Name: Mirna Faris Slewa

Class: Third Year

SURGE IRRIGATION: INFILTRATION

1-Introduction

S urge irrigation, the intermittent application of water to furrows or borders in a

series of surges with constant or variable time spans, has been shown to have
benefits over conventional irrigation (continuous application of water). The
advantage of surge IrrIgation is faster advance along the field, thereby improving
distribution uniformity along the length of the furrow or border. Surge irrigation
also reduces tailwater and deep percolation losses, thus improving water
application efficiency. This implies a reduction in total volume of irrigation water
required and provides a means of automating the management and control of water
application to the field (Bishop et aI., 1981; Podmore and Duke, 1982).
Several studies have been conducted and many theories have been proposed to
explain the surge phenomena. This mechanism includes consolidation of the top soil
matrix, which reduces the hydraulic conductivity of the surface layer (Samani, 1983;
Farahani,1991). Surface sealing, caused by either the deposit of sediment on the
surface or by the migration of soil particles into pore spaces, results in substantial
reduction of the hydraulic conductivity of the thin surface (Stieb, 1983; Kemper et
aI., 1'I8S, Trout, 1990). Other mechanisms, such as air entrapment, in which air is
isolated III soIl pores or as a layer between water layers, also has great effects on
infiltration rate (Duke, 1988; Seymour, 1990; Lep, 1981; Farahani, 1991). Another
mechanism of surge irrigation is decreased furrow roughness and more stable cross
section during infiltration of water between surges (Coolidge et aI., 1982; Trout,
1990). There are other mechanisms, such as redistribution of water during the off-
time and hysteresis, which reduce the hydraulic gradient (Izadi et aI., 1988, Levin et
aI., 1979; Killen et aI., 1987)
Although there has been some pro!o'Tess in J.mderstanding the mechanisms of
surge irrigation, the process of infiltration still IS not completely understood. The
effects of surge irrigation are soil type related and seem to be soil texture dependent
(Alemi et aI., 1988; Saleh and Hanks, 1989). Surge irrigation has also responded to
management variables (AISaud et ai., 1993; Izuno et ai., 1985; Blair et ai, 1985), and
has been found to cause changes in soil properties (Senzanje, 1994; Saleh et al..
1989. Farahani 1991). AI-Saud et al. (1993) found that the surge effects increased
with increased off-time up to an optimum off-time. Senzanje (1994) tried to predict
infiltration rates under field conditions using a bulk density response surface for
clay loam and sandy loam, but obtained low correlation.
2- Problem Definition and Background

While surge irrigation is indeed a good strategy for improving the performance of
surface irrigation management in order to increase efficiency, design and management of
surge flow irrigation is not a straightforward issue and, to date, the results are not readily
predictable. There is a need for a more reliable design and management procedure to
determine optimum on-time, off-time and cycle ratios, in order to manage surge
ImgatJOn for given field conditions such that the surge effect and beJlefit are
ma),.lllllzed. From literature discussions, it can be summarized that in surge Irrigation
there are surge mechanisms that affect advance and infiltration; these mechamsms operate
and influence advance and infiltration differently in different soils. The mechanisms
probably act simultaneously and are affected by or respond to management parameters
such as on-time and off-time. The effectiveness of surge irrigation depends on the
interaction between management and physical soil parameters, such as soil bulk density,
porosity, hydraulic conductivity, and others. The fundamental question is, for a given soil
type, what response can be predicted? Or what infiltration reduction can be predicted?
The answer is that multiple surge mechanisms occur simultaneously in a given situation
or field condition. The prediction of surge irrigation performance under a set of field
conditions and management practices seems difficult. In any irrigation event, a
combination of different mechanisms bnngs about the surge effects. Killen (1987)
concluded that surface sealing and consolidatIOn were not satisfactory explained
improved advance times in sandy soils.

