Arab J Geosci DOI 10.1007/s12517-013-1084-1
ORIGINAL PAPER
Contribution of shallow groundwater rapid fluctuation to soil salinization under arid and semiarid climate Mohamed Khaled Ibrahimi & Tsuyoshi Miyazaki & Taku Nishimura & Hiromi Imoto
Received: 2 March 2013 / Accepted: 22 August 2013 # Saudi Society for Geosciences 2013
Abstract Rising saline shallow groundwater and associated soil salinization problems are widespread especially in arid and semiarid areas. There have been numerous studies on groundwater-associated salinity, but more information is required on the effects of groundwater frequent and high fluctuations on soil salinization. In the present study, laboratory experiments and numerical simulations using HYDRUS-1D model were carried out for this purpose. The experimental and modeling results showed that groundwater fluctuation caused not only the accumulation of more salt in the soil profile compared to stable groundwater, but also an enhancement of the mechanism. Water table fluctuation induced a much greater spreading of the bromide (Br) tracer within the column than the constant water table. The Br content was on average five orders of magnitude greater under a fluctuating water table than under a constant one. Further, the numerical simulations showed that an increase in the groundwater fluctuation frequency brought about an increase in soil surface salinization under the same evaporation boundary conditions. Additional simulations with HYDRUS-1D were used to study the effects of various management strategies on soil salinization induced by shallow groundwater. Hence, by reducing the evaporation rate through the application of surface mulching, a significant reduction of salt concentration at the soil surface was observed. Moreover, frequent irrigations with small quantities were effective to reduce soil surface salt accumulation induced by saline shallow groundwater.
Keywords Groundwater fluctuation . Salt accumulation . Salinity management . HYDRUS-1D M. K. Ibrahimi (*) : T. Miyazaki : T. Nishimura : H. Imoto Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan e-mail:
[email protected]
Introduction “Global problem,” “white death,” “silent flood,” and “growing menace” are among the many slogans describing the hazard of soil salinization which is considered a great threat to the environment and sustainable agriculture, jeopardizing the well-being of many people over wide geographic areas. Despite considerable improvement in our knowledge of soil salinization and much progress in its management and control, the problem has not disappeared. According to recent reports, this phenomenon occurs within the boundaries of more than 100 countries around the world with a variety of extents, nature, and properties (Rengasamy 2006). For instance, Mohamed et al. (2012) found that soil salinization is one of the main causes of land degradation in the northeastern part of the Nile Delta, Egypt. Shirokova et al. (2000) reported that about 50 % of irrigated land in Central Asia is saline, 29 % of which has a strong to moderate degree of salinity. Although salt accumulation can occur under any climate, it is mainly found in arid and semiarid environments (e.g., Matinfar et al. 2011; Abdel-Kawy and Belal 2013). Indeed, it was stated by the Food and Agriculture Organization (FAO 2002) that the major saline regions of the world are generally found in semiarid and arid and relatively low-lying, poorly drained lands where the water table is at or near the soil surface. In these conditions, the fluctuation of saline shallow groundwater is the key process of salt accumulation at the soil surface. Soils underlain by shallow water tables have a high salinization risk due to capillary upflow transporting salts into the root zone (e.g., Al-Garni and El-Kaliouby 2011). Saline shallow groundwater contribution to soil salinization was usually attributed to capillary rise whenever the water table level reaches a critical depth (usually considered to be 1.5 m) below ground surface (e.g., Kara 2002; Northey et al. 2006). Hence, much of the research on soil salinization induced by shallow groundwater has been conducted with a
Arab J Geosci
fixed depth to the water table. Basically, at laboratory scale, the water table was maintained at a specified depth for the duration of the experiment (e.g., Gowing et al. 2006), and in field situations, where the water table depth cannot be controlled, the depth to water was characterized as an average depth (e.g., Werner and Lockington 2004). Less attention, however, has been given to the effect of rapid and frequent water table fluctuations which have been shown to occur in most of shallow unconfined aquifers in response to recharge events (O’Brien 1982; Cartwright et al. 2006; Ibrahimi et al. 2010). Neglecting the water table fluctuation in the assessment of soil salinization risk can lead to an underestimation of the salinity risk in shallow groundwater systems and consequently to ineffective salinity management in these areas. To prevent or at least control the phenomenon of shallow groundwater-associated salinity, appropriate management strategies are required such as reducing the potential evaporation by using surface mulching and improving irrigation management. The first technique is based on the fact that the intensity of the incoming radiation at soil surface can be reduced and consequently the evaporation rate. For instance, covering or mulching the soil surface with vapor barriers or with reflective materials can reduce the intensities of the incoming radiations and reduce the evaporation in the first stage of drying (Lal and Shukla 2004). In arid and semiarid regions, mulching is a common and effective practice to reduce evaporation artificially (Yamanaka et al. 2004). Yuan et al. (2009) found that the gravel mulches dramatically reduced the evaporation from bare soil surface, particularly when soil water contents were at high levels. Fekri and Kasmaei (2013) proved experimentally the effects of windy sand and light expanded clay aggregate mulches on decreasing evaporation from the soil surface. Similarly, Al-Dhuhli et al. (2010) concluded through field experiment that mulching with shredded date palm leaves proved as an effective management practice to conserve soil moisture, reduce salt accumulation in soil, and control increase in surface soil temperature. As for the irrigation management technique, it is used to control the salinity levels of agricultural lands by maintaining some downward movement of water and salts out of the root zone to the soil beneath (Kara 2002). This is generally achieved by adding a leaching fraction to the amount of irrigation water needed. For such a practice to be successful, an effective drainage system has to be installed in the irrigated area In the present study, laboratory experiments were specifically designed to simulate soil salinization under fluctuating and stable water tables using soil typically found in Metouia Oasis (South Tunisia). These laboratory investigations aimed to improve the understanding of salt accumulation phenomenon induced by saline shallow groundwater in unconfined aquifer systems. In addition, numerical simulations using the model HYDRUS-1D were carried out in order to investigate the effect of saline shallow groundwater fluctuations on soil
surface salt accumulation and to evaluate some preventive and curative management solutions.
