In the United Arab Emirates, the share of the agricultural sector in the overall gross domestic production (GDP) was only 3.8% in 1999 [1
] and declined to less than 1% in the year 2013 [2
]. Despite its limited contribution to the GDP in the UAE, the agricultural sector used around 70% of the total water demand in the Emirate of Abu Dhabi in the year 2012 [3
]. In the UAE, irrigation water is either from conventional water resources like groundwater and springs or from non-conventional water resources like desalination plants and treated wastewater facilities. The development of the agriculture sector in the UAE faces two main challenges. The first challenge is related to the prevailing arid atmospheric conditions which make that the region receives the very low amount of rainfall that is essential for the development and expansion of the agricultural sector. Second, the soil and land surface conditions in the region led to an overall low suitability of the land which also limited the potential agricultural sector and its productivity in the UAE. The latter is the focus of this study as it is important to investigate the spatial pattern of land suitability in the country and determine if it has been used to its full potential or not. The comparison of the determined land suitability maps to the current extent of lands used in farming in the UAE should indicate areas where future agriculture development should be done.
Lands that are suitable for irrigated agriculture represent only 6.81% of the total land area of UAE. Arable and permanent cropland represent 0.77% and 2.39% of land area respectively [4
]. During the period of 2000–2009, the annual growth of arable and permanent croplands increased by 4.88% (654,000 ha) and 1.5% (1,939,000 ha) respectively, in comparison to 7.79% (420,900 ha) and 37.43% (640,000 ha) during the period of 1990–2000 [4
] marking therefore a decline in the expansion of cropland in the country. Forest area represents 3.8% of the total land area in the UAE, with an annual growth of only 0.24% (3,122,000 ha) during 2000–2009. So, the limited expansion of land that is used for agriculture in the UAE has certainly made achieving food security in the country challenging. In fact, local agriculture production currently satisfies less than 10% of local needs [5
On the other hand, a substantial increase in water desalination and non-conventional water resources development in the country has been observed in the last few years. Almost 1.7 billion cubic meters of desalinated water was pumped in 2011 which made the UAE the second largest producer of desalinated water in the world after Saudi Arabia. This amount accounts for 14% of the world’s total output of desalinated water, which is a remarkably high proportion [6
]. About AED 12 billion is being spent per year on water desalination, from about 70 major seawater desalination plants [6
]. The government has initiated several programs to encourage the reuse of TSE in agriculture, forestry and urban design sectors. TSE is used in a controlled range of agricultural production in several piloting farms in the UAE [7
]. However, it is being heavily used in forest plantations. In 2013, desalination and treated sewage effluent contribute about 35% to the total water demand [8
So, in the UAE, on one hand, the agriculture sector is facing serious challenges related to the aridity of the climate and the low suitability of the land. On the other hand, the country is recording a significant increase of non-conventional water production which may help to boost productivity in the agriculture sector. It is essential to carefully analyze and aggregate all the available information to optimize the sustainable use of land and water and accurately determine its suitability. Proper management of the limited resources maximizes agricultural productivity and helps in achieving food security in the country. This is the main goal that we propose to pursue in this study through the determination of land suitability in the Emirate of Abu Dhabi that accounts among other for non-conventional water resources.