Infiltration is the process of water entry into the soil matrix through the soil
surface. Infiltration is a complex physical process, and difficult to describe for real
anisotropic and heterogeneous conditions. Infiltration in surge Inigation is the most
difficult to predict or model, due to the on-time and off-lime combination of water
application. The relationship between on-time and off-time, and infiltration is unknown
for most soils. The infiltration function under surge also varies with different soil
conditions (Alemi, 1988) and is difficult to predict (Kemper et aI., 1988). Surge
infiltration has been shown to be influenced by the intake opportunity time, off-time
(Saleh and Hanks, 1989), wetted perimeter (Walker and Willardson, 1983; Samani,
19x3), spatial location (Wallander and Bautista, 1983), and cycle times and cycle ratio
(Blair, 1984). Other factors include sediment transport, furrow stream velocity, and
chemical and biological conditions. The IIlfiltration rate is important because it is the
dominant factor governing the performance of a surface irrigation system.
Studies show that the infiltration rate under intermittent applications of water is
less than that under continuous applications of water. Levin and Rooyen (1977) found
that the horizontal flow of pulsed-in trickle irrigation was 13 to 32 percent greater than
a continuous non-ponding application rate. Malano (1982) found that intake rates under
intermittent application decreased more rapidly than in continuous application. Testezlaf
et a1. (1987) found that surge flow caused a 33 to 66 percent greater reduction in
quasisteady infiltration rate than continuous flow.

3- Irrigation Management

Physical and chemical methods in combination have proved to be the most

effective approaches to solving water infiltration problems. However, these require
extensive and continuing annual investment in both time and money to be effective.
Many users try to complement these methods with irrigation practices to make the
water infiltration problem easier to solve or manage. Several practices are discussed
here.

i. Irrigating more frequently is a simple and effective approach especially for

soils having an initially high infiltration rate but for which the rate drops rather
quickly due to low salinity or high SAR. The objective is to supply the crop
with adequate water at all times without secondary problems developing
(waterlogging, poor aeration).

Irrigating more frequently maintains a higher average soil water content and
reduces the possibility of a water stress that might result if irrigations were
spaced further apart. If the crop is not stressed for water between irrigations,
increasing the irrigation frequency does little good.

ii. Pre-plant irrigation can be relied upon to fill the rooting depth to field capacity
at a time when there is little chance of causing crop damage. In some difficult
soils a pre-plant irrigation is the only opportunity to wet the deeper part of the
crop root zone. It is also an effective method for wetting soils with a very slow
infiltration rate.
iii. Extending the duration of an irrigation applies more water and is beneficial
provided that soil aeration, waterlogging, runoff and surface drainage
problems do not result. Many irrigators try to extend the irrigation by reducing
the volume of flow to a field and holding the water on the field for a greater
period of time. Careful management and monitoring is needed to maintain
water use efficiency and to keep runoff to a minimum. Excessive runoff is
frequently collected in a pond at the low side of the irrigated field and is
pumped back up slope through a pipeline to be re-circulated into the irrigation
stream. These recirculation (return-flow) systems are becoming common in
surface irrigated areas and can aid greatly in efficiently irrigating a soil with a
low infiltration rate. In a few instances this system is installed following a
comprehensive land levelling or grading programme to improve water use
efficiency. By collecting and re-circulating water, both the total water use
efficiency and depth of penetration can be more easily controlled.
iv. Changing irrigation systems may be necessary on more difficult soils. For
instance, changing from a surface irrigation system to one which applies
 water more precisely (sprinklers for sandy soils and localized (drip) irrigation
for heavier clayey soils) may allow the user to approach the soil intake rate
more closely. These changes require large capital expenditures and
additional power to operate, but the system can be designed to apply water at
the rate desired. If runoff occurs with sprinklers or localized (drip) irrigation,
the application rate is too high. Changing the rate of application after
installation may be difficult and complete redesigning of the system may be
needed. In some cases an existing sprinkler or localized irrigation system can
be intermittently operated to match the infiltration rate more closely, stopping
irrigation at the time runoff begins and re-irrigating every few hours until the
desired depth of applied water is reached. This technique does allow the use
of an existing sprinkler or localized irrigation system, but will probably use a
little more water, thus increasing production cost, and it may also need more
investment in equipment to offset idle time.

Sprinklers apply water in droplets, some quite large. On impact, these large
droplets can disperse the soil surface particles and aggravate or cause an
infiltration problem accompanied by excessive runoff. Application rates
normally vary from 3 mm to 6 mm per hour over the irrigated area. Sprinklers
are well adapted to sandy and loamy soils but less so to heavy or clayey type
soils. Localized drip or trickle irrigation systems are better adapted to loamy
or clayey soils and apply water through many small outlets (emitters) at a rate
of 2 to 4 litres per hour. At these low rates they do not disperse the soil
particles as do sprinklers. They are less well adapted to sandy soils.