Material and methods The soil The soil used in the experiments was collected from Metouia Oasis (33°58′00.12″N–10°00′02.55″E) located in Southeast Tunisia (Fig. 1). It is a 266-ha irrigated area which is increasingly being impacted by high soil salinity, leading to yield losses and ultimately to farmland being abandoned. In this area, shallow groundwater rise is considered to be the major factor responsible for salt accumulation in soils (Ibrahimi et al. 2010). Metouia Oasis has experienced dramatic change in the shallow groundwater table and soil salinization. The arid climate conditions prevailing there, characterized by high evapotranspiration reaching about 2,000 mm/year and low rainfall rate averaging 200 mm/year, have led to the application of excessive saline irrigation water (4.5 dS/m) for several years. This led to a rising saline shallow groundwater and consequently to soil surface salt accumulation. High salinity in the root zone caused reduction of crop productivity and limited their diversity. This led to a decrease in the economic viability of crop production and ultimately to the abandonment of many farmlands in the oasis. Soil samples taken from Metouia Oasis were brought to the laboratory for physical and chemical analysis. Table 1 summarizes the main physical and chemical properties of Metouia soil. The water retention curve of Metouia soil (Fig. 2) was determined using the “hanging method” for the low range of matric pressure head (0 to −200 cm) and a pressure plate system for higher range of suction (100 to 1,500 kPa). This curve was well fitted with the van Genuchten function (van Genuchten 1980). The hydraulic parameters of Metouia soil were optimized in the retention equation of van Genuchten (1980) using the RETC nonlinear optimization code (van Genuchten et al. 1991) from water retention measurements using soil samples. The van Genuchten parameters were the following: residual water content θ r =0.096, saturated water content θ s =0.418, α =0.016, n =2.015. The saturated hydraulic conductivity was measured in the laboratory using the falling-head hydraulic conductivity measurement method of Klute and Dirksen (1986). Soil samples were also subject to a thorough chemical analysis to determine the major ions present in the soil solution. The analysis showed that the dominant cations are in the order Ca2+ > Na+ > Mg2+ > K+, and the dominant anions are in the order SO42− > Cl− > HCO3−. In Metouia soil, sulfate, calcium, and chloride prevailed and constituted 1,800, 670, and 530 ppm, respectively. Based on the chloride-to-sulfate ratio (Cl−/SO42−), the soil in Metouia has a sulfate–chloride salinity type.
Arab J Geosci Fig. 1 Location map of Metouia Oasis, South Tunisia
Experimental apparatus and procedure Two experiments (experiment 1 and experiment 2) were carried out to simulate soil salinization under fluctuating and stable saline shallow water tables, respectively. In both experiments, a plastic column, 60 cm in length and 7.5 cm in diameter, was packed as uniformly as possible to a bulk density of 1.38 g/cm3 using soil collected from Metouia Oasis. This soil was air-dried, passed through a 2-mm grid, and then packed in 2-cm layers inside the soil column. A glass filter, 0.5 cm in height and with an air entry value greater than 80 cm, was placed at the bottom of the column to prevent air invasion through the bottom. The outlets of the column were connected with a tube to a water reservoir for drainage and to a
Mariotte flask used to control groundwater level inside the soil column. This column was equipped with eight tensiometers positioned at 5, 9, 13, 17, 26, 34, 46, and 58 cm below surface and ten thermocouples inserted horizontally to the center of the column at 0.5-, 1.5-, 3-, 5-, 7-, 15-, 22-, 30-, 46-, and 54cm depths (Fig. 3). During experiment 1, the water table was fluctuating between 30- and 60-cm depths. The groundwater fluctuation was simulated by performing periodic drainage– imbibition cycles. The rising of the water table was performed within 24 h using the Mariotte flask while its lowering by
Table 1 Summary of physical and chemical properties of Metouia soil Properties
Metouia soil
Water content (%) Bulk density (g cm−3) Particle density (g cm−3) Saturated hydraulic conductivity (cm s−1) Particle size distribution Sand (%) Silt (%) Clay (%) Salinity (EC1:10, dS/m) pH SAR
24.3 1.38 2.76 7.27×10−4 87.8 4.2 8.0 3.39 8.30 20.40
Fig. 2 Measured and fitted (using the van Genuchten model, 1980) soil water retention curve for the Metouia soil
Arab J Geosci
deionized water in a 1:5 soil/water fraction, then shaken for 1 h and allowed to stay overnight. The supernatant was filtered through a Whatman 0.2-μm filter, and the Br concentration was measured by using an ion chromatograph (Shimadzu HPLC). Two heat lamps (150 W) placed 35 cm above the columns provided energy for evaporation. The lamps were on 12 h during each 24-h period. The experiments were conducted in a room where the ambient temperature was maintained at 25±2 °C.
