East Uweinat area is an important area of agricultural reclamation in the southwestern
part of the Western Desert of Egypt, which attracted enormous investments of
governmental and private sectors in the last three decades. The Nubian Sandstone
aquifer system (NSAS) is the sole water resource used for all development purposes.
The land and groundwater resources of this area cover the requirements needed
for the reclamation and implementation of new communities (GPC,
1984; GARPAD, 1994). The regional study area of East
Uweinat is located between latitudes 22° 00' 24.68''-23° 27' 35.53''
N and longitudes 27° 59' 50.62'' -29° 13' 36.72'' E, with an area of
about 20, 438 km2 (Fig. 1, 2).
The area of study is generally a flat plain with hill ridges and scarps, which
are mostly rugged and rough. The ground surface elevation varies from about
225-420 m above sea level (masl) (Fig. 3). It is characterized
by high temperature and low relative humidity during summer. It has also has
a long dry rainless summer and short rainless winter.
|| Location map of the study area
Wind velocity increases sometimes causing sandstorms, especially during the
Khamasine periods. The daily mean temperature is 22.86°C. It varies between
12.8°C in January to 30.7°C in July.
|| ETM+satellite image (bands 7 4 2) of the study area (acquired
|| Location of groundwater wells, east Uweinat area
The average maximum temperature is 30.7°C. It varies between 17.1°C
in January to 39.95°C in June, while the highest recorded temperature was
46.0°C. The average annual minimum temperature is 14.5°C.
|| Geologic map of east Uweinat area (CONOCO,
1987 NG36SW-Luxor sheet)
It varies between 5.2°C in December to 22.1°C in July. The annual average
relative humidity is 28.8%. It reaches the highest values in January (42.8%)
and the lowest values are during May (18.5%).
Geologically, the dominant rock units occupying this area extend from Cretaceous
to Quaternary (Fig. 4). Meanwhile, the Nubian Sandstone (Six
Hills Formation) is considered as the main exploitable water-bearing formation.
This aquifer system is used for the agricultural and socio-economical development
of the Western Desert of Egypt. The Nubian Sandstone Aquifer System (NSAS) was
formed by the local infiltration during the past wet periods (pluvial periods),
which ended in the Northeastern Sahara at about 8000 years ago, while ended
in the south (Nubian Desert; northeast of Sudan) at about 4000 years ago (Heinl
and Thorweihe, 1993). Sedimentary units of the East Uweinat and southern
areas of the Western Desert have been extensively studied for their freshwater
potential, among them (GPC, 1984; Nour,
1996; Heinl and Thorweihe, 1993; Robinson
et al., 2007; Elewa et al., 2010;
Nahry et al., 2010). The oldest exposed rocks
in the studied area are the crystalline basement rocks of Precambrian age. Overlying
basement rocks, there are different rock units with different ages, with well-authenticated
unconformable relationships (GSE, 1987) (Fig.
4). The tectonic movements, which affected the area, probably, started early
in the Precambrian. They sculptured the southern part of Egypt into three major
intracratonic basins and relatively small and shallow basins. These basins occupy
now the Nubian Plateau, Dakhla and Barket El-Shab areas and they have in between
the Nakhlai-Aswan High and Tarfawi Kharga High. Two main sets of faults were
noticed, a north-south set observed in the vicinity and an E-W set, which crosses
the area for a long distance. Supplementary sets of northeast and northwest
trends are also noticed. It is possible that these faults are associated with
the uplift of the basement in this area. At the study area, a thick subsurface
section of Nubian formations exists. This section was investigated by different
authorities, such as the General Petroleum Company (GPC,
1984) and the General Authority for Rehabilitation Projects and Agricultural
Development (GARPAD, 1994). The maximum fully penetrated
thickness of this section reaches 723 m in Well No. 27 (GARPAD,
1994). It is assigned to Pre-Cenomanian.