4- Snrge Irrigation Mechanism/

The advantages of surge irrigation are direct results of reduced infiltration rates. These
mechanisms by which reduced infiltration rates are produced include consolidation,
surface sealing, redistribution, air entrapment, cracking, and hysteresis. Trout (1990)
proposed that, while each of these mechanisms affect the infiltration rate, they are not
additive. For the purposes of this study, we will explore how these mechanisms are
affected by soil texture and whether or not they are additive.

4.1 Consolidation
Consolidation is defined as the volume change of the soil mass. Coolidge et al.
(1982) and Kemper et al. (1988) have demonstrated consolidation during surge irrigation
to be the volume change of the soil due to suction force during the off-time. Figure 2.1
illustrates a microscopic view of consolidation at the soil surface with surge irrigation
(Seymour, 1990).
The effect of consolidation is significant when it causes a substantial change in the
saturated hydraulic conductivity. Saleh and Hanks (1989) evaluated the effects of
consolidation on hydraulic conductivity and bulk density in field-simulated surge studies.
The soils tested were a silty clay loam, a silty loam, and a sandy loam. They found that
during the off-time of surge irrigation, consolidation occurred on the surface of silty loam
and sandy loam soils, significantly reducing the saturated hydraulic conductivity.
In their study, Saleh and Hanks used off-times of 10, 20, and 30 minutes for three
different discharges. On the silty loam, the reductions of the saturation hydraulic
conductivity were 28 percent, 40 percent, and 31 percent respectively, between the first
and second surge for the off-time. However, under continuous irrigation, the decrease
was only about six percent. On the sandy loam, the saturated hydraulic conductivity
reductions were 35 percent, 45 percent, and 55 percent respectively, while the reduction
under continuous irrigation was five percent. For both silty loam and sandy loam, then,
the reductions were significantly higher compared to continuous irrigation. On the silty
clay loam there was no significant change in the saturated hydraulic conductivity, as there
was no measurable compaction. Moreover, the effect was most pronounced during the
first off-time of water application, and mostly occurred during the first 10 to 20 minutes
of off-time. The model simulations predict a reduction of infiltration rate on the next offtime
caused by the reduction of saturated hydraulic conductivity during the first off-time.
The higher the reduction of saturated hydraulic conductivity, tile hIgher the reduction of
infiltration rate.
Samani (1983) found that consolidation with surge irrigation occurs in soils
where the pore pressure is changed due to pressure change from a positive to a negative.
The saturated hydraulic conductivity decreased by about 90 percent for the silty clay
loam soil that Samani used. He suggested that the amount of consolidation and change
in saturated hydraulic conductivity would vary according to soil type and depended on
the magnitude of the capillary pressure developed. Kemper et al. (1988) investigated
consolidation in tilled furrows, and concluded that consolidation was a significant
mechanism tied in with particle movement, which increases the density of a furrow bed,
thereby reducing the infiltration rate.

Seymour (1990) conducted laboratory experiments and computer simulations to

evaluate the combined effects of consolidation and air entrapment. Five surface layer
thicknesses were modeled. She found that soil consolidation had a limited contribution
to surge effects for the sandy loam soil. When water was applied from the top, bulk
density decreased, reversing the consolidation process as air bubbles rushed out through
the soil surface during wetting. This loosened the soil, as the trapped air had no other
escape. This phenomenon was observed by Farahani et al. (1993).
The fact that the surge effect is observed after the first off-time lends credence to
the idea that consolidation plays a part in creating the surge effect (Duke, 1988). It has
also been shown that as the off-time increases, the infiltration reduction also increases up
to an optimum off-time (AI-Saud et aI., 1993). This indicates that there are other
mechanisms which playa more important role than consolidation, and it is of interest to
us how these mechanisms work, and if they are additive. Trout (1990) noted that the net
effects of surface sealing and consolidation are not additive.