Numerical simulations
Fig. 3 Experimental setup for groundwater fluctuation and soil salinization experiments in Metouia soil columns. Fluctuating and stable water tables were simulated using the Mariotte flask. The lamp was intermittently switched off automatically
drainage took between 4 and 5 days. These features were aimed to reproduce the groundwater fluctuation at field scale in Metouia Oasis. Throughout experiment 2, the water table was maintained constant at 60 cm depth using a connected Mariotte flask. For the purpose of soil sampling during the experiments, sets of 20 sample access holes (4 mm diameter) were made throughout the column at depths of 3, 11, 22, 34, and 46 cm for the first ten holes and at 5, 13, 26, 38, and 50 cm for the second ten holes (Fig. 3). The sampling technique involved removing approximately 2 g of soil from the column using a coring stainless steel tube (Shimojima et al. 1990). This technique allowed measurement of both tracer (Br) concentration and water content during the course of evaporation with minimal disturbance to the experimental column. The hole formed in the soil by sample extraction was immediately filled with a 4-mm-diameter rod and resealed. The soil water content was determined by weighting wet and oven-dried soil samples. For Br measurements, dry soil samples were mixed with
The results obtained from column experiments dealing with stable saline shallow groundwater (experiment 2) were used for further numerical simulations. The widely used HYDRUS-1D computer model version 1.4 (Šimůnek et al. 2008) was used to simulate vertical isothermal variably saturated flow in the one-dimensional soil column. In brief, HYDRUS-1D uses linear finite elements to numerically solve the Richards equation for saturated–unsaturated water flow and Fickian-based advection–dispersion equations for both heat and solute transport. The heat transport equation considers conduction as well as advection with flowing water. The solute transport equations assume advective–dispersive transport in the liquid phase and diffusion in the gaseous phase (Šimůnek et al. 2008). In order to solve these equations, we need to define initial and boundary conditions. In our simulations, initial pressure head values and soil temperatures (variable with depth) were determined from measured values on day 1 of the experiment. In the tracer transport model, the initial condition was represented by a zero concentration profile. Evaporative losses at the soil surface represent a major driving force to drive soil salinization (e.g., Shimojima et al. 1996), and as such, an evaporative boundary condition is adopted at the upper model limit. Evaporation from the soil surface was computed by Hargreaves formula implemented in HYDRUS-1D and was used as upper boundary condition. A thermocouple located at the soil surface provided continuous measurement of air temperature above the soil profile. With respect to Br transport during evaporation, we implemented a no-flux boundary condition at the soil surface because soluble salts are not volatile. Observed daily concentrations, obtained by linear interpolation from biweekly measured data, were used to define a time-variant bottom boundary condition. The soil profile was considered to be 60 cm deep, with observation nodes located at depths of 0.5, 5, 9, 13, 26, 38, and 58 cm for comparing calculated temperatures, pressure heads, and concentrations with measured values. A constant nodal spacing of 2 mm was used, leading to 302 discretization nodes across the problem domain.
Arab J Geosci 30
Experimental results Pressure head variations The pressure head variations observed during the experiments with the fluctuating and the stable water tables are illustrated in Figs. 4 and 5, respectively. The groundwater fluctuation at the bottom of the soil column during experiment 1 caused a fluctuation of pressure head values throughout the soil profile. All tensiometers from the top to the bottom of the soil profile responded approximately with similar pressure head magnitude. When the water table level was maintained stable during experiment 2, the pressure head values were almost constant for the whole monitoring period. Considering the pressure head–water content relationship of the soil, these observations suggest that the shallow groundwater has a great effect on the vertical distribution of soil water content and soil water potential. In addition, with respect to soil water content, it can be inferred that under a fluctuating groundwater, soil water profiles also fluctuate. Thus, an increase of the water table level would bring about an increase of the soil moisture content toward the soil surface and vice versa. On the other hand, under a constant groundwater, the soil water content would be constantly depleted, at the low and soil-controlled rate, due to the drying of the soil profile induced by evaporation.
