Spatiotemporal distributions of water consumption for various land use-cover types over the Eastern province of Saudi Arabia were estimated using Surface Energy Balance Algorithm. Water consumption of various land use and cover classes shows similar seasonal dynamic trends. The spatial distribution of annual actual evapotranspiration (AET) shows low values in the Empty Quarter (231–438 mm/yr), and moderate values in the Eastern Province borders (439–731 mm/yr). Very high AET values were observed in irrigated croplands in the Northern plains, Hafar Al-Batin, the central coastal lowlands, and the southern coastal lowlands, where annual AET ranged from 732 to 1,790 mm/yr, representing the majority of the study area agricultural land. Evaporative behavior of land use-cover types indicated that irrigated cropland, which occupies 0.37% of the study area, has an average daily AET ranging from 9.2 mm/day in January to a maximum value in April (30 mm/day). Average annual water use by irrigated cropland is relatively very high and it is roughly 1,786.9 mm/yr, while water bodies, which cover 0.023% (121.2 km2) of the study area, also had relatively high mean AET (660.8 mm/yr). Overall, AET rates for irrigated cropland are much higher than for other land uses.

INTRODUCTION

Fundamental information on irrigation practices and groundwater resources is perceived as a missing part for building the water sector strategy of the Kingdom of Saudi Arabia (KSA). Water is declining rapidly in rural and urban areas, generally because of the growth in agricultural requirements (Mahmoud & Alazba 2016). Evapotranspiration (ET) constitutes a major part of the hydrological cycle, and is considered to be a means of exploiting irrigation water (Papadavid et al. 2011). For the agricultural production, water is a fundamental component, as well as critical for future global food security. During the past few decades, fast growing world population, urbanization, and industrialization caused depletion of fresh water resources. Agricultural water demand in Saudi Arabia comprised about 83–90% of the total water demands during 1990–2009 (Chowdhury & Al-Zahrani 2015). In order to better assess and manage irrigation water, estimating the amount of water consumed for agriculture-related activity is very important because it is a primary consumer of water. Since a large amount of irrigated water is lost through ET, its accurate estimation can be helpful for efficient management of irrigation water. Improved irrigation water management requires accurate scheduling of irrigations, which in turn requires an accurate calculation of daily ET (Bausch 1995). Determination of actual evapotranspiration (AET) from remote sensing data is based on the assessment of the surface energy balance components with the support of limited meteorological data (Ayenew 2003). Over the past decades, a large fraction of hydrological research activities has been directed at accurately estimating ET rates using various spatial and temporal scales. One of the most significant developments in the area of remote sensing hydrology was the determination of distributed a real AET from satellite data, based on the energy balance approach (Menenti 1984; Parodi 1993; Bastiaanssen 1995; Bastiaanssen et al. 1998a).

The surface energy balance algorithm for land (SEBAL) method (Bastiaanssen 1995) enables the calculation of actual and potential ET during the daytime of the satellite data recording. The method calls for complex procedures and determination of a number of variables from the different spectral bands of the satellite data, such as surface kinetic temperature, Normalized Difference Vegetation Index (NDVI), emissivity and albedo. Bastiaanssen (1995) described the method in detail. The widely adopted SEBAL model was generally utilized in farming areas (Bastiaanssen et al. 2005), where the flat landscape minimizes the spatial heterogeneities of the energy fluxes if compared with the effects of morphology on meteorological variables. For these agricultural contexts, numerous applications can be found in the literature (e.g., Bastiaanssen 2000; Minacapilli et al. 2009), chiefly focusing on the establishment of model outputs at the platform overpass time via simultaneous micro-meteorological measurements. Moreover, some applications were carried to regional (e.g., Sánchez et al. 2008) or basin (e.g., Teixeira et al. 2009) scales, by altering the assessment of instantaneous fluxes to account for the effects of complex morphologies.