From 1980s to 2010s, exploration and exploitation drilling campaigns along
the southern part of the Western Desert of Egypt were conducted to assess the
water resources of the NSAS (GPC, 1984; GARPAD,
1994; Nour, 1996; Elewa et
al., 2010) (Fig. 3). The transboundary NSAS covers
an area of about 2.4 million km2 and extends in south Libya, Egypt,
north Sudan and northeast Chad (Thorweihe, 1990) and
also extends as far north as to Sinai. It is bounded from east, south and south
east by basement outcrops, while in north it is bounded by a fresh-saline interface
occurring in the vicinity of Qattara Depression in the northern part of Egypt
(Ezzat, 1974) (Fig. 1). It overlies
basement rocks that are crosscut by an extensive E-W fault system in southern
Egypt that caused the vertical uplift (Issawi, 1978).
In Egypt, it is classified into six distinct geological units ranging in age
from Jurassic to Upper Cretaceous (Klitzsch and Lejal-Nicol,
1984) (Fig. 4).
However, the planned area to be cultivated in 1980s was 210.03 km2
from a total area of 2100.32 km2 representing the top priority of
land capability (GARPAD, 1994). Up till 2003, the total
cultivated lands are about 118.81 km2. The ongoing strategy of development
and cultivation project of 924.14 km2 in East Uweinat needs about
1826 wells beside 58 wells in El-Ain Village (Ministry of Agriculture and Land
Reclamation Personal communications). Until 2006, about 383 wells were drilled
by the Egyptian Government (Robinson et al., 2007).
For more than three decades, governmental experts prepared a strategy for development
of East Uweinat area and national plans were settled up. The study area is about
20,437 km2, which covers the expected future expansion of groundwater
exploitation activities. The number of wells drilled in the 1980s had increased
dramatically through the 1990s and 2000s periods. For this reason, tracing the
spatial distribution of aquifer attributes using the same wells is not available
as outgrowths in both lands and well numbers with time, is a matter of fact.
However, due to the vast area available for reclamation in East Uweinat, different
well spots in different time periods are favorable for determining the aquifer
spatial variation. Following this variation will be of prime importance in determining
available areas for further horizontal expansion of reclamation projects. Most
of the previous studies concentrated their work upon investigation of the aquifer
hydrogeological characteristics, but no studies were performed for determining
the expected areas suitable for the horizontal expansion of groundwater abstraction
projects. Additionally, it had been stated that water management is connected
with a series of difficulties such as insufficient information. To overcome
these difficulties, the present work was performed to fulfill this information
gap through building up a Geographic Information System (GIS) and running-up
binary-weighted spatial suitability models (BSSM-WSSM) on the constructed GIS
data layers. The BSSM-WSSM models help in determining the priority areas for
sustainable hydrogeological development and the possible regions available for
future horizontal expansion. The GIS is composed of superposed thematic multilayered
system of several decisive maps, which would comprehend in performing the objective
of the present work.
MATERIALS AND METHODS
East Uweinat study area occurs at the southwestern part of Egypt, where the
water bearing horizons of the NSAS is encountered at shallow depths. This is
the only part of this huge aquifer system where groundwater occurs under unconfined
conditions in an area where the Nubian sandstone crops out and is underlain
by shallow basement rocks; in this area groundwater has no thermal characteristics.
The aquifer system has a relatively high hydraulic conductivity and the preliminary
assessment of the groundwater resources has indicated that groundwater can be
extracted at a rate of 4.7x106 m3 day-1; the
long-term economics of extraction that can sustain large-scale development projects
(Nour, 1996). The direction of groundwater flow is generally
northeastwards but is distorted at faults and fracture zones.
Field and office work: In designing our field study, we considered differences
in groundwater consumption history and TDS variation. Accordingly, representative
well spots were chosen to reveal the regional spatial distribution of the selected
modeled parameters. The hydrogeological background data of the NSAS are undertaken
according to the previous published and non-published works and the recently
collected data from the inventory carried out for some wells in 2007-2008 (NARSS,
2008). Subsequently, laboratory and office works were conducted, which included
mapping and building-up of a multi-layered GIS for topographic elevations, geologic
units, basement relief, water chemical analyses and multi-temporal water levels
(Table 1-3). The recent well inventories
of 2007-2008 included the in situ measurement of pH value, Electric Conductivities
(EC), depth to water (mbgl) and collection of water samples for major cations
and anions determination and recording the geographic locations of wells by
Geographic Positioning Systems (GPS).