Groundwater contribution to salt accumulation Br was used as a tracer to investigate the groundwater contribution to salt accumulation in Metouia soil under fluctuating and constant water table conditions. An artificial groundwater was prepared by dissolving 50 mmol KBr in 1 l distilled water. Figures 6 and 7 depict the Br content profiles at different sampling times in the case of fluctuating and stable water tables, respectively. It is worth noting that in our laboratory
Pressure head (cm)
Results and discussion
10 -10 -30 -50 -70 0
10 Time (days)
T2 (9 cm) T6 (34 cm)
T3 (13 cm) T7 (46 cm)
15
20 T4 (17 cm) T8 (58 cm)
Fig. 5 Pressure head variation throughout the monitoring period in the case of stable water table. T1 through T8: tensiometers. The tensiometer position is shown between parentheses
experiments, the lateral groundwater flow which may cause Br transportation was not included. We managed that only the vertical movement through the soil column is considered. The experimental results showed that in the case of the fluctuating groundwater, the rise of water table from 60 to 30 cm depth brought about a rise of Br content toward the soil surface (day 3). A decrease in the water table level from 30 to 60 cm induced a decrease of Br content with concentrations, however, higher than the initial condition observed on day 2. In the case of the constant water table, Br concentrations were only detected during the last sampling time (day 30) (i.e., 1 month after the starting of the experiment). The comparison between the fluctuating and the constant water tables, in terms of groundwater contribution to salt accumulation, showed the enhancement effect of groundwater fluctuation on salt concentration throughout the soil profile. Water table fluctuation induced a much greater spreading of the tracer within the column than the constant water table. In fact, throughout the soil profile, Br reached the depth of 12 cm below surface under fluctuating water table conditions and did
Br content (mmol/kg dry soil)
0
5
10
15
20
0
50
day 2, WT at 60 cm
30
10
10
day 3, WT at 30 cm day 8, WT at 60 cm
-10
Depth (cm)
Pressure head (cm)
5
-30 -50
20
day 18, WT at 60 cm
30 40
-70 0
5
10 Time (days)
T1 (5 cm) T6 (34 cm)
T2 (9 cm) T7 (46 cm)
15
20 T4 (17 cm) T8 (58 cm)
Fig. 4 Pressure head variation throughout the monitoring period in the case of a fluctuating water table. T1 through T8: tensiometers. The tensiometer position is shown between parentheses
50 60
Fig. 6 Br content profiles at different sampling times in the case of the fluctuating water table
Arab J Geosci 10
Br content (mmol/kg dry soil)
1
2
3
0 day 1, WT at 60 cm
Depth (cm)
10 20
Simulated
4
day 18, WT at 60 cm day 30, WT at 60 cm
30 40 50 60
Fig. 7 Br content profiles at different sampling times in the case of the stable water table
not exceed 50 cm depth under constant water table conditions. The Br content is on average five orders of magnitude greater under a fluctuating water table than under a constant one. The experimental results obtained from experiments 1 and 2 find their application in many shallow groundwater regions, such as wetlands and lowlands in river valleys where, as was stated by Chen and Hu (2004), a high groundwater table and significant hydraulic gradients between the saturated zone and the root zone lead to continuous supply of groundwater to the root zone. In those regions, the role of groundwater in variations of the root zone soil moisture (Chen and Hu 2004) and consequently salt content (Qiang et al. 2009) becomes essential. Model validation The results of the second experiment simulating salt accumulation under a stable water table were used for model validation by comparing simulated and measured data. Calculated pressure head variations, temperatures, and Br concentrations were compared to observed values throughout the experimental period. A number of statistical criteria were used for model performance evaluation. These are maximum error (ME), root mean square error (RMSE), coefficient of determination (CD), and modeling efficiency (EF). The formula, as well as the significance of these measures, is detailed in Loague and Green (1991). The lower limit for the ME, RMSE, and CD statistics is zero. The maximum value for EF is one. If EF is less than zero, the model-predicted values are worse than simply using the observed mean (Loague and Green 1991). Measured and simulated pressure head values at depths 9 and 26 cm during the simulation period are shown in Fig. 8. Generally, at both depths, the simulated pressure head variations were in good agreement with the measured values for the entire monitoring period, with slight overestimation at depth 9 cm. At the deeper depth (26 cm), the model predicted
Pressure head (cm)
0
Observed
-10 depth 26 cm
-30 -50
depth 9 cm
-70 -90 0
5
10
15
20
25
30
35
Time (days)
Fig. 8 Simulated and measured pressure head values at two depths (9 and 26 cm) throughout the soil column
reasonably well the observed pressure head data. The statistics ME, RMSE, CD, and EF listed in Table 2 confirm these observations. For instance, the calculated RMSE was 3.98 for depth 9 cm and 2.38 for depth 26 cm, indicating a better model prediction at this depth. The observed and predicted pressure head values were almost constant during the simulation period suggesting a very slight change in soil water content throughout the soil profile. Figure 9 depicts simulated and measured soil temperature at two depths (0.5 and 58 cm). The model successfully captured the diurnal courses of soil temperatures and temperature behavior with depth. The simulated soil temperatures reproduced almost perfectly the measured ones until day 15. Then, the observed temperatures were overestimated at both depths, with much better fit at 0.5 cm compared to the 58-cm depth. These observations