In recent years, an increasing number of satellite algorithms based on the surface energy balance equation have been developed to estimate actually a real averaged ET by incorporating remotely sensed variables such as land surface temperature (LST), surface albedo, vegetation cover fraction, and land cover (LC) type in conjunction with field observations (Bastiaanssen 2000; Boronina et al. 2005; Senay et al. 2007; Wang et al. 2007; Singh et al. 2008; Elhaddad et al. 2010; Liu et al. 2010). Wang et al. (2007) proposed a simple regression equation to estimate ET using surface net radiation, air, surface temperatures, vegetation indices and ground-based measurements available at eight enhanced surface facility sites located throughout the Southern Great Plains area of the United States. Their work concluded that AET can be reasonably predicted using a correlation coefficient that varies from 0.84 to 0.95 under a wide range of soil moisture conditions and LC types. More recently, Singh et al. (2008) conducted a study to assess the operational characteristics and performing large-scale mapping of spatial distributions and seasonal variation in south-central Nebraska climates. The outcomes of their work showed that the model was capable of predicting growing season cumulative daily corn ET reasonably well within 5% of the measured values, and they concluded that SEBAL can be a feasible instrument for generating ETc maps to assess and quantify the spatiotemporal distribution of ET on large scales as well as estimating surface energy fluxes. Elhaddad et al. (2010) used the energy balance-based model to estimate the spatial and temporal variability of AET in the Arkansas River Basin in Colorado, South Platte River Basin in Colorado, and Palo Verde Irrigation District in California along with a 1-day ET estimate for Texas. Another study was conducted by Liu et al. (2010) to estimate AET and determine the variation with regard to varying types of land use and LC in urban settings. They implemented the SEBAL model to estimate AET at a higher spatial resolution using Landsat 5 satellite images and field observations. Their results revealed that the lowest observed AET was for developed urban areas and the highest for open water bodies. The AET difference is also presented from two contrasting counties. The results also indicated that agricultural fields have higher AET values than those of urban settings for all LC types except open water bodies.

The SEBAL model has been tested under several irrigation conditions in Egypt, India, Sri Lanka, Pakistan, Afghanistan, Argentina, Turkey, Niger, China, Brazil, United States, Sudan and Spain to diagnose the uniformity in crop consumptive use, crop water stress and irrigation performance (Wang et al. 1995; Bastiaanssen et al. 1998a, 1998b; Bastiaanssen 2000; Senay et al. 2007; Bastiaanssen et al. 2012; Wu et al. 2015). The first attempt to assess actual irrigation applications based on remotely sensed ET observations was presented by Ramos et al. (2006). The authors used the SEBAL model to assess the AET and to compute net water volumes and net irrigation volumes by introducing a water application efficiency factor. They reported that it is possible to assess actual irrigation applications based on AET observations. A similar study was conducted by Senay et al. (2007) using a simplified surface energy balance model to monitor and measure the performance of irrigated farming in Afghanistan. They used a combination of 1-km thermal data and 250-m NDVI data, both from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor. Seasonal AET over a period of 6 years for two major irrigated basins was computed by analyzing up to 19 cloud-free thermal and NDVI images from each year. Senay et al. (2007) reported that one of the major advantages of the energy-balance approach is that it can be applied to measure the spatial extent of irrigated areas and their water-use dynamics independently of the source of water, as opposed to the water balance model, which requires knowledge of both the magnitude and temporal distribution of rain and irrigation applied to fields.

In the present study, the spatial and temporal distributions of reference and AET over the Eastern province of Saudi Arabia during 1992–2013 were estimated using a surface energy balance based algorithm. The zonal statistics for each sort of land-cover type was also carried out, in order to interpret the water consumptions of different land use-cover types.

MATERIAL AND METHODS

Description of the study area

The Eastern Province (Figure 1) is the largest province of Saudi Arabia and it occupies an area of 518,300 km2. It borders the Persian Gulf, which contains the province's only maritime boundary and borders five countries on earth. Aside from this watery border, the majority of the states that the Eastern Province borders are on solid ground, and those are Iraq, Kuwait, Oman, Qatar, and the United Arab Emirates. The largely uninhabited Rub' al Khali (Empty Quarter) desert occupies more than half of the region. Geologically, Tertiary sediments characterize the whole district, with significant areas covered by eolian sands and late Tertiary to Quaternary gravel sheets. The entire region is divided into eight subregions. They are Northern plains, the central coastal lowlands, the southern coastal lowlands, northern summan, southern summan, northern Dahna, southern Dahna and the Rub al-Khali. The Northern plain is a linear depression bordering Hafar Al-Batin. It is a large, lifeless, flat plain rising very gradually from sea level in the east to about 400-m elevation in the west. The region is magnificent in many respects, being the largest continuous sand sheet of the world. Some of the sand dunes in this part of the desert are as high as 250 m. The clime of the Eastern Saudi Arabia is mostly arid, with temperature rising from 15 °C in January to a maximum of about 42 °C in August-September period. Annual rainfall means range from around 113–275 mm/yr in the north and northwest to less than 30 mm in the Empty Quarter (Figure 2). The vegetation of the Eastern regions is generally sparse. Annual plants, which comprise approximately 60% of the vegetation cover, are dominating in wadis and other shallow depressions.
Figure 1

Location map of the study area.

Figure 1

Location map of the study area.

Figure 2

Mean annual rainfall (1992–2013).

Figure 2

Mean annual rainfall (1992–2013).