The assessment of land suitability for agriculture is a complex, multidisciplinary and multi-criteria process which entails land topography, climate, water resources available for irrigation, soil capabilities and current management practices including land use and land cover [9
]. This complexity calls for the application of appropriate decision support tools, such as the multi-criteria decision making (MCDM). MCDM is one of the most widely used methods of overcoming the difficulties in defining relative weights of several criteria involved in decision-making on land suitability [11
]. An integrated suitability assessment for land use planning and sustainable development purposes has been developed in the Emirate of Abu Dhabi [5
] and used four soil related criteria namely salinity, depth, texture and moisture followed by sequential assessments of the topography and water availability criteria. An integrated Soil Information System in the United Arab Emirates (UAESIS) using multi-criteria decision-making approach was developed in 2014 to define land suitability for date palm production [14
]. The study results show that 14.03% and 16.29% of total lands in the Emirate of Abu Dhabi are highly and moderately suitable for date palm plantation, respectively. However, non-conventional water resources were not used in [14
There have been many MCDM methods used to assess land suitability like ordered weighted average [15
], simple additive scoring [17
], outranking methods [18
], logic scoring of preference [19
] and the analytical hierarchical processes (AHP) [21
]. The latter has been largely used to solve complex decision-making processes which include multiple criteria, sub-criteria and alternatives [22
]. AHP involves ranking relative criteria into a hierarchical structure, assessing the importance of these criteria per each level, comparing all alternatives for each criterion and determining an overall ranking of the alternatives [23
]. Moreover, the determination of land suitability requires the integration of geospatial information from multiple sources among others on land cover, land use practices, soil type, soil nutrient content, pollution, weather condition, water resources and the technology used in agriculture. Such integration should be done in a geographic information system (GIS) [24
] framework where diverse layers of information could be aggregated and processed to identify the most suitable location for agriculture for specific crops [25
], coupling therefore AHP-GIS.
The AHP-GIS integrated method has been increasingly used in recent years as a powerful spatial decision support system in different fields; for land suitability assessment for agriculture [26
], irrigated agriculture [21
], eco-tourism purposes [28
] and land-use suitability assessment [29
]. In the AHP-GIS integrated method, assessing goal, criteria and alternatives need to be identified in relation to the purpose of the study. In selecting assessment criteria, attention should be paid to consider only those relevant to the decision-making process and contribute to the final goal. A spatial layer that includes all suitability classes with respect to the specific criterion in a specific location presents one evaluation criterion. The suitability classes then need to be rated and aggregated according to their relative importance based on the contribution of each criterion, to achieve the intended goal or objective.
This study addresses two original aspects. First, we combine in the context of an arid region a large number of agronomic and climatic factors into management, water resources and socio-economic factors, (16) criteria and (80) sub-criteria to define land suitability for seven of the Emirate’s most critical crops as identified by the Government of the UAE in the national food security [30
] and food diversification strategies [31
]. Second, we introduce non-conventional water resources in the analysis, namely, desalinated water and treated sewage effluent as the main sources of water for irrigation for specific crops (non-edible and climate resilient crops) to ensure the sustainability of irrigated agriculture under current and future climate.
1.1. Study Area
The UAE is located on the eastern corner of the Arabian Peninsula to the north of the Sultanate of Oman and the Kingdom of Saudi Arabia. It is composed of seven coastal Emirates, six of which are located on the coast of the Arabian Gulf (they stretch over more than 650 km), while the Fujairah emirate lies on the Gulf of Oman stretching over ~90 km. The “83,600 km2
” total area of the UAE is mostly covered by sandy soils. The highest point in the UAE is Jebel Yibir at 1527 m, while the lowest point is the coastal area on the Arabian Gulf. The desert sand dunes dominate the UAE western and southern parts and merge into the Rub’ Al Khali desert of Saudi Arabia as illustrated in Figure 1
. The desert also includes two important oases—Al-Liwa Oasis (Mezaira) near the border with Saudi Arabia and Al-Buraymi Oasis shared with the Sultanate of Oman.
This study focuses on the Emirate of Abu Dhabi, the largest emirate in the UAE because of data availability constraints. The Emirate occupies more than 85% of the UAE’s mainland area covering an area of over 67,000 km2
]. The climate is extremely hot and humid in the Emirate mainly in summer. The average annual rainfall is estimated at 124 mm in the east coast, 131.9 mm in the mountain region, 107.7 mm in the gravel plain and 74.9 mm in the lowland desert [33
]. The northeastern mountains receive the highest amount of 160 mm, while the southern desert receives less than 40 mm per year, with prevailing winds from North West. The evaporation rate is high, up to 2–3 m per year on average [34
]. The mean daily air temperature ranges from 14 °C in winter to more than 45 °C in summer which is not suitable for many crops. The relative humidity can also be very high and may reach up to 100% in certain areas, especially near to the coast [35
]. The study area is also affected by steep increasing trends of the number of dust events [36
] which might have an adverse effect on agriculture activities. The relative humidity can also be very high and may reach up to 100% in certain areas, especially near to the coast [37
The study area has limited freshwater resources (only groundwater). Therefore, agriculture is heavily dependent on groundwater which has been depleted due to over-exploitation [38
] and salt-water intrusion [39
]. Estimates for 2011 indicate that groundwater reserves in the Emirate amount to 635.6 billion cubic meters (BCMs), out of which only 3% (19.1 BCMs) is fresh water [40
]. The depth of groundwater ranges between 5 and 100 m [41
]. However, the over-exploitation of groundwater throughout the years has led to a severe depletion of the reserve at a rate of 1.5 to 5 m per year. The current groundwater storage is insufficient for large-scale agricultural plantation.