can be seen by reviewing the statistical measures listed in Table 2. The statistics appear better at the first half of the simulation period (i.e., day 1 to day 15). For example, for the 58-cm depth, the RMSE is equal to 0.72 during the period starting from day1 to day 15 and to 2.15 during the last half of the simulation period. The discrepancy between the measured and the simulated soil temperature may be the result of many factors such as an uncertainty in input parameters, incorrect initial soil moisture, the assumption of a constant daily temperature amplitude, or external atmospheric effects. Another possible reason for the observed discrepancies between predicted and measured soil temperatures is the progressive formation of a thick salt crust at the soil surface. This explanation is supported by the fact that the deviation of the simulated values started from day 15, corresponding approximately to the consolidation of the salt crust (checked by daily visual observations) when the soil surface was drying out. The salt crust acts as a surface barrier to the incoming radiation at the soil surface and consequently will cause a decrease of soil temperature. In HYDRUS-1D, there is no consideration for feedback between concentrations and soil temperature variation with time (Šimůnek, personal communication). In this context, Noborio and McInnes (1993) showed experimentally that the addition of salt to clay loam and loamy sand soils did reduce their thermal conductivity (λ)
Arab J Geosci Table 2 Model performance statistics for pressure head, temperature, and concentration values Depth Period MEa RMSEa CDa EFa
a
ME, RMSE, CD, and EF would yield 0, 0, 1, and 1, respectively, in case of a perfect fit between predicted and observed data
Pressure head
Temperature
9 cm 1∼30 6.19 3.98 0.49 −1.76
0.5 cm 1∼15 7.01 2.87 1.18 0.79
26 cm 1∼30 3.02 2.38 2.20 −0.27
compared to salt-free samples. They pointed out also that the λ values of soils treated with different salts generally did not significantly differ from one another. Because of the general lack of information about the effects of salt crust formation on soil temperature, we conducted a study under controlled conditions. Two soil columns with 8 cm length and 5 cm inner diameter were packed homogenously with Metouia soil at a bulk density of 1.38 g cm−3. For each column, two thermocouples were inserted at 1- and 7-cm depths. Two lamps placed 35 cm above each column were used to heat the soil surface. After saturating the two columns with pure water and waiting for equilibrium, the lamps were turned on, and the experiments started. The salt crust progressive formation was tracked daily by visual observation of the soil surface. In one column, the crust was allowed to form while in the other, the surface soil was carefully scratched to prevent salt crust formation. Figure 10 depicts the soil temperature variation at 1 cm depth of soil columns with and without surface salt crust. The damping effect of the salt crust on temperature values
Temperature (oC)
50
depth 0.5 cm
40 30
Concentration
16∼30 9.55 3.43 1.17 0.71
58 cm 1∼15 1.92 0.72 1.07 −1.03
16∼30 3.01 2.15 0.03 −36.33
0–60 cm 1∼30 0.01 0.04 2.12 0.89
compared to those obtained for soil with no crust at the surface can be clearly seen. For the deeper depth (not shown), the salt crust effect on temperature reduction was observed but was less evident due most likely to the short period of experimentation (∼10 days). This experiment suggests that the effect of the salt crust on reducing soil temperature is likely the reason for the discrepancies between measured and simulated temperatures observed in this study. With respect to solute concentration, Fig. 11 shows the measured and the simulated soil Br concentrations obtained at the end of the experimental investigation. It can be seen that the model reproduced fairly well the observed concentrations. The good performance of the model in predicting Br concentrations can be seen through the ME and RMSE values close to zero and the EF value close to unity (Table 2). Overall, the HYDRUS-1D model for water, heat, and solute transport applied under laboratory conditions performed reasonably well during salt accumulation experiments when modeled values were compared with experimentally measured ones despite some discrepancies that were caused by either model limitations or input parameter uncertainty. The spatial distributions of pressure head, temperature, and Br concentrations agreed generally well with measurements. Therefore, the HYDRUS-1D model is validated for further simulation scenarios which are difficult and time consuming if performed at laboratory scale.
20 Observed
10
Simulated
0 0
5
10
15
20
25
30
45
35
depth 58 cm
40 Temperature (oC)
Temperature ( o C)
Time (days)
30 20 Observed
10
40 35 30 25 20 no crust
15
crust
Simulated
10
0 0
5
10
15
20
25
30
35
Time (days)
Fig. 9 Simulated and measured soil temperature at two depths (0.5 and 58 cm) throughout the soil column
0
2
4
6 8 Time (days)
10
12
Fig. 10 Observed soil temperature variation at 1 cm depth in the case of soil columns with and without salt crust
Arab J Geosci 0 Observed Simulated
-10
Depth (cm)
-20
-30
-40
-50 -60
-70 0.0
0.2 0.4 0.6 Concentration (mmol/cm3)
0.8
Fig. 11 Simulated and measured Br concentrations obtained on day 30 throughout the soil profile
Simulation scenarios In order to simulate the effect of a fluctuating water table on soil surface salt accumulation, the bottom boundary condition was set as variable pressure head in the HYDRUS-1D model. The groundwater is assumed to have an angular fluctuation frequency ω [T −1] which equals p/t 0 where p is the number of groundwater peaks during a period of time t 0. Two groundwater fluctuation frequencies were considered, namely WTf1 and WTf2 with values of ω =0.22 day−1 and ω =0.46 day−1, respectively. Figure 12 depicts the simulated salt concentration profiles induced by a stable water table (WTs) and fluctuating water