Reference evapotranspiration

Climatic data was obtained from the meteorological department, Ministry of Agriculture and Ministry of Water and Electricity. Reference evapotranspiration is the ET from a hypothetical surface. In the case of the United Nations Food and Agriculture Organization (FAO) method (Monteith 1965), this surface consists of a grass reference crop of 0.12-m in height that is actively growing and adequately watered. It should have a surface roughness of 70 (s/m) and an albedo of 0.23. The latent heat flux of this reference surface is estimated according to the Penman–Monteith equation. In this study, 63 years of daily climate data (1992–2013) collected from weather stations located across the study region were utilized. The Penman–Monteith method (Monteith 1965) was used for estimating ET0 on a grid-by-grid basis using Equation (1). 
formula
1
where
  • ETo = reference evapotranspiration (mm/day)

  • Rn = net radiation at the crop surface (MJ/m2/day)

  • G = soil heat flux density (MJ/m2/day)

  • T = mean daily air temperature at 2 m height (oC)

  • u2 = wind speed at 2 m height (m/s)

  • es = saturation vapor pressure (kPa)

  • ea = actual vapor pressure (kPa)

  • Δ = slope vapor pressure curve (kPa/oC)

  • γ = psychrometric constant (kPa/oC)

  • es-ea = saturation vapor pressure deficit (kPa).

The data include elevation, latitude and longitude of the station, total monthly rain, maximum and minimum air temperature, relative humidity, wind speed at 2-m height, and daily sunshine hours. Penman–Monteith calculations were implemented within the main program of GIS prior to calculating ET. The inverse distance weighted (IDW) interpolation function was utilized to interpolate the values of the input data for each Julian day throughout the year.

The SEBAL model

AET retrieval, either by means of remote sensing-based approaches or by micro-meteorological measurements, is based on the application of the surface energy balance equation to the soil–plant–atmosphere system. SEBAL is one of the residual methods of energy budget, developed by (Bastiaanssen et al. 1998a). It combines empirical and physical parameterization. The remote sensing data used in this study included MODIS atmospheric profile data (MOD07). Near surface temperature values were extracted at the pressure level closest to the ground-surface described by the region's digital elevation model (DEM). Daily data were averaged to generate monthly averages. Data were further improved by relating MODIS-Ts to climate-station temperatures using MODIS LST data (MOD11A2). Emissivity values were derived by averaging MODIS-bands 31 and 32 emissivities, NDVI (MOD13Q1), and MODIS products combined with BRDF-albedo products (MCD43B3). Shuttle Radar Topographic Mission DEM (gap-filled) were obtained from the consortium of spatial data and information. The inputs also included local weather data and satellite data. From the input data, the (net solar radiation), , albedo, roughness length, and G (soil heat flux) were calculated. The Solar Analyst tool in ESRI's ArcGIS-environment was used to model solar radiation (Fu & Rich 1999). This tool is capable of generating both point and area-integrated estimates of solar radiation for variable time intervals. This balance, under the hypothesis that advection, energy required for photosynthesis, and canopy and air storages are negligible, can be written as: 
formula
2
where Rn (W/m2) is the net radiation, G (W/m2) is the soil heat flux at the surface, H (W/m2) is the sensible heat flux and (W/m2) is the latent heat flux, which represents the energy amount used for the surface evaporation and plant transpiration processes. The instantaneous surface energy balance (Equation (2)) at satellite overpass time is solved for by applying a residual approach. In the framework of SEBAL modeling, the net radiation, Rn, is computed based on the balance between downwelling and upwelling shortwave and long-wave radiation: the equation to calculate the net radiation flux is given by 
formula
3
where Rn is the net radiation at the surface, is the surface albedo, is the incoming short-radiation, is the incoming long-wave radiance, is the outgoing long-wave radiation, and is the surface emissivity. Albedo is the ratio of reflected light to total incident sunlight for a given area of the land surface. It is a fundamental property controlling the energy flux at the surface of the Earth; it can provide information on the biophysical characteristics of the land surface, including the structure of vegetation canopies, soil moisture, and urbanization. Broadband albedos are calculated using spectral-to-broadband conversion algorithms. The MODIS albedos represent the best quality data possible for each 16-day period (Zhou et al. 2003). Wanner et al. (1997) provided a detailed description of the MODIS albedo product, its creation, and its validation. These products were developed based on atmospherically corrected, cloud-cleared reflectance observations from the MODIS sensors on NASA's Aqua and Terra satellites.
DEM-based methods to estimate net shortwave radiation prove sound for clear sky conditions (Ruiz-Arias et al. 2009). In the present study, we used the Solar Analyst tool in ESRI's ArcGIS-environment to model solar radiation. The tool is capable of generating both point and area-integrated estimates of solar radiation for variable time intervals. 
formula
4
where is the cosine of the solar incidence angle, d is the inverse squared relative earth–sun distance, and is the atmospheric transmissivity. The atmospheric transmissivity values were determined based on clear sky and relatively dry conditions obtained using the elevation-based relationship (Allen et al. 1998), where Z is the elevation of the weather station above sea level (m).
The incoming long-wave radiance and the outgoing long-wave radiation are derived from an application of the Stefan–Boltzmann equation and MODIS-based Ts and Ta as input, i.e. 
formula
5
 
