Saudi Arabia is an arid country where summer air-temperature
often exceeds 45°C. Improved irrigation systems such as sprinkler,
drip (surface or sub-surface) and the trickles face clogging problem mainly
due to salt accumulation resulting from water evaporation in and around
the water supply points. The terms drip, trickle and spray irrigation,
have been supplanted by the term micro irrigation, by the American Society
of Agricultural Engineers. Micro irrigation includes all methods of frequent
water application, in small flow rates, on or below the soil surface.
Ideally, the volume of water is applied directly to the root zone through
small outlets (emitters) in quantities that approach the consumptive use
of plants for optimal production. During the dry season in humid areas,
or in arid climates, micro irrigation significantly affected the quality
and quantity of yield, pest control and harvest timing. Also, limited
soil wetting permits orchard cultural operations and minimized the labour
scheduling problems (Nakayama and Bucks, 1986).
Normally water infiltration takes place in the region
directly around the emitter, which is small compared with the total soil
volume of the irrigated field. Previous studies on water infiltration
into soils under a point-source dealt mainly with the soil moisture content
or wetting front advance pattern during or just after the termination
of the irrigation process following a continuous irrigation (Warrick,
1974; Oron, 1981; Mostaghimi et al., 1982; Wu et al., 1999;
Elmaloglou and Malamos, 2006). On the other hand, many studies developed
techniques to describe continuous trickle irrigation, whereas only few
studies focused on the pulse trickle irrigation. Levin et al. (1979)
and Mostaghimi et al. (1981), studied the effect of discharge rate
and intermittent water application by a surface point-source. Revol et
al. (1996) observed free water pond formation under a trickle irrigation.
Revol et al. (1996) and Cote et al. (2003) investigated the effect
of intermittent water application on the wetting front advance in subsurface
trickle irrigation. They concluded that advance in wetting front is significantly
affected by the total amount of water applied.
The distribution pattern of soil water resulting from
trickle irrigation is quite different from those resulting from the conventional
modes of irrigation. It was observed that If many emitters are close together
on a line, the result is an effective line or strip source and the pattern
of wetting will be two dimensional (horizontal in the direction perpendicular
to the source and vertical) rather than three dimensional (Hillel, 1982).
Numerical solutions were developed to analyze multidirectional
infiltration under trickle sources (Bresler et al., 1971; Brandt
et al., 1971; Levin et al., 1979; Ragab et al., 1984;
Taghavi et al., 1984; Svehlic and Ghali, 1984; Ghali and Svehlik,
1988; Lafolie et al., 1989). Bresler et al. (1971) and
Elmaloglou and Malamos (1999) studied soil water movement and distribution
in homogenous soil profiles under various rates of water application from
a surface trickle line source. They found that for shallow rooted and
widely spaced row crops, higher water application rates are advisable,
while for deep rooted and closely row crops, lower application rates are
recommended. Also, Elmaloglou and Malamos (2003) analyzed the local infiltration
from a surface line source of trickle irrigation in two homogeneous and
unsaturated soils, using three wetting rates in each soil. They concluded
that the empirical model was successful and proved as a useful tool for
predicting the depth of the wetting front throughout the soil profile
under a surface line source of trickle irrigation.
The discharge of emitter was calculated by Roberson and
Crowe (1993) as follow:
||Discharge through orifice (m3
||Area of cross-section of the orifice (m2),
||Acceleration due to gravity, 9.81 (m sec-2),
Depth of water over the centre
of the orifice (on the upstream side) in case of free flow orifice,
or the difference in elevation between the water surface at the
upstream and downstream faces of the orifice plate in case of
submerged orifices (m).
The principles of pulsing were first set out by Karmelli
and Peri (1974). They described a pulse as consisting of an operating
phase (to), during which water is applied to the soil at a
discharge (Qp) and a resting phase (tr) when the
flow is zero. The average pulsed discharge (Qa) over the irrigation
period (to+tr) would be:
The average pulsed discharge (Qa) is equivalent
to a continuous discharge (Qc). Thus the same amount of water
would be applied with a continuous discharge (Qc) as in the
pulsed regime provided the irrigation period was (to+tr).
Several investigations demonstrated the effect of pulsing
flow compared with continuous applications.
Mostaghimi and Mitchell (1983) showed that pulsing significantly
reduced the water losses from deep percolation in sandy soils and increased
the lateral spread of water in the soil. Levin and Van Rooyen (1977) and
Levin et al. (1979) also reported similar advantages in reducing
percolation losses but also showed that wetting patterns for pulsed and
continuous flow applications were almost identical at very low discharges
for point sources.