4.2 Air Entrapment

Air entrapment in surge irrigation is produced in two ways. First, air residing in
the soil pores may be completely isolated by water during rewetting. Thus the
discontinuous air phase creates entrapped air. The second type of entrapped
air is that which is trapped between the water layers (confined air) of successive surges
(Duke, 1988).
Air entrapment is dependent on soil pores, geometry, and water content. Entrapped air
occurs when some pores fill before others an affect of the distribution of water velocities
within a soil that has pores of varying size. Entrapped air has been shown to affect soil
moisture characteristics, saturated hydraulic conductivity, saturation, and infiltration
(Davidson et aI., 1966; Bond and Collis-George, 1981). There has been some work with
surge irrigation regarding air entrapment. Lep ( I % I ) conducted intermittently and
continuously ponded one-dimensional infiltration tests USlllg silty loam soil in
airreleased and air-tight soil columns. He found that both instantaneous and cumulative
intake were significantly lower in the air-tight than in the air-released columns,
apparently due to increased air pressures impeding the flow in the former case. He also
found that lower air pressures were found below the wetting front in the intermittently
ponded simulations than in the continuously ponded ones. The intermittently ponded test
had higher wetting front advance rates compared to' the continuously ponded. Lep
obtained higher intake rates and lower pressures ahead of the wetting front in both
columns when they were wetted intermittently than when wetted continuously. As Lep
admits, these results are inconsistent, which might imply that the soil air phase was not
in fact air tight, causing the increased infiltration (Lep, 1981).
 Seymour (1990) investigated the role of air entrapment in an intermittently
wetted sand in laboratory simulation. The flow simulations indicated that a change in
natural saturation due to entrapped air of 10 percent or greater, and an entrapped air depth
of at least 25 cm, would be necessary to give a significant step reduction in infiltration
rate. It was found in the laboratory columns that when sand was wetted intermittently,
initially from the bottom and then from the top each time, natural saturation decreased
(air entrapment increased) from the first to the second on-time. The experiment gave
mixed results when the samples were wetted from the top both initially and dunng the
wetting phases of intermittent flow. It was noted that when samples were wetted from the
top, a significant amount of entrapped air rushed through the soil surface during the
wetting, reducing the bulk density of the surface layer and reversing consolidation
(Farahani, 1991). Air entrapment is in fact a potential factor leading to a reduction in
infiltration rates under surge irrigation. It is more pronounced in sands than fine textured
soils.

4.3 Infiltration Function

The empirical function introduced by Kostiakov (1932) to describe the infiltration process
is the most useful equation in surface irrigation.

z = k t'

where
z = infiltration depth after time t [L],
t = time of infiltration, the opportunity time [T], and
k, a = empirical constants.

This power function indicates that the infiltration rate goes to zero after long infiltration
times. Since the infiltration time on the AI- Saud et al. (1993)
experiment was short, this equation fit the field data very well. A parameter for surge one
was used in the infiltration equation for surge two. We conclude that a was soil
dependent and k was management dependent. The infiltration rate does not approach
zero in the field, but goes to the steady state infiltration or basic infiltration rate.
Therefore, the next Equation is a modification of Kostiakov equation, which does not
approach zero at a long time and is widely used.

z=kt'+ct

where
z = infiltration depth after time t [L],
t = time of infiltration [T],
k, a = empirical constants, and
c = basic infiltration rate.

The infiltration functions applied to surge irrigation are variations of the

Kostiakov or modified Kostiakov equations. The infiltration rate of surge lITIgatIOn
changes during on-time and off-time periods, thus two infiltration equations are
necessary to describe infiltration under surge flow (Walker and Lee, 1981). For the
initially dry section of the furrow, the modified Kostiakov Equation 2.2 is used. Ihe
same basic equation is used for the previously wetted section, except for changes in the
empirical parameters k' and a', and the basic infiltration rate, c', yielding:

where
,
z = infiltrated depth after time t [L],
t = infiltration opportunity time [I),
a', k' = empirical parameters, differ from those of the initially dry furrow, and
c = basic infiltration rate for surge conditions.

As already discussed, basic infiltration rate under surge irrigation is usually less
than that under continuous flow. Podmore and Duke (1982) found that surge flow
produced steady state infiltration rates that were half of thos,) under continuous irrigation.
Izuno et aI., (1985) approximated infiltration under surge irrigation using a step drop
from the infiltration rate predicted by the Kostiakov equation for continuous flow. Ihis
result was based upon data from infiltration tests in a one-meter blocked furrow
infiltrometer. Ihey also developed an algorithm from Clemrnens (1981) branch function
to characterize infiltration rate under surge irrigation. Ihe infiltration functions suggested
are:

z= kta t< = t'

z = k't,a, + c'(t-t') t > t'

where:

z = infiltration depth after time t [L],

t = cumulative intake opportunity time [I),
t' = duration ofthe first surge (on-time) [I],
a', k' = empirical parameters different from those of the initially dry furrow, and
c = basic infiltration rate.