formula
6
where is the mean surface temperature (K), which is obtained from the satellite data. Daily Ts data were averaged to generate monthly averages. Data were further improved by relating MODIS-Ts to climate-station temperatures.
is the Stefan Boltzmann constant (5.67 × 10−8 W/m2K2); is the broad-band surface emissivity; the surface emissivity was computed using an empirical equation where the NDVI > 0 by inputting the NDVI and the LAI as described in the SEBAL user manual; is the emissivity of the atmosphere; emissivity of the atmosphere was calculated from Equation (8). 
formula
7
Substituting (8) into (7) provides the following Equation (8): 
formula
8
Soil heat flux cannot be directly determined from satellite sensors and requires empirical formulation. Remote sensing derivable parameters that influence soil heat flux are used; these include NDVI, surface temperature and albedo. The soil heat flux (G) has been estimated using the following relation developed by Bastiaanssen (1995): 
formula
9
Sensible heat flux is the rate of heat loss to the air by convection and conduction, due to a temperature difference. It was computed using the following equation for heat transport: 
formula
10
where is the air density (kg/m3), is the specific heat of air (1,004 J/kg K), is the near surface temperature difference (K), and is the aerodynamic resistance to heat transport (s/m), where 
formula
11
The aerodynamic resistance to heat transport (rah) is computed for neutral stability as: 
formula
12
where Z1 is a height just above the zero displacement distance height of the plant canopy set to 0.1 m for each pixel, and Z2 is the reference height just above the plant canopy set to 2 m for each pixel, u* is the friction velocity (m/s), and k is the von Karman constant (0.4). 
formula
13
Surfaces of wind-velocity were generated by interpolating climate-station recorded wind velocities. The IDW method was used to interpolate wind velocities recorded 2 m above the ground surface. Elevation differences were used as weights in the interpolation of wind velocity. The initial value for friction velocity (u*) was computed for neutral stability with data from the local meteorological station. The average height of vegetation (m) was surveyed in the site nearby the meteorological station and was used to calculate the land surface friction, Zm. Then, Zm for each pixel was calculated by a regression equation according to the pixel's NDVI value. The near-surface wind speed was converted to a value at the blending height (200-m) where the effects of the land surface roughness could be eliminated. The initial estimated values of u* were used to infer the first values of aerodynamic resistance (rah). Corrections for atmosphere stability were obtained iteratively for each pixel. This required a series of iterations to determine new values of the corrected friction velocities and the corrected aerodynamic resistances before obtaining numerical stability based on the criterion of Monin–Obukhov length.
Once sensible heat flux and the available energy (Rn-G) were determined, estimation of instantaneous could be computed pixel by pixel using Equation (2). The evaporative fraction (EF) describes the partitioning of the surface energy balance as the latent heat flux/net available energy, with the net available energy being defined as the difference in net radiation and soil heat flux. In this study, we used the concept ‘ETo fraction’ (ETof), which represents the ratio of ET of each pixel to the reference ET as computed by the Penman–Monteith method. ETof is the same as the crop coefficient, Kc, and was calculated using Equation (14). Then, daily values of ET were computed using Equation (15). 
formula
14
 
formula
15
The computation of the monthly AET involved extrapolating the SEBAL daily AET, within each month, proportional to the reference ET, where the latter was derived from standard meteorological measurements. The procedure involved determining the cumulative reference ET between successive satellite images before computing the ratio (Km) of the cumulative reference ET to the average potential ET over the period. The monthly ET (AETm) was then computed as follows: 
formula
16

RESULTS AND DISCUSSION

Developing LC and reference ET

A Landsat 5/7 TM/ETM image was obtained in the year 2013 from the King Abdul-Aziz City for Science and Technology. This image was incorporated with collected data from the specified region, and ultimately utilized in categorizing LC. ERDAS Imagine software was used to mosaic the collected satellite images. Geometric corrections were made using an already rectified SPOT-5 ‘2.5 m PSM’ satellite image of 06/03/2000. Iso Cluster unsupervised classification and Maximum likelihood classification methods were used for the unsupervised and supervised classification. Training samples were collected during field surveys to create spectral signatures for the supervised classification to identify what the cluster represents. The LC map was classified into six main classes: irrigated cropland, shrubland, sparse vegetation, built-up areas, bare soil and water bodies. The extent of the LC distribution in 2013 is presented in Figure 3. The areal extent of each class and its LC are presented in Table 1. The Eastern Province is the largest province of the KSA and it occupies 24.1% of its total area. Results revealed that bare soil represents the largest portion of the area (99.45%), and only 0.37% of the total area is ‘irrigated cropland’, while water bodies, sparse vegetation and shrubland represent 0.023%, 0.037% and 0.0127% of the total area, respectively. Built-up area accounts for about 494 km2 or 0.1% of the total province area (Table 1).
Figure 3

LC map of the study area.