Among the various problems associated with the trickle
irrigation, Emitter clogging is considered as the most serious problem
in trickle irrigation. The emission uniformity reduces greatly when this
happens and the crop damage may occur before the clogging is detected.
Obviously, the effect of pulsing depends not only on
the soil type but also on the discharge and operating and resting times
chosen. This study discusses the results of a laboratory study to show
that pulsing discharges, within the range of most commercial trickle systems,
can produce similar wetting patterns as compared to those of low continuous
discharges. Also to find out the possibility for increasing the emitter
sizes in order to minimize the tendency of emitter clogging under arid
MATERIALS AND METHODS
The experiments were conducted at the Agricultural Engineering Laboratories
of Agricultural and Veterinary Training and Research Station, King Faisal
University, Al-Hassa, Kingdom of Saudi Arabia during 2005-2006. The soils
of the research station are mostly coarse textured (sandy soils). The
physical and chemical properties of soil used can be shown in Table
The experimental apparatus consisted of a soil container
and a water application system. The frame of the soil container was made
of aluminum with overall dimensions of 1x1x0.2 m length, depth and width,
respectively. The soil profile was designed in such a way that its thickness,
which coincides with the y-axis, was smaller than its length and depth
whose directions were along the x-axis and z-axis, respectively. To observe
record the advance of the wetting front, the front face
of the container was constructed with a 20 mm thick glass plate supported
by an aluminum frame while the back face was made of fiber plate. The
soil water contents were sampled through 25 mm diameter holes drilled
on a grid layout on the fiber plate panel to the rear. These holes were
plugged with rubber bungs during the course of the experiment and could
be removed when soil samples were taken. The remaining two sides forming
the thickness of the container and the bottom side were made of hard fiber
sheets firmly fixed to the frame to prevent water leakage. A 2 cm grid
system was drawn on the outer side of the glass plate using a waterproof
felt pen for plotting the wetting fronts easily and accurately.
The water application system consisted of a water supply reservoir and
the drippers. The water supply reservoir was a bucket of 20 L capacity
and connected to the drippers by means of rubber tubing. The emitter used
is 125 mm PVC pipe in which 19 needles were fixed 10 mm apart and placed
centrally across the soil tank to simulate a line source that would be
used to irrigate a row crop. By using a line source, water movement along
the y-axis was minimized so that the lateral and vertical movement of
water along the x-axis and z-axis, respectively could be easily monitored.
The water was supplied to the reservoir from a water supply tap by means
of a rubber tubing connection. The discharge of the emitters from the
reservoir was controlled by adjusting the height of the bucket above the
soil sample container by means of a pulley system and using a valve. Uniform
discharge was attained by means of an overflow siphon (Fig.
Before each test, the soil was air-dried, mixed thoroughly
to obtain a homogeneous mixture, then placed in the tank and compacted
to a constant bulk density.
Four pulse application regimes were selected within the
range of most commercially available trickle equipment for comparison
with a continuous flow of 8 L min-1 h-1. The flow
rates used were as follows:
||1.33 cm3 min-1 cm-1
applied continuously (equivalent to 8 L min-1 h-1).
||2.66 cm3 min-1 cm-1 applied as
30 min on, 30 min off pulse (equivalent to 16 L min-1
||4 cm3 min-1 cm-1 applied as 20
min on, 40 min off pulse (equivalent to 24 L min-1 h-1).
||8 cm3 min-1 cm-1 applied as 10
min on, 50 min off pulse equivalent to 48 L min-1 h-1).
||16 cm3 min-1 cm-1 applied as
5 min on, 55 min off pulse equivalent to 96 L min-1 h-1).
The total quantity of water applied came to 12.8 L in
each case which is equivalent to water application of
||Soil container and water supply system
(not to scale)
64 L min-1 run along a trickle lateral. The
total running time for all the water applications came to 8 h. The soil
wetting patterns were observed at regular time intervals. During each
test, soil samples were taken at the end of each irrigation to determine
soil moisture contents. Asymmetrical water distribution was assumed in
the whole soil profile.
RESULTS AND DISCUSSION
The data in Fig. 2 and 3 show that
the shape of the wetted soil zone was significantly affected by the discharge
rate of different treatments. An increase in the water application rate
from control (8 L min-1 h-1) to 16, 24 and 48 L
min-1 h-1 resulted in 9.1, 14.5 and 19.1% increase
in vertical advance and 4.1, 8.3 and 12.5% decrease in the horizontal
advance, respectively as compared to the control (continuous) treatment.
