to AGR potential zones [21, 23, 24].

Drainage Stream Density.

DSD of the study area ranged from 0 to .58 per km2 Fig. 3(b). The parameter is defined as the capacity of the total water drained through stream channels within a watershed represented as the ratio of total stream length of all orders by the total area of the drainage basin. Permeability of water decreases if DSD is higher [17, 26]. It is one of the fundamental parameters for AGR demarcation. SRTM (Shuttle Radar Topography Mission) DEM spatial data were used to develop the thematic layer [23]. Increased regional DSD can be noticed in the eastern part of the Sharjah Emirate, north and northeastern part of Ras-al-Khaimah and Umm-al-Quwain, respectively [21, 23].

Geomorphology.

Figure 3(c) represents the geomorphology map of the study area: fan deposits, high and low dunes, mountain, sand, urban areas, and vegetation [21, 23]. Geomorphological patterns allow us to understand the water flow and movement below the ground level and explains the water storage capacities depending on permeability and porous landforms [17, 29]. Spatial data from Landsat8 ETM+ of 30 m spatial resolution was collected and processed to develop the thematic layer [23]. This study showed that fan deposits are most favorable for AGR in the UAE and accordingly the highest rank has been assigned to this criteria. As the majority of the country is covered with desert sand, high dunes also come in a higher ranking with respect to AGR determination, refer.

Geology.

Landsat 8 was downloaded and shortwave infrared band were geo-processed to develop the geology thematic layer. The geology of the study area comprises: alluvium, limestone, gabbro, metamorphic, sand and ophiolite [21, 23]. Sand covers more than 45% of the study area. Recharging volume can be determined by understanding the porosity and assigning spaces for water holding [17, 29]. Alluvium holds more water compared to other classes. This study marked the alluvium class with the highest rank followed by sand.

Total Dissolved Solids.

Historical data from the Ministry of Environment and Water, UAE (2015) were considered to prepare the spatial thematic layer of TDS. TDS has a great impact on the quality as it exceeds the turbidity of water which eventually determines the fit and unfit for potable use. Permeability of water with high TDS can clog pores stopping the flow through the aquifer and risks of pathogen activities increases leading to many diseases to humankind. Regions with lower values of TDS are considered to be fit for AGR zonation [30, 31]. Generally, TDS values are enormously high near to the coastal shorelines. TDS map of this study conductively portrayed that regions closer to Gulf of Oman have values of 38000 mg/l whereas the west of the study area which is covered by Arabian Gulf have TDS values of 50000 mg/l. This demonstrates the outcome that Arabian Gulf is more saline compared to Gulf of Oman.

Groundwater Level.

In-situ data from bore-wells were collected within the study area and combined together making interoperability in the GIS platform to deduce the thematic layer, refer Fig. 3(d). Inverse distance weighted interpolation technique was used in ArcGIS Pro to develop the layer. The units are meters above sea level (masl). This factor helps in analyzing the hydraulic gradient of the region which eventually depends on pore pressure and atmospheric pressure at the surface level [17, 27, 32]. Higher levels of groundwater is less likely suitable for AGR [17, 32]. The thematic layers shows that the southeast Sharjah and west of Ras al-Khaimah have higher values of groundwater levels. One of the primary explanation behind this is that the region is in close proximity to the foothills and receives more rainfall which up levels the groundwater. The IDW equation is as follows [13, 21, 33]:

$${\rm Z}_{0}=\frac{{\sum }_{i=1}^{N}{z}_{i} x {d}_{i}^{-n}}{ {\sum }_{i=1}^{N}{d}_{i}^{-n}}$$
(1)

Where Z = calculated value of Z at o;

zi = observed value at sample point I; di is the distance between sample point i and o; N is the number of sample points used to estimate the value at o; n is a distance decay parameter [34, 35].

Elevation.

SRTM DEM of 30m spatial resolution were downloaded and used for developing the thematic layer. The elevation is as low as zero meters above sea level (masl) near shore lines and as high as 1112 masl in the mountains for the study area, Fig. 3(e). Elevation is inversely proportional to AGR zonation which demonstrates that inferior elevation values are more appropriate for AGR [9, 27, 35,36,37]. It analyses the accumulation capacity of water and its flow direction through the aquifers. Few regions within Fujairah and Ras-al Khaimah have elevation above 1000 masl.

Lineament Density.

Linear features such as fractures and folds were extracted from satellite images and then spatial maps were processed. It has direct proportionality to AGR suitability. Availability of groundwater and potential available aquifers can be determined from his parameter. It also determines the secondary porosity. Higher values of LD can be witnessed in Fujairah. Generally, from previous research it has been concluded that a zone of around 300m closer to any folds and fractures of the earth composition can be considered suitable for AGR [29].

Distance from Residences.

To develop this layer, shapefiles of residential areas were mapped and then Euclidean distance was calculated to prepare the thematic layer, refer Fig. 3(f). For sustainable AGR sites the distance from residences factor needs to be considered. AGR needs to be designed at optimum distance from the densely located city considering water pumping factors to required places as well as closer proximity to pipelines [8, 35]. Higher the distance from residences more suitability for AGR.