Figure 3

LC map of the study area.

Table 1

Areal distribution of different LC classes

LC classArea (%)Area (km2)
Irrigated cropland 0.37 1,910 
Shrubland 0.0127 66 
Sparse vegetation 0.037 191.8 
Built-up areas 0.1 494 
Bare areas 99.45 515,517 
Water bodies 0.023 121.2 
LC classArea (%)Area (km2)
Irrigated cropland 0.37 1,910 
Shrubland 0.0127 66 
Sparse vegetation 0.037 191.8 
Built-up areas 0.1 494 
Bare areas 99.45 515,517 
Water bodies 0.023 121.2 
In the present study, various meteorological data, including normal maximum temperature, minimum temperature, sunshine hours, relative humidity, wind speed and rainfall, have been obtained from the meteorological department, Ministry of Agriculture and Ministry of Water and Electricity. The FAO Penman–Monteith method was employed to model the spatial distribution of ETo in the region. The IDW interpolation function was used to interpolate each month, with the result being 12 separate maps of mean monthly ETo (mm/day). The monthly distribution of ETo showed a gradual increase from January to July. On the other hand, there was a subsequent decline in monthly ETo values from August to December (Figure 4). The lowest values were in January, when ETo ranged from 2.46 to 3.56 mm/day in the study area. The highest values were observed in July, when ETo ranged from 6.67 to 12 mm/day. From March to October, ETo is found to be maximum in the Northern plains, Hafar Al-Batin, the central coastal lowlands, and the southern coastal lowlands. From December to February there is a shift of maximum ETo towards the Empty Quarter, which occupies more than half of the province, and the Eastern Province borders on land with Iraq, Kuwait, Oman, Qatar, and the United Arab Emirates. This may be due to the seasonal variation in the relative humidity and wind velocity in the study area. Mean annual ETo (1992–2013) ranged from 2,060 to 2,530 mm/yr. The spatial distributions of annual ETo are shown in Figure 5, which reflects a combined effect of all climatological factors. The lowest annual ETo values are located in the Empty Quarter and the Eastern Province borders with Iraq, Kuwait, Oman, Qatar, and the United Arab Emirates (2,060–2,340 mm/yr). The maximum values of ETo rates are observed mainly over the Northern plains, Hafar Al-Batin, the central coastal lowlands, and the southern coastal lowlands (2,350–2,530 mm/yr).
Figure 4

Spatial distribution of monthly reference evapotranspiration (mm/day).

Figure 4

Spatial distribution of monthly reference evapotranspiration (mm/day).

Figure 5

Spatial distribution of mean annual reference evapotranspiration.

Figure 5

Spatial distribution of mean annual reference evapotranspiration.

Spatial–temporal distribution of AET

Figure 6 shows average monthly AET of different land uses. The highest monthly values were in April (5.22–30 mm/day). The monthly distribution of AET showed a gradual increase from January to April. In January, daily AET ranged from 0 to 9.18 mm/day; in February, AET ranged from 1.6 to 14 mm/day; in March, daily AET ranged from 3.7 to 26 mm/day; while the highest observed value in April ranged from 5.2 to 30 mm/day. On the other hand, there was a subsequent decline in daily AET values from May to December. In May, daily AET ranged from 3.63 to 27.6 mm/day, while in June daily values varied from 2.1 to 13.7 mm/day, and in July and August daily AET ranged from 0.8 to 7.12 mm/day and from 0.59 to 4.5 mm/day, respectively. The decline in daily AET values is clearly seen from September to December, where the daily AET range dropped from 0.59–4.5 mm/day in August to 0–2.53 mm/day in October. There was a slight increase to 0–4.28 mm/day in November and 0–7.53 in December. Nevertheless, this increase is only in irrigated cropland. In general, water consumption of various land use/cover classes shows similar seasonal dynamic trends throughout the years. The value of AET started to rise rapidly in February, reached peak values in April, and then declined to the lowest levels in September and October, with a slight increase in November and December. The maximum monthly AET values in the study area were observed for irrigated cropland in the Northern plains, Hafar Al-Batin, the central coastal lowlands, and the southern coastal lowlands during the 12 months of the year, while the lowest values were observed in the Empty Quarter and the Eastern Province borders on land with Iraq, Kuwait, Oman, Qatar, and the United Arab Emirates. The actual monthly ET for all land uses follows the monthly rainfall pattern in the study area, as areas with high monthly and annual rainfall rate have high AET. This is clearly seen in the Empty Quarter, which receives annual rainfall less than 30 mm/yr, while the Northern plains, Hafar Al-Batin, the central coastal lowlands, and the southern coastal lowlands receive means annual rainfall range from around 113–275 mm/yr. This is in good agreement with Kongo et al. (2011), who found that the actual monthly ET for all land uses follows the monthly rainfall pattern.
Figure 6

Spatial distribution of average monthly AET (1992–2013).