0
-10 WTs WTf1
Depth (cm)
-20
WTf2 -30
-40
-50
-60 0.0
0.5
1.0 1.5 Concentration (mmol/cm 3)
2.0
2.5
Fig. 12 Simulated salt concentration profiles induced by stable water table (WTs), and fluctuating water tables, WTf1 and WTf2, with fluctuation frequencies of 0.22 and 0.46 day−1, respectively
tables with two different frequencies (WTf1 and WTf2). Quite clearly, regardless of the fluctuation frequency, the fluctuating groundwater caused more salt to accumulate at soil surface than that caused by a stable groundwater. In fact, at the end of the simulation period (60 days), the Br concentrations at the soil surface were approximately 0.5, 1.7, and 2.1 mmol/cm3 induced by WTs, WTf1, and WTf2, respectively. The detection of salt accumulation at the soil surface in the case of the stabilized water table at 60 cm indicates that this depth is not sufficient to limit soil surface salt accumulation. Furthermore, the comparison of the time needed for Br to reach the soil surface under stable and fluctuating water tables revealed that the groundwater fluctuation caused an enhancement of the soil surface salt accumulation. For instance, in the case of the stable water table, after 42 days, there was nearly zero salt concentration at the soil surface while in the case of the fluctuating water table, the Br originating from groundwater reached the soil surface after 36 days of simulation. These observations of predicted soil salt distributions provide evidence that the fluctuating groundwater transports salts faster and higher into the vadose zone than the stable groundwater. The numerical simulations showed also that the frequency of groundwater fluctuation affects salt accumulation throughout the soil profile. Hence, the water table WTf2 brought about a higher magnitude of soil surface salt accumulation in comparison to WTf1 which has lower fluctuation frequency. At the end of the simulation (day 60), the Br concentration at the soil surface was about 2.1 mmol/cm3 in the case of WTf2 and was about 1.7 mmol/cm3 under WTf1. The mechanisms involved in salt flux in the liquid phase include different transport processes: advective flux, diffusive flux, and hydrodynamic dispersive flux (e.g., Hillel 1998). Advection represents mass flow of solute caused by the flux of water. If solutes were moving into a soil from a source of constant concentration at the surface, this would produce a square-wave distribution of solute concentration with depth known as piston flow (Radcliffe and Simunek 2010). The diffusion is the movement of solute molecules in response to differences in concentrations. Hydrodynamic dispersion is a nonsteady, irreversible process that consists of mechanical dispersion and molecular diffusion (Tindall et al. 1999). In our simulations, the differences between salt profiles obtained under a fluctuating and a stable groundwater as shown in Fig. 12 can be explained by a difference in transport mechanisms. In the case of a stable water table, the salt is transported toward the soil surface mainly by capillary rise of saline groundwater and secondarily by molecular diffusion. Under a fluctuating water table, in addition to these transport mechanisms, the groundwater flux triggers rapid and high mass flow of salt toward the soil surface. Considering the case of many saline semiarid and arid lands where the water table is at or near the soil surface, the above discussed numerical simulation results suggest that the groundwater fluctuation is most
Arab J Geosci 0 25 %
50 %
0%
12 %
-10
Depth (cm)
-20
-30
-40
-50
-60 0.0
0.1
0.2 0.3 0.4 0.5 Concentration (mmol/cm 3)
0.6
0.7
Fig. 13 Simulation of surface mulching effect on soil surface salt accumulation. The percentages represent the intercepted temperature reduction
respectively. Such values are comparable to the field conditions in arid and semiarid areas. During three simulation scenarios, the surface mulching was simulated by reducing the temperature by 12.5, 25, and 50 %. The results were compared to the initial simulation under observed temperatures. Figure 13 depicts the Br concentration profiles obtained at the end of the simulation under various temperature conditions. It can be seen that the effect of less evaporation (through temperature reduction) on soil salinity dynamics was considerable. In fact, the simulation results showed that a temperature reduction of 12.5, 25, and 50 % brought about a reduction of the Br concentration at the soil surface of 0.2, 0.3, and 0.4 mmol/cm3, respectively, compared to a mulching-free soil. Nevertheless, this measure did not prevent salt to reach the soil surface even with 50 % temperature reduction. It would be more effective in limiting soil surface salt accumulation if combined with another technique. Irrigation management options
likely a major driving process of salt accumulation enhancement in these areas. Management strategies In this section, two management strategies are evaluated by numerical simulations: (1) reducing the potential evaporation by using surface mulching and (2) improving irrigation management. Surface mulching (reducing the evaporation) In our numerical simulations, evaporation was used as surface boundary condition. The potential evaporation was generated using the Hargreaves formula which requires maximal (T max) and minimal (T min) daily air temperatures. Throughout the experimental period, T max and T min averaged 40 and 20 °C,