The ArcGIS Spatial Analyst extension of ArcGIS 9.1.1® software
(ESRI, 2007) was used to develop two spatial models Binary
Spatial Suitability Model (BSSM) and Weighted Spatial Suitability Model (WSSM)
using the constructed GIS thematic maps. The aquifer transmissivity (T) (m2
day-1) and hydraulic conductivity (K) (m day-1) were compiled
from the results of pumping tests performed by GPC (1984)
and GARPAD (1994). These values were plotted on the on
the landsat satellite ETM+ image and WSSM map to reveal their spatial distribution
and to validate the models results.
|| Data recorded during 2008 inventory
|| Data recorded from 1980s period
|| Data recorded from 1990s period
The logic steps of suitability spatial modeling could be summarized as:
||Defining the parameters of prioritization (or goals)
||Decide on evaluation criteria
||Define weights for criteria
||Calculate ranking model results
||Mapping and results evaluation
The priority areas for hydrogeological development and areas suggested for further horizontal expansion should address the following main input criteria:
||Aquifer saturated thickness (m) is adequately enough for the
long-term consumption policies
||Depth to water (mbgl) is reasonably shallow and easy to be economically
||Water total dissolved solids (ppm) are satisfactory low for domestic and
||Water Sodium Adsorption Ratio (SAR) is satisfactory low for irrigation
|| Flowchart of methodology
The flowchart of methodology is illustrated in Fig. 5.
One of the most common functions in vector GIS spatial analysis is the classification
function. This is used to transform a relatively complex set of vector values
to a simpler one. In many respects, this is like using lookup tables in raster
classifications, but it accommodates the fact that there is actually only one
attribute in a grid that can be used for analysis (Talbot,
1998; Mitchell, 1999; Malczewski,
1999). A range of suitable criteria was chosen and used for classification
to create a new grid where all of grid cells are 1 and the rest are 0. In the
present work, binary data (presence/absence, spatial distribution of conditional
criteria of suitability) are generally assessed using logistic regression methods
(Collett, 1991; Miska and Jan, 2005).
Logistic regression is a form of Generalized Linear Model (GLM) in which the
relationship is expressed as a probability surface, the expected error structure
is binomial and a logic transform (logistic link) is applied to the data (Trexler
and Travis, 1993; Sokal and Rohlf, 1995; Miska
and Jan, 2005). The logistic link means that the probability of obtaining
a positive response (meeting the prioritization criteria) is a logistic, s-shaped
function when the linear predictor is a first-order polynomial and for second-order
polynomials will approximate a bell-shaped function (Crawley,
In logistic regression, the binary nature of the response variable variation
is the basis of parameter estimation and thus the logistic regression models
will not produce inappropriate values (π(X)>1 or π (X)<0) for
the probability of presence.
||Ranks and weights for data layers and their influencing classes
used for groundwater prioritization mapping
Logistic regression has the form (Hosmer and Lemeshow, 1989)
where, α is the constant and β is the coefficient of the respective independent variables. The probability of presence π (ranging from 0 to 1) is given as a function of the vector of this model and becomes apparent after the logistic transformation, giving the form Eq. 2:
where, (ln) denotes to the natural logarithm (Sokal and Rohlf,
1995). A more technical and detailed review of logistic regression is presented
by McCullagh and Nelder (1989) and Collett
On the other hand, the data manipulation in a Weighted Spatial Suitability
Model (WSSM) implies the integration of all thematic layers within the WSSM.
However, it was assumed that all these layers have the same magnitude of contribution
on the process of suitability determination. Accordingly, the previously mentioned
priority factors are assigned equal weights of 25% but with different rates
(ranks) and degrees of effectiveness (Table 4). Also some
factors work negatively while others work positively in groundwater prioritization,
like that in water TDS, depth to water, water SAR, which work negatively in
the mapping process, whereas the aquifer saturated thickness works positively
in such task. For these reasons, each layer was assigned a specific weight of
effect on prioritization mapping. The given weights were adopted, depending
on the field observations, the literatures (Akther et
al., 2009; Peuquet, 1986; Malczewski,
1999) and the authors experience. Therefore, the integrated layers
in this study were given the weights (Wf), average rates (Rf)
and degree of effectiveness (E) of each GIS data layer, as shown in Table
4. The degree of effectiveness was obtained for each priority class according
the following equation Eq. 3:
RESULTS AND DISCUSSION
The reported historical data were used to elaborate the change in potentiometric levels, Total Dissolved Solids (TDS) and important hydrogeological characteristics since the implementation of developmental activities in the study area since 1980.