Figure 6

Spatial distribution of average monthly AET (1992–2013).

The modeled average annual AET by SEBAL model is presented in Figure 7. The spatial distribution of annual AET shows low values in the Empty Quarter (231–438 mm/yr) and moderate values in the Eastern Province borders on land with Iraq, Kuwait, Oman, Qatar, and the United Arab Emirates (439–731 mm/yr), while it shows very high AET values for irrigated cropland in the Northern plains, Hafar Al-Batin, the central coastal lowlands, and the southern coastal lowlands, where annual AET ranged from 732 to 1,790 mm/yr, representing the majority of the study area agricultural land. Figure 7 shows that built-up areas and bare soil present low values of AET, while irrigated cropland shows the highest values. In general, AET rates for irrigated cropland are much higher than for other land uses. Shrubland and sparse vegetation are the next highest, with water bodies slightly lower, and very low AET rates from built-up areas and bare soil. This is in good agreement with Dunn & Mackay (1995), who found that irrigated croplands transpire at a slightly higher level than other vegetation, as they maintain a canopy cover throughout the year.
Figure 7

Estimated mean annual AET.

Figure 7

Estimated mean annual AET.

Estimating consumptive water use for agriculture in Saudi Arabia has always been a challenge for farmers, water managers, researchers and government agencies. Quantification of consumptive use allows more informed water use practices, consistent measures of the allocation among stakeholders, and improved planning. Daily and annual water consumption for different land use and cover classes were computed. A summary of the values obtained for each land use and cover types is given in Table 2. The evaporative behavior or consumptive water use of the different land use and cover types in the study area as inferred by the SEBAL model indicates that irrigated cropland, which occupies 0.37% of the study area, has an average daily AET that ranges from 9.2 mm/day in January to its peak in April (30 mm/day) and drops to 7.5 mm/day in December. Mean annual water consumption by irrigated cropland is relatively very high, about 1,786.9 mm/yr, while water bodies, which cover 0.023% (121.2 km2) of the study area, also had relatively high mean AET (660.8 mm/yr). However, they cover only a limited area. Evaporative behavior in the largest continuous sand sheet of the world has a similar trend, and it follows monthly rainfall pattern in the area. Bare soil, which represents 99.45% of the total area, had a low evaporative behavior, where the total annual AET is about 608 mm/yr.

Table 2

Monthly and annual AET of different land use and cover classes

 Jan.Feb.Mar.Apr.MayJuneJulyAug.Sept.Oct.Nov.Dec.Mean annual AET
LC classAET (mm/day)AET (mm/season)
Irrigated cropland 9.2 14 26.1 30 27.6 13.7 7.1 4.5 2.4 2.5 4.3 7.5 1,786.9 
Shrubland 4.5 6.7 12.1 13.6 9.5 4.9 1.6 0.7 0.7 0.9 1.8 704 
Sparse vegetation 4.8 7.4 12.5 14.3 10.3 5.2 3.4 0.8 0.8 1.5 2.7 760.8 
Built-up areas 2.37 5.79 7.92 9.4 9.2 4.49 2.8 1.47 1.1 0.9 0.86 1.5 469 
Bare soil 3.6 3.94 6.72 12.8 7.63 3.94 2.25 1.17 0.5 0.6 0.5 608 
Water bodies 4.8 7.4 12.3 14.3 9.8 5.1 3.4 0.7 0.8 1.5 2.6 660.8 
 Jan.Feb.Mar.Apr.MayJuneJulyAug.Sept.Oct.Nov.Dec.Mean annual AET
LC classAET (mm/day)AET (mm/season)
Irrigated cropland 9.2 14 26.1 30 27.6 13.7 7.1 4.5 2.4 2.5 4.3 7.5 1,786.9 
Shrubland 4.5 6.7 12.1 13.6 9.5 4.9 1.6 0.7 0.7 0.9 1.8 704 
Sparse vegetation 4.8 7.4 12.5 14.3 10.3 5.2 3.4 0.8 0.8 1.5 2.7 760.8 
Built-up areas 2.37 5.79 7.92 9.4 9.2 4.49 2.8 1.47 1.1 0.9 0.86 1.5 469 
Bare soil 3.6 3.94 6.72 12.8 7.63 3.94 2.25 1.17 0.5 0.6 0.5 608 
Water bodies 4.8 7.4 12.3 14.3 9.8 5.1 3.4 0.7 0.8 1.5 2.6 660.8 