In the case of Metouia Oasis, the irrigation management within the irrigated area is based on water allocation. The irrigation network operates on rotation delivery with an interval of 10 to 15 days between two consecutive irrigations in the same farmland. The flood irrigation is the main irrigation method practiced by farmers in Metouia Oasis. In this context, during the intervals between irrigations when there is no downward flow of soil water, significant amounts of water can move upward by capillary forces. The loss of water by evaporation causes much of the salt to accumulate at or near the surface. Based on the facts mentioned above, numerical simulations were carried out in order to evaluate various irrigation management strategies to prevent soil salinization induced by saline shallow groundwater. For this objective, three simulation scenarios were conducted: (1) no irrigation—in this case,
0
-10
Continuous irrigation
Intermittent irrigation
No irrigation t= 30 t= 42
Depth (cm)
-20
t= 60 -30 -40
-50
-60 0.0
0.2 0.4 0.6 Concentration (mmol/cm3)
0.0
0.2 0.4 0.6 Concentration (mmol/cm3)
0.0
0.2 0.4 0.6 Concentration (mmol/cm3)
Fig. 14 Br concentrations obtained for three different simulated irrigation options: no irrigation, intermittent irrigation, continuous irrigation
Arab J Geosci
the salt accumulation is caused by saline shallow groundwater; (2) intermittent irrigation at 10 days interval by applying 5 cm/day at once—this case is similar to what it is applied in Metouia Oasis; and (3) continuous irrigation by applying 0.5 cm/day—in this case, instead of applying 5 cm/day at once, this quantity is divided and equally distributed on a daily basis. The simulation results of the three irrigation scenarios are illustrated in Fig. 14. When there was no irrigation, salt accumulated at the soil surface due to capillary rise from saline shallow groundwater and the evapoconcentration process. Since the soil was assumed initially to be salt free, the inflow of salt from the groundwater is the main mechanism controlling soil surface salt accumulation. By applying intermittent irrigation, the salt concentrations were reduced throughout the soil profile compared to the “no irrigation” scenario. However, this irrigation practice did not prevent salt to reach the soil surface. This was observed on days 54 and 60. A reasonable explanation of the simulation results of the intermittent irrigation scenario could be that during the intervals between irrigation, when there is no downward flow of soil water, significant amounts of water can move upward by capillary forces. Capillary flow brings water to the surface soil, and evaporation causes the surface soil to be at lower water potential than that at static equilibrium. The lower water potential of the surface soil causes water flow from the water table toward the soil surface. The inflow of saline water at the bottom of a soil column controls the addition of salt to the soil column. The loss of water by evaporation causes much of this salt to accumulate at or near the surface. The daily irrigation using small irrigation amounts was the most effective scenario for preventing salt to accumulate at the soil surface. In this case, salt concentration was nearly zero down to 40 cm below the soil surface. It seems that the continuous supply of water by irrigation maintained enough moisture at the surface layer to meet the atmospheric evaporativity. Consequently, the soil surface remains at higher water potential than that at static equilibrium which causes water flow to be from the top to the bottom of the soil profile.
Conclusion Shallow groundwater systems are known for their rapid and large fluctuation in response to recharge events. Elucidating the effects of such behavior on soil salinization, so far less investigated, was the main objective of the present study. Laboratory and modeling investigations were carried out for this purpose. The experimental results showed that the increase of salt concentration toward the soil surface was relatively slow and progressive under a constant water table and was notably more rapid and with higher magnitude in the case of a fluctuating
water table. These findings were substantiated by numerical investigations which revealed that groundwater fluctuation caused more salt to accumulate at the soil surface compared to that caused by a stable groundwater. In addition, less time was needed for salt to reach the soil surface under a fluctuating water table. It was further shown that soil surface salt accumulation seems to be a function of the water table fluctuation frequency, as increasing the fluctuation frequency increased soil surface salinization under the same evaporation boundary conditions. Management strategy analyses through numerical simulations revealed that reducing the evaporation rate through a decrease of the temperature intercepted at the soil surface brought about a significant reduction of salt concentration at the soil surface. With respect to the appropriate irrigation scenario, numerical simulations showed that when there was no irrigation, salt accumulated at the soil surface due to capillary rise from the saline shallow groundwater and the evapoconcentration process. In this case, the inflow of salt from groundwater is the main mechanism controlling soil surface salt accumulation. By applying intermittent irrigation, the salt concentrations were reduced throughout the soil profile compared to the “no irrigation” scenario. The continuous irrigation with small amounts was the most effective scenario for preventing salt to accumulate at the soil surface. Hence, for Metouia Oasis, a different irrigation strategy than that currently applied would bring about better results for the environment. Overall, the results of the present study proved the importance of groundwater fluctuation as a contributor to soil salinization and suggest that the influence of groundwater fluctuations on top soil salinity must be appreciated and evaluated if a better understanding of groundwater-induced soil salinization is to be accomplished.