Groundwater movement and potentiometric surfaces: The delineation of
groundwater movement was performed GIS data layers of old and recent records
of potentiometric levels since the commencement of groundwater well drilling
and development activities in the mid-eighties (Fig. 6). These
GIS thematic layers constitute the baseline for the assessment of the current
status, anticipating changes and forecasting trends in groundwater quantity
and quality due to the natural and anthropogenic impacts in time and space.
The constructed flow nets clearly show the presence of a change in groundwater
movement, regionally and locally. The direction of groundwater movement during
eighties period (at the early stage of groundwater abstraction) was generally
from the south and the northwest to the east (Fig. 6a).
||Potentiometric levels data through: (a) Eighties period, (b)
Nineties period and (c) 2008
|| Structure contour map on top of basement rocks (according
to the data of drilled wells)
In nineties period, there was little change where there is a limited movement
of groundwater to the west in the southern part of the study area (Fig.
6b). But the dramatic change in the groundwater movement was revealed in
the map of 2008, where the potentiometric low area in the middle of the region
(bounded by contour 220 (masl)) was formed as a result of the long-term consumption
period and the widespread increase in the number of drilled wells. This long-term
over drafting led to a drastic depression in potentiometric levels and a reversal
in groundwater movement from south and north towards the central part of the
mapped area was established (Fig. 6c).
In other words, the maps revealed a certain decline in aquifer water levels, where general stabilized levels prevailed in eighties and nineties periods but with a progressive drawdown in groundwater potentiometric levels from nineties to 2008. Low consumption rates with scarce activities of water well drilling were the reasons behind the slight decline in eighties - nineties periods. The potentiometric levels of the recorded data of the old developmental area during eighties period was about 250 masl, depleted to about 245 masl in nineties period, whereas in 2008, this level became about 230 masl (Fig. 6).
Aquifer saturated thickness: The Six Hills Sandstone Formation represents
the sole water-bearing unit in the area, which belongs to the Pre-Aptian age
(Said, 1990) (Fig. 4). The presence
of semi-permeable layers (clay and siltstone) is discontinuous due to the rapid
lateral facies changes. The semi-permeable layers are hydraulically connected.
The Six Hills Formation acts as a water bearing bed of sandstone, which is named
commonly as the NSAS. The Six Hills Sandstone aquifer is hydraulically connected
with the underlying Precambrian fractured basement rocks, as a result of the
fracture system, which constitutes secondary porosity zones, but only for shallow
depths within the basement.
The estimated thicknesses of this aquifer in subsurface ranges from 186-706 m with an average thickness of about 446 m. From the present field measurements, the thickness variation of this aquifer is mainly due to the structural setting and Precambrian basement relief (Fig. 7), where it occurs at low elevations due south-southwest and northwest (-50-150 masl) which is represented by the effect of numerous fault systems with trends of NE-SW and NW-SE, which configured the basement relief. However, this basement relief variation controls the variation in NSAS saturated thickness, where a substantial increase in thickness is noticeable at the northwest and towards-southwest parts of the study area (Fig. 8), which confirms the previous clue given by the structure contour map constructed on top of basement rocks (Fig. 7).
|| Saturated thickness of water bearing six hills aquifer system
by GIS techniques
Groundwater total dissolved solids (TDS): Total dissolved solids (TDS)
are a measure of all constituents dissolved in water. The inorganic anions dissolved
in water include carbonates, chlorides, sulfates and nitrates. The inorganic
cations include sodium, potassium, calcium and magnesium. Thus, sulfate is a
constituent of TDS and may form salts with sodium, potassium, magnesium and
other cations (Table 3). As reflected from the data collected
in 2008 (Table 1), the salinity content in East Uweinat area
is mainly governed by the location of each well. It ranges between 281 mg L-1
(Wells 1 Army, 8/1) (analyzed in 2008) and 915 mg L-1 (Well S1) (Fig.