CONCLUSION

This study was designed to model spatial and temporal distribution of reference and AET over the Eastern province of Saudi Arabia during 1992–2013 using the Surface Energy Balance based algorithm. The monthly distribution of reference evapotranspiration (ETo) showed a gradual increase from January to July and a subsequent decline in monthly ETo values from August to December, while the monthly distribution of AET showed a gradual increase from January to April. On the other hand, there was a subsequent decline in daily AET values from May to December. Water consumption of various land use/cover classes shows similar seasonal dynamic trends throughout the years. The maximum monthly AET values in the study area were observed for irrigated cropland in the Northern plains, Hafar Al-Batin, the central coastal lowlands, and the southern coastal lowlands during the 12 months of the year, while the lowest values were observed in the Empty Quarter and the Eastern Province borders on land with Iraq, Kuwait, Oman, Qatar, and the United Arab Emirates. The spatial distribution of annual AET shows low values in the Empty Quarter (231–438 mm/yr) and moderate values in the Eastern Province borders on land with Iraq, Kuwait, Oman, Qatar, and the United Arab Emirates (439–731 mm/yr), while it shows very high AET values for irrigated cropland in the Northern plains, Hafar Al-Batin, the central coastal lowlands, and the southern coastal lowlands, where annual AET ranged from 732 to 1,790 mm/yr, representing the majority of the study area agricultural land. Overall, AET rates for irrigated cropland are much higher than for other land uses. Shrubland and sparse vegetation are the next highest, with water bodies slightly lower, and very low AET rates from built-up areas and bare soil. Mean annual water consumption by irrigated cropland is relatively very, about 1,786.9 mm/yr, while water bodies, which cover 0.023% (121.2 km2) of the study area, also had relatively high mean AET (660.8 mm/yr). Evaporative behavior in the largest continuous sand sheet of the world has a similar trend, and it follows monthly rainfall pattern in the area. Bare soil, which represents 99.45% of the total area, had a low evaporative behavior, where the total annual AET is about 608 mm/yr. The successful application of SEBAL for estimating AET proves its value within the hydrological community in the study area and similar regions, and it can be used to help water managers devise sustainable water use planning and management in Saudi Arabia.

ACKNOWLEDGEMENTS

The project was financially supported by King Saud University, Vice Deanship of Research Chairs.