References Abdel-Kawy WAM, Belal A (2013) Use of satellite data and GIS for soil mapping and monitoring soil productivity of the cultivated land in El-Fayoum depression, Egypt. Arab J Geosci 6:723–732. doi:10. 1007/s12517-011-0371-y Al-Dhuhli HS, Al-Rawahy SA, Prathapar S (2010) Effectiveness of mulches to control soil salinity in sorghum fields irrigated with saline water. A monograph on management of salt-affected soils and water for sustainable agriculture, 41–46. 2010 Sultan Qaboos University Al-Garni MA, El-Kaliouby HM (2011) Delineation of saline groundwater and sea water intrusion zones using transient electromagnetic (TEM) method, Wadi Thuwal area, Saudi Arabia. Arab J Geosci 4:655– 668. doi:10.1007/s12517-009-0094-5 Cartwright N, Baldock TE, Nielsen P, Jeng DS, Tao L (2006) Swashaquifer interaction in the vicinity of the water table exit point on a sandy beach. J Geophys Res 111:1–13 Chen X, Hu Q (2004) Groundwater influences on soil moisture and surface evaporation. J Hydrol 297:285–300
Arab J Geosci FAO (2002) Biodrainage—principles, experiences and applications. Knowledge Synthesis Report No. 6–May 2002. International Programme for Technology and Research in Irrigation and Drainage Food and Agriculture Organization of the United Nations Fekri M, Kasmaei LS (2013) Effects of windy sand and LECA mulches on decreasing evaporation from soil surface. Arab J Geosci 6:163– 168. doi:10.1007/s12517-011-0333-4 Gowing JW, Konukcu F, Rose DA (2006) Evaporative flux from a shallow water table: the influence of a vapour–liquid phase transition. J Hydrol 321:77–89 Hillel D (1998) Environmental soil physics: fundamentals, applications, and environmental considerations, 1st edn. Academic Press, San Diego, 771 pages Ibrahimi K, Miyazaki T, Nishimura T (2010) A high measurement frequency based assessment of shallow groundwater fluctuations in Metouia Oasis, South Tunisia. Hydrol Res Lett 4:75–79. doi:10. 3178/HRL.4.75 Kara T (2002) Irrigation scheduling to prevent soil salinization from a shallow water table. Proc. IS on Salination for Hort. Prod. Eds. U. Aksoy et al. Acta Hort. 573, ISHS: 139–151 Klute A, Dirksen C (1986) Hydraulic conductivity of saturated soils. p. 694–703. In: A. Klute et al. (ed.) Methods of soil analysis. Part 1. Physical and mineralogical methods. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI Lal R, Shukla MK (2004) Principles of soil physics. Marcel Dekker, Inc, New York, 716p Loague K, Green RE (1991) Statistical and graphical methods for evaluating solute transport models: overview and application. J Contam Hydrol 7:51–73 Matinfar HR, Panah SKA, Zand F, Khodaei K (2011) Detection of soil salinity changes and mapping land cover types based upon remotely sensed data. Arab J Geosci 6:913–919. doi:10.1007/s12517-0110384-6 Mohamed ES, Belal A, Saleh A (2013) Assessment of land degradation east of the Nile delta, Egypt using remote sensing and GIS techniques. Arab J Geosci 6:2843–2853. doi:10.1007/s12517-0120553-2 Noborio K, McInnes KJ (1993) Thermal conductivity of salt affected soils. Soil Sci Soc Am J 57:329–334 Northey JE, Christen EW, Ayars JE, Jankowski J (2006) Occurrence and measurement of salinity stratification in shallow groundwater in the Murrumbidgee Irrigation Area, south-eastern Australia. Agric Water Manag 81:23–40
O’Brien AL (1982) Rapid water table rise. J Am Water Resour Assoc 18(4):713–715 Qiang L, Baoshan C, Zhifeng Y (2009) Dynamics of the soil water and solute in the sodic saline soil in the Songnen Plain, China. Environ Earth Sci 59:837–845 Radcliffe DE, Simunek J (2010) Soil physics with Hydrus: modeling and applications. CRC press, Boca Raton, 373 pp Rengasamy P (2006) World salinization with emphasis on Australia. J Exp Bot 57(5):1017–1023 Shimojima E, Curtis AA, Turner JV (1990) The mechanisms of evaporation from sand columns with restricted and unrestricted water tables using deuterium under turbulent airflow conditions. J Hydrol 117:15–54 Shimojima E, Yoshioka R, Tamagawa I (1996) Salinization owing to evaporation from bare-soil surfaces and its influences on the evaporation. J Hydrol 178:109–136 Shirokova Y, Forkutsa I, Sharafutdinova N (2000) Use of electrical conductivity instead of soluble salts for soil salinity monitoring in Central Asia. Irrig Drain Syst 14:199–205 Šimůnek J, van Genuchten MT, Šejna M (2008) Development and applications of the HYDRUS and STANMOD software packages and related codes. Vadose Zone J 7:587–600. doi:10.2136/vzj2007. 0077 Tindall JA, Kunkel JR, Anderson DE (1999) Unsaturated zone hydrology for scientists and engineers. Prentice-Hall, New Jersey, 624 pp van Genuchten MT (1980) A close-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci Soc Am J 44: 892–898 van Genuchten M Th, Leij FJ, Yates SR (1991) The RETC code for quantifying the hydraulic functions of unsaturated soils, version 1.0. EPA Report 600/2-91/065, U.S. Salinity Laboratory, USDA, ARS, Riverside, California Werner AD, Lockington DA (2004) Dynamic groundwater and salt transport near a tidal, partially penetrating estuary. In computational methods in water resources, proceedings of the XVth international conference on computational methods in water resources. Dev Water Sci 55:1535–1547 Yamanaka T, Inoue M, Kaihotsu I (2004) Effects of gravel mulch on water vapor transfer above and below the soil surface. Agric Water Manag 67:145–155 Yuan C, Lei T, Mao L, Liu H, Wu Y (2009) Soil surface evaporation processes under mulches of different sized gravel. Catena 78:117– 121