9). The variation in groundwater salinity from the year of 2002 until present
day was traced by the historical data, where salinities recorded in 2002 and
2008 were mapped (Fig. 9). Sources of TDS in groundwater in
the study area originate naturally from the dissolution of rocks and minerals
and can also be from the septic tanks and agricultural runoff resulted from
the reclamation activities in the last three decades. In the deeper horizons
of water bearing sediments, which occur at the northwestern, south and southwestern
parts of the study area, the basin-like troughs underlying these parts contain
groundwater entrapped in the deeper zones of the basin. These portions of the
basin typically contain denser water with higher TDS than the shallower zones
(Fig. 7, 8 and 9b-c).
Pumping shallow wells may draw up deeper poor quality water into the wells.
Accordingly, the noticeable trend is generally towards the relative salinization
from 2002 to 2008, which is due to water level depletion and withdrawal of deeper
low-quality water in addition to the seepage from the agricultural drainage
water (Fig. 9).
Priority areas for hydrogeological development by GIS binary and weighted
modeling: To investigate the spatial relationships (topological relationships)
of four prioritization or effective hydro-economical attributes that has their
own bearing on determining the priority areas for sustainable development and
future horizontal expansion, the model is logically based on the previously
discussed decision criteria. Topological notions include continuity, interior
and boundary, which are defined in terms of neighborhood relations (Egenhofer,
1993). If topological aspects have been part of previous and up-to-date
investigations, which constitutes the model input parameters or layers (i.e.
depth to groundwater, aquifer saturated thickness, groundwater total dissolved
solids and water Sodium Adsorption Ratio (SAR), the definitions of topological
relationships have been based upon, or mixed with, other concepts such as metric
(Peuquet, 1986) or order (Jungert,
1988; Chang et al., 1989; Lee
and Hsu, 1992). Consequently, the Binary model codes cells 1 for first priority
area, 0 for second priority area, or in other words, it emphasizes the topological
relationships by the intersections of boundary and interior of the modeled parameter
with the binary values meet and non meet of the prioritization mean value of
each input model parameter.
||Groundwater total dissolved solids, (a) 2002, (b) 2007 and
The model distinguished areas of first and second priority for the hydrogeological
development of the regional east Uweinat area (Fig. 10) as
a first step. Thus, the priority areas could be described as integrated roles
of these criteria. For a variety of reasons, these priority areas may be determined
to be of highest priority for environmental protection and development. For
immediate development of the groundwater, priority areas are determined to satisfy
the following conditions:
||Water quality segregations (TDS) for agricultural use is satisfactory
low (< 620 ppm for the 1st priority area and >620 ppm for the 2nd
||Economically optimum groundwater depth (< 27 mbgl for the 1st priority
area and >27 mbgl for the 2nd priority area)
||Sodium Adsorption Ratio (SAR) as a value determining the water suitability
for agriculture (< 2.5 for the 1st priority area and >2.5 for the
2nd priority area)
||Aquifer saturated thickness is big enough for the long-term consumption
(>280 m for the 1st priority area and <280 m for the 2nd priority
The running of the GIS binary spatial suitability modeling (BSSM) technique
elucidated two (1st and 2nd) priority areas (subclasses) for hydrogeological
development. The 1st priority area, which meets the predetermined criteria of
prioritization, occupies mostly the south-central and southwestern parts of
the study area with an area of 1,046 km2 which represents 5.12% of
the total study area.
||Determining priority areas for sustainable hydrogeological
development by BSSM technique, according to the data acquired in 2007-2008
The 1st priority area occupies mostly the intensively developed area characterized
by heavy consumption of groundwater and pivotal irrigation schemes, which reflects
the accuracy of the model results. However, most of the mapped area percentage
is covered by the 2nd priority area (94.88% of the total study area) (Fig.
In the second step, the spatial analysis was performed on these layers through a WSSM to discriminate more subdivisions for the 2nd priority area resulted from the BSSM, or, in other words, determining the suitable areas available for further horizontal expansion in reclamation activities (Fig. 11). Accordingly, as our aim is to identify the suitable areas for new horizontal expansions in development activities, so the 2nd priority area should be classified or subdivided into more classes using the weighted overlay GIS model. From this point, the BSSM technique was used at the early stage of the GIS spatial modeling. Subsequently, it was followed by the weighted overlay tool (ArcGIS Spatial Analyst Extension) using the same input criteria described previously, within the Raster Calculator of ArcGIS 9.3.1® software platform. Equal weights were assigned to each thematic input data layer (25% for each layer) (Table 4).