REFERENCES

REFERENCES
Allen
R. G.
Pereira
L. S.
Raes
D.
Smith
M.
1998
Crop evapotranspiration-guidelines for computing crop water requirements. FAO Irrigation and Drainage Paper 56
.
Food and Agriculture Organization, Rome
.
Bastiaanssen
W. G. M.
1995
Regionalization of Surface Flux Densities and Moisture Indicators in Composite Terrain, a Remote Sensing Approach Under Clear Skies in Mediterranean Climates, Report 109
.
Agricultural Department
,
Wageningen
,
The Netherlands
.
Bastiaanssen
W. G. M.
Menenti
M.
Feddes
R. A.
Holtslag
A. A. M.
1998a
A remote sensing surface energy balance algorithm for land (SEBAL): I. formulation
.
Journal of Hydrology
212
,
198
212
.
Bastiaanssen
W. G. M.
Pelgrum
H.
Wang
J.
Ma
Y.
Moreno
J. F.
Roerink
G. J.
Van der Wal
T.
1998b
A remote sensing surface energy balance algorithm for land (SEBAL): 2. validation
.
Journal of Hydrology
212
,
213
229
.
Bastiaanssen
W. G. M.
Noordman
E. J. M.
Pelgrum
H.
Davids
G.
Thoreson
B. P.
Allen
R. G.
2005
SEBAL model with remotely sensed data to improve water-resources management under actual field conditions
.
Journal of Irrigation and Drainage Engineering-ASCE
131
(
1
),
85
93
.
Bastiaanssen
W. G. M.
Cheema
M. J. M.
Immerzeel
W. W.
Miltenburg
I. J.
Pelgrum
H.
2012
Surface energy balance and actual evapotranspiration of the transboundary Indus Basin estimated from satellite measurements and the ETLook model
.
Water Resources Research
48
(
11
).
Chowdhury
S.
Al-Zahrani
M.
2015
Characterizing water resources and trends of sector wise water consumptions in Saudi Arabia
.
Journal of King Saud University-Engineering Sciences
27
(
1
),
68
82
.
Elhaddad
A.
Garcia
L. A.
Chávez
J. L.
2010
Using a surface energy balance model to calculate spatially distributed actual evapotranspiration
.
Journal of Irrigation and Drainage Engineering
137
(
1
),
17
26
.
Fu
P.
Rich
P. M.
1999
Design and implementation of the solar analyst: an ArcView extension for modeling solar radiation at landscape scales
. In:
Proceedings of the Nineteenth Annual ESRI User Conference
,
Chicago
, pp.
1
31
.
Kongo
M. V.
Jewitt
G. W. P.
Lorentz
S. A.
2011
Evaporative water use of different land uses in the upper-Thukela river basin assessed from satellite imagery
.
Agricultural Water Management
98
(
11
),
1727
1739
.
Liu
W.
Hong
Y.
Khan
S. I.
Huang
M.
Vieux
B.
Caliskan
S.
Grout
T.
2010
Actual evapotranspiration estimation for different land use and land cover in urban regions using Landsat 5 data
.
Journal of Applied Remote Sensing
4
(
1
),
041873
.
Menenti
M.
1984
Physical Aspects and Determination of Evaporation in Deserts Applying Remote Sensing Techniques, Report 10 (Especial Issue)
.
Institute of Land and Water Resources Management
,
Wageningen
,
The Netherlands
, p.
202
.
Minacapilli
M.
Agnese
C.
Blanda
F.
Cammalleri
C.
Ciraolo
G.
D'Urso
G.
Iovino
M.
Pumo
D.
Provenzano
G.
Rallo
G.
2009
Estimation of actual evapotranspiration of Mediterranean perennial crops by means of remote sensing based surface energy balance models
.
Hydrology and Earth System Sciences
13
(
7
),
1061
1074
.
Monteith
J. L.
1965
Evaporation and environment
.
Symposia of the Society for Experimental Biology
19
,
205
234
.
Papadavid
G.
Hadjimitsis
D.
Toulios
L.
Michaelides
S.
2011
Mapping potatoes crop height and LAI through vegetation indices using remote sensing, in Cyprus
.
Journal of Applied Remote Sensing
5
,
053526
.
Parodi
G.
1993
An Up-to-date Inventory of Remote Sensing Potentially in the Energy Balance Equation Approach for Actual Evapotranspiration Mapping
.
Basic Theory and Prospects. Doctoral dissertation, MSc thesis
,
ITC
,
Enschede
,
The Netherlands
.
Ramos
J. G.
Kay
J. A.
Cratchley
C. R.
Casterad
M. A.
Herrero
J.
López
R.
Martínez-Cob
A.
Domínguez
R.
2006
Crop management in a district within the Ebro River Basin using remote sensing techniques to estimate and map irrigation volumes
.
WIT Transactions on Ecology and the Environment
96
,
365
377
.
Ruiz-Arias
J. A.
Tovar-Pescador
J.
Pozo-Vázquez
D.
Alsamamra
H.
2009
A comparative analysis of DEM-based models to estimate the solar radiation in mountainous terrain
.
International Journal of Geographical Information Science
23
(
8
),
1049
1076
.
Sánchez
J. M.
Scavone
G.
Caselles
V.
Valor
E.
Copertino
V. A.
Telesca
V.
2008
Monitoring daily evapotranspiration at a regional scale from Landsat-TM and ETM + data: application to the Basilicata region
.
Journal of Hydrology
351
(
1
),
58
70
.
Singh
R. K.
Irmak
A.
Irmak
S.
Martin
D. L.
2008
Application of SEBAL model for mapping evapotranspiration and estimating surface energy fluxes in south-central Nebraska
.
Journal of Irrigation and Drainage Engineering
134
(
3
),
273
285
.
Wang
J.
Ma
Y.
Menenti
M.
Bastiaanssen
W. G. M.
Mitsuta
Y.
1995
The scaling-up of processes in the heterogeneous landscape of HEIFE with the aid of satellite remote sensing
.
Journal of the Meteorological Society of Japan
73
(
6
),
1234
1244
.
Wang
K.
Wang
P.
Li
Z.
Cribb
M.
Sparrow
M.
2007
A simple method to estimate actual evapotranspiration from a combination of net radiation, vegetation index, and temperature
.
Journal of Geophysical Research: Atmospheres (1984–2012)
112
(
D15
).
Wanner
W.
Strahler
A. H.
Hu
B.
Lewis
P.
Muller
J. P.
Li
X.
Schaaf
C. L.
Barnsley
M. J.
1997
Global retrieval of bidirectional reflectance and albedo over land from EOS MODIS and MISR data: theory and algorithm
.
Journal of Geophysical Research-Atmospheres
102
(
D14
),
17143
17161
.
Zhou
L.
Dickinson
R. E.
Tian
Y.
Zeng
X.
Dai
Y.
Yang
Z. L.
Schaaf
C. B.
Gao
F.
Jin
Y.
Strahler
A.
Myneni
R. B.
2003
Comparison of seasonal and spatial variations of albedos from moderate-resolution imaging spectroradiometer (MODIS) and common land model
.
Journal of Geophysical Research: Atmospheres (1984–2012)
108
(
D15
).