Groundwater salinity is another critical issue in the Emirate as it ranges between 0 and 500 ppm and 125,000–160,000 ppm. Desalinated water is the main source of water for domestic use in the UAE. Treated wastewater, also widely known as TSE, is used in a controlled range of agricultural production in several farms in the UAE [42
]. According to the UAE Ministry of Environment and Water each tree needs between 18 and 30 L of water per day [43
]. The UAE treated around 450 MCM wastewater per year out of which only 60% is reused [44
About 80 percent of the soil in the UAE is sandy soil with low organic matter [7
] and therefore, as is, it is marginally suitable for agriculture activities. Soil classification in the UAE is based on the “Keys to Soil Taxonomy” that was developed by the International Center for Biosaline Agriculture in collaboration with the Environment Agency-Abu Dhabi (EAD) in 2013 [45
]. The soil salinity in the Emirate is divided into four major classes with salinity values ranging between 0 and <2 dS/m EC for non-saline soil and equal to and greater than 40 dS/m EC for highly saline soil [46
]. There are two soil orders in the Emirate—Aridisols and Entisols [47
]—and six suborders, eight groups and 89 soil families [45
]. The study area contains 14 different soil textures (clay, clay loam, coarse sand, coarse sandy loam, fine sand, fine sandy loam, loam, silty clay loam, loamy fine sand, loamy sand, sand, sandy clay, silty clay and silt). The soil in the study area is divided into four main sub-criteria in relation to soil moisture wet, humid, dry and very dry. Soil depth differs significantly among different soil types [49
]. The study area is divided into five main soil depth categories; common hardpan, moderate probability, low probability, none or rare and not mapped areas. The land slope is not a critical factor in the UAE as more than 90% of the study area has a slope below 5.3% percentage. Land elevation ranges between −1 and 1132 m above sea-level.
1.2. Datasets Sources and Processing
Sixteen datasets are gathered and used in this study under five main categories, namely, climate, water resources, topography, soil capabilities and management data sets. Meteorological data for a 30-year period is used to create temperature and relative humidity layers using METAR data, while the precipitation layer is generated using the National Center for Meteorology (NCM) database for the years 2003 to 2015. Slope, elevation and aspect maps are derived from the UAE-ASTR-Digital Elevation Model [50
] with 30-m resolution. Numerical data sets are converted to spatial layers to create the distance to desalination plants and wastewater treatment facilities. Desalination plants information were retrieved from the DesalData website (www.DesalData.com
), while the treated sewage facilities database was provided by EAD. The Emirate of Abu Dhabi land-use map is generated from the Moderate Resolution Imaging Spectroradiometer (MODIS) Land Cover Data Base (www.eros.usgs.gov/land-cover
). The spatially aggregated data for each year in the period 2001–2012 is used at 0.5° × 0.5° resolution. Table 1
presents the description of the data sets used in this study and their sources.
Data collection and preparation in GIS is one of the fundamental steps in land suitability analysis. Different GIS techniques are used including interpolation, model building, reclassification, recalculation and weights overlay functions. All datasets including the numerical and spatial layers were converted into raster layers at a spatial resolution of 105 m, which is the coarsest resolution of the available spatial layers. Raster reclassification is used to reclassify all spatial layers into the five sub-criteria classes as integer raster representing different suitability levels. Then, all layers were projected or re-projected into Abu Dhabi Transverse Mercator using ArcGIS Spatial Analyst. The new spatial datasets were processed in ArcGIS. The produced layers are presented in Figure 2
. Then, all layers were recalculated using the weights assigned to each sub-criterion based on the AHP analysis, before applying the Weighted Overlay function. The resulting weighted overlaid raster contains the five suitability classes.