Fig. 3.
figure 3figure 3

Thematic Layers [19].

Full size image

3.2 Analytical Hierarchical Process

AHP is a structured multi-criteria decision making (MCDM) procedure. It utilizes experts’ opinions to determine the weights and rank of each parameter. AHP has been used by researchers in various groundwater studies and site selection studies [8, 21, 23, 33, 38,39,40,41]. This study employs AHP for selecting potential zones for the AGR. An important step in AHP is prioritizing the parameters and assigning them weights, as discussed below.

Weighting the Parameters.

The influencing factors were arranged as a structured hierarchy and the pairwise comparison matrix were formed to confirm the consistency of the weights. The weights of the all 9 selected parameters were placed in a square matrix keeping all the diagonal values as 1. The diagonal of the pairwise matrix was kept 1 when both the parameters have equal importance over the other. The relative importance of the parameters were analyzed using the principal eigenvalue along with the normalized right eigenvector of the matrix [21, 23, 42, 43]. On a scale of 1–9 of relative importance by Saaty [44] all the parameters were ranked. Highest rank 9 was assigned to criteria with highest influence on the decision. Measurements of consistency were done by checking the randomized and consistency index as well as the consistency ratio.

Consistency Ratio.

CR allows the glance for the subjectivity of determined weights through a pairwise matrix. In order to confirm the consistency of the pairwise comparison matrix, consistency index (CI), consistency ratio (CR) and randomized index (RI) were obtained. CR is defined as the degree of consistency of the comparison matrix prepared with respect to parameters and its weights. The value of CR must be less than 0.01 for the consistency of the matrix to be maintained [21, 23, 43]. The CR can be derived using the following equations [44]:

$$\mathrm{CI }= \frac{\lambda max - n}{n-1}$$
(2)
$$\mathrm{RI }= \frac{1.98 \times (n-1)}{n}$$
(3)
$$\mathrm{CR }=\frac{\text{CI}}{\text{ RI}}$$
(4)

CI is a consistency index, RI is a randomized index (average of CI values of the comparison matrix), CR is a consistency ratio, λmax is the maximum eigenvalue of a comparison matrix and n is the order of the comparison matrix. The calculated CR equals.007 < .01, which supports the weighting model and the AHP technique [21, 45]. Fig. 4 Demonstrates the weights of the utilized thematic layer associated with AGR in percentage.

Fig. 4.
figure 4

Weights of utilized thematic layers for AGR zonation in percentage [19].

Full size image

4 Results and Discussion

The potential AGR was estimated and mapped using AHP approach, refer Fig. 5. The map was categorized on the basis of ordinal scale into 6 classes: “very high”, “high”, “moderate high”, “moderate low”, “low”, and “very low”.

Fig. 5.
figure 5

Potential zones of AGR.

Full size image

Following points were concluded from the AGR map:

  1. (i)

    The very high zone is located in the central part of Ras al Khaimah. The properties of the input parameters have geology- alluvium, geomorphology -fan deposits, precipitation of 91 mm, groundwater level 102 masl, drainage density at .38 per km2, TDS 1862 mg/l, distance from residences .14 m, elevation of 87 m and lineament density of .05 per km2.

  2. (ii)

    The “high” zone is located in the north-eastern and central-eastern part of Sharjah. This region also comprises geomorphology- fan deposits, geology-alluvium. This region is near the foothills of the Al-hajjar Mountains and thus comprises many suitable qualities of geology and geomorphological characteristics for AGR. TDS of 1300 mg/l were noted in this region.

  3. (iii)

    Table 1 demonstrates the area in percentage for each class of AGR.

Table 1. Percentage of area for AGR classes.
Full size table

5 Conclusion

The main objective of this paper was to determine locations which are suitable for AGR, delineate them using AHP and be able to get an estimate of the proportion of study area which can benefit from AGR techniques. To go about this approach, 9 main parameters were found to be governing AGR site suitability. A weighted AHP process was utilized to rank these parameters and then a weighted overlay analysis was done to obtain the final AGR suitability map. The map was clustered in 6 areas ranging from very high suitability to very low suitability. Several inferences were then drawn from the resulting map. The north-central part of Ras Al Khaimah was observed to be lying in the most highly suitable zone while another major portion of Ras Al Khaimah and Sharjah were seen to be lying in the zone characterized as high suitability for AGR. Another reason was the high elevation and mountainous region which received high amounts of precipitation and this in turn caused high amounts of fan and alluvial deposits in these regions. On the other hand, both the shorelines were found not suitable for AGR due to the presence of high salinity. Similarly, dense populations in the urban locations of the eastern parts also made these areas unsuitable for AGR. Out of the total study area, 20% was estimated to be a highly suitable category, 30% as moderate-high, and 12% as low. This was done using ArcGIS Pro and utilizing pixel sizes as the deciding factor. Semi-arid and arid countries such as those in the gulf can highly benefit from this research methodology to demarcate and locate potential locations and zones suitable for AGR techniques. As the world moves into a stage where water becomes more and more a scarcity, the methodology described in this paper can serve as a guiding tool at the least to help find suitable locations for AGR and reduce the stress caused by water scarcity.