This would mean that pulsed discharges up to six times the equivalent
continuous discharge caused only minor changes in the soil wetting zone
(wetting front) both vertically and horizontally. On the other hand, increasing
the water application rate to 96 L min-1 h-1 caused
63.6% increase in the vertical advance and 40% decrease in the horizontal
advance when compared with the continuous treatment. This suggests that
as the level of pulsed flow increased up to twelve times the continuous
flow, it reduced the
of the wetting front from a trickle line source
of the wetting front from a trickle line source
deep percolation and increased the horizontal spread.
These findings support the work of Mostaghimi and Michell (1983), who
reported that an increase in pulsed flow caused significant reduction
in deep percolation and appreciable increase in horizontal spread. The
ratios between the maximum width and the maximum depth of the wetted profiles
were 0.92, 1.04, 1.15, 1.25 and 2.5 for 8, 16, 24, 48 and 96 L min-1
h-1 discharge rates, respectively. Mathematical analysis of
the vertical and horizontal advance show that these can be expressed as
a power function.
V = a V0c
H = b V0d
||Vertical advance (mm),
||Horizontal advance (mm),
|a, b, c and d
||Product of the discharge, L min-1 h-1 and
time in min.
A strong correlation was found between the total water applied and the
vertical and horizontal advances at different water application rates
(Table 2) and can be expressed by power function.
A major factor limiting the use of high pulsed discharges
is the surface water ponding. Once this exceeds the initial infiltration
capacity of the soil, the surface ponding spreads rapidly. At low discharges,
surface ponding extended only 50 mm for the continuous flow and 305, 493
and 503 mm for the pulsed discharges of 16, 24 and 48 L min-1
h-1, respectively. Although at a discharge rate of 96 L min-1
h-1, the flow pulses lasted only for five minutes but the horizontal
infiltration continued at the same rate for several minutes after the
water supply was switched off. All this process could be subjected to
ponding which acted as a reservoir on the soil surface. This may not only
change the wetting pattern significantly but may also result in run-off
losses and soil erosion. The results agree with findings of Mostaghami
and Mitchell (1983) and Levin et al. (1979), who concluded that
pulsing water discharge could significantly reduce deep percolation water
losses in sandy soils and increase the lateral spread of water in soils.
Soil water contents measured across the wetted soil zone at a depth of
150 mm (Fig. 4) indicate that as the flow rates increased
under a pulsed regime, the volume of wetted soil increased for the same
volume of applied water. The results are in agreement with the findings
of Elmaloglou and Malamos (1999), who found that for shallow rooted and
widely spaced row crops, higher water application rates are advisable,
while for deep rooted and closely spaced crops, lower water application
rated are recommended. In conclusion, pulsing water discharge produced
favorable soil moisture regime by maintaining higher soil water contents
than the continuous flow in the active crop root zone area.
The results show that there is a good possibility to
maintain and establish optimal wetting patterns in soils under trickle
irrigation using pulsed discharges which are significantly higher as compared
to the equivalent continuous flow. This could lead to the use of large
emitter sizes thus reducing the emitter clogging problems and the soil
water filtration. The data also indicate that the emitter sizes could
be increased up to 2.4-3.5 times and the emitter cross section areas up
to 6-12 times which would reduce the clogging tendency of emitters.
Although pulsing discharges might reduce emitter clogging
but it may increase the cost of the system as
Soil water content
with horizontal distance from trickle line source
equations describing the vertical and horizontal wetting front
advance under different applications as a function of water applied
large lateral pipes will be required to supply the large
flows and automatic sequencing valves might be needed to pulse the flows.
Moreover, this cost could be offset by reduction in maintenance and cleaning
or replacing emitters. The costs of pumping should not be affected provided
the pulsed discharge (Qp) is similar to that required in a continuous
flow system (Qc). However, this can be achieved by careful selection of
the number of trickle laterals required for efficient operation during
The use of pulsed water supply reduced the vertical wetting
front in soils thus reducing the deep water percolation losses in coarse
textured soils. On hourly cycle basis, a pulsed flow rate up to six times
the continuous flow rate did not significantly change the shape of the
wetting bulb. From design point of view, water discharge rates between
48 and 96 L min-1 h-1 could be used and this will
allow the use of large emitter sizes, reduce emitter clogging problems
and appreciably reduce soil filtration. A strong correlation was obtained
between the water application rates and the vertical and horizontal advances
which could be expressed as power function. Pulsing the water supply wetted
the soil more uniformly and produced more favorable moisture regime than
continuous supply. The significance of surface ponding width in increasing
the wetting bulb was noticed particularly in the 96 L min-1
h-1. The increase in the high pulse rate than the soil infiltration
capacity caused rapid increase (spread) in the surface ponding thus showing
a major factor limiting the use of high pulsing rates in trickle irrigation.
The author thanks the SABIC for financial support in
carrying out this study and to procure necessary equipments for this research.