The selection of the relevant criteria was based on the received feedback of local experts from The International Center for Biosaline Agriculture, Food and Agriculture Organization of the United Nations, EAD and the UAE Ministry of Climate Change and Environment. A comprehensive analysis of the literature [11
] showed that similar regional and global studies have demonstrated the importance of using the following criteria: precipitation [10
], temperature [19
], relative humidity [51
], groundwater salinity, groundwater table [52
], soil texture [10
], soil moisture [54
], soil depth [10
], soil salinity [52
], aspect [19
], slope [19
], elevation [19
], land use [53
] and soil capabilities [19
]. In this study, we are introducing the use of desalinated water and TSE for irrigation as a supplementary source–a significant omission from previous investigations. The sub-criteria of these main criteria are the different distances to the desalination plants and wastewater treatment facilities. Stations with private ownership or very small capacities are neglected from the study.
This study successfully implemented an integrated AHP-GIS model as an advanced and comprehensive MCDM approach to evaluate and define a land capability for irrigated agriculture suitability. The study was used to derive several AHP structures and suitability maps based on a significant number of criteria encompassing climate, water resources, topography, soil characteristics and land management. The inputs criteria were evaluated using AHP pairwise comparison matrix per crop. The highest capability weights, as presented in the resulted maps, were mostly found in areas of highly capable soils and in areas close to water resources (groundwater, desalination facilities and TSE). Moreover, comparisons of the date palm derived map and the Environment Agency-Abu Dhabi map indicated small disagreements, demonstrating the effectiveness of the AHP-GIS-based method.
Land in the central and eastern part of the Emirate has the highest capability weights (scores), confirming that current agricultural farms were developed on highly capable land. The selection of seven different crops provided a good indication of how suitability maps vary as a result of the different crop requirements, which are presented by different decision-making perspectives (AHP). For vegetables and cereals plantation, output maps indicate that areas of excellent suitability are limited and not exceeding 7% for each crop. Those suitable lands are close to agricultural zones, water resources and proper soil texture and moisture. Jojoba and sorghum output maps indicate the largest area of excellent suitability for those two crops productions, due to their ability to adapt to harsh environmental conditions and their limited need for continuous and excessive irrigation. When comparing all output maps, jojoba and sorghum present the most suitable crops, followed by date palm, fruits and forage and finally vegetable and cereals.
The inclusion of non-conventional water resources altered the classification of some areas making them suitable for irrigated agriculture. Consideration of defining a suitable area for large-scale plantation for specific crops that are tolerant to a harsh climate like jojoba is also incorporated.
Finally, this study is the first of its kind in the UAE to define a set of findings that will determine whether large-scale plantations are recommended for the UAE. It is the first study to identify the kind of crops that have the highest potential to adapt to the hot and dry weather without affecting the crop yield that ensures the sustainable use of the limited groundwater resources. According to the government’s draft food diversification strategy (planned to be published soon), there is an urgent need to develop land suitability maps for irrigated agriculture in the UAE. This work corresponds directly to the requirements of the government of Abu Dhabi and provides a good evidence of the applicability of the AHP-GIS method to encompass a very large number of input criteria (16) and sub-criteria (80) in comparison with other commonly used multi-criteria methods. To our knowledge, this is the first paper to use AHP-GIS to cover 16 criteria, 80 sub-criteria and for seven crops.
Future work of this study is to cover other plants and Emirates. Additional plants such as Jatropha, Sporobolus virginity and Distichlis spicate should be considered in future research. Furthermore, a detailed analysis should be undertaken on the potential of these plants for carbon sequestration, energy production for desalination plants in remote areas and their contribution to evapotranspiration and large-scale climate processes. In addition, the focus should be on expanding the analysis to the rest of the UAE and the Arabian Peninsula and analyzing the impact of changes in climate conditions in the region on the suitability maps of different plantations, which should lead to a better understanding of the impact on food security and the sustainable use of natural resources in regional and local level.