The WSSM output identified three more priority classes derived from the 2nd priority area, namely, the 2nd (7793 km2), 3rd (9976 km2) and 4th (1622 km2) priority classes, which represent more subdivisions for the 2nd priority area resulted from the BSSM. The geographic locations of the 2nd priority class represent new proposed suitable areas for development far from the old reclaimed ones (Fig. 11).
To validate the BSSM-WSSM results, some values of the Transmissivity (T) and
hydraulic conductivity (K) has been collected (GPC, 1984;
GARPAD, 1994) and plotted on the WSSM map (Fig.
11). It was found that the spatial distribution of these values is consistent
with the distribution of proposed priority areas for horizontal expansion. Accordingly,
the 2nd and 3rd subclasses are characterized by high T values (i.e., 1533 m2
day-1 in Well S14; 2225 m2 day-1 in Well Prod-2;
3392 m2 day-1 in Prod-4-site 4). These priority subclasses
are suitable for future horizontal expansion and are characterized by reasonable
aquifer saturated thickness (i.e., 350, 150, 230 and 425 m, respectively). However,
the WSSM also pinpointed to reasonable TDS values in these areas (i.e., 450,
600, 550 and 625 ppm, respectively.
Special attention is given to connectivity of the boundaries of these priority areas to preserve the hydrogeological characteristics of this aquifer system. The results of WSSM will guide planners to maintain these hydro-environmentally sensitive areas. The model predicts areas that can economically accommodate the future sustainable development and horizontal expansion of East Uweinat area.
||Determining priority areas for sustainable hydrogeological
development by WSSM techniques, according to data acquired in 2007-2008
The Nubian Sandstone is a significant aquifer system that is used to supply water for agricultural and domestic purposes in East Uweinat area. The present work aims to determine the hydrogeological priority areas of development suitable for further horizontal expansion. For such a critical purpose, the hydrogeochemical data of the most economically effective attributes are analyzed and digitized within a multilayered GIS. The attributes that had been taken into consideration are depth to water, aquifer saturated thickness, water TDS and SAR. The database of recorded water levels included those measured in the 1980s, 1990s and 2008, which showed that the potentiometric levels had a general noticeable decline from 1980s through 2008. The upgradable constructed GIS and binary-weighted spatial suitability models (BSSM-WSSM) were performed for determining the priority areas for sustainable hydrogeological development and the possible areas available for future horizontal expansion of reclamation projects. The BSSM models output map elucidated that the 1st priority area comprises mostly the intensively developed old reclaimed area that is characterized by a big number of production wells and consequently, heavy consumption of groundwater. Most of the study area was categorized as the 2nd priority area (94.88%), which designates to a lesser aquifer quality and characteristics in terms of the given prioritization criteria. For more detailed classification, the WSSM identified three more priority subclasses subdivided from the 2nd priority area, namely, the 2nd (7,793 km2), 3rd (9,976 km2) and 4th (1,622 km2) priority subclasses. The geographic locations of the 2nd priority class represent new proposed development areas relatively far from the present crowded ones, which could relieve the high pressure exerted upon the aquifer system in such areas. However, additional expansion in reclamation projects due north and east directions needs additional researches to assess the exact responses of the aquifer to the overgrowing withdrawal policies. Further accelerated drawdown in future is highly expected if continuing the present-day heavy consumption rates in the same present day consumption areas or even if the reclamation and agricultural projects are expanded without control on the kinds of crops to be planted.
As a consequence of the growing investments and populations, many new villages
and communities had been evolved in East Uweinat area, which led to an urgent
need for a well-developed water economy. The water economy is characterized
by limited opportunities for new water impoundments, rising incremental water
supply costs, intensified competition between diverse users and increased interdependencies
among water users. Also there is an urgent need for providing information for
improvements in the planning, policy and management of groundwater resources.
Furthermore, groundwater in East Uweinat developmental area is mostly found
under unconfined (water table) conditions thereby is liable to contamination
by the agrochemicals and domestic waste water leaks.
The author acknowledge the National Authority for Remote Sensing and Space Sciences (NARSS) for funding an internal research project from which, the data of the present work was derived.