Surge flow irrigation: field experiments under short dimension field conditions in egypt

Surge flow irrigation: field experiments under short dimension field conditions in egypt Irrigation gravitaire par vagues: expériences dans les conditions de parcelles de dimensions réduites en Egypte Saleh. M. Ismail ABSTRACT Surge flow irrigation is a surface irrigation method that can be used to improve the efficiency of water applied by furrows. Several studies have shown that surge flow irrigation offers the potential of increasing the distribution uniformity, thereby increasing the efficiency of surface irrigation. Most of these studies were conducted in fields with long furrows, but the effects of surge flow in short fields, like in Egypt, are still not well known. In order to investigate the effect of surge flow irrigation in short fields a series of experiments have been carried-out in two different locations in Egypt. The first location with a clay soil was situated at the Agriculture Experimental Station, Assiut University, Assiut. The second location with a sandy soil was situated at the Assiut University Experimental Station for Desert Land, El-Wadi El-Assuity, Assiut. The furrow length was 7 m and the furrow width was.7 m for both locations. The furrows had a blocked end. Three discharges were selected for each soil type, namely.46,.74 and.9 l/s for clay soil and.73, 1. and 1.4 l/s for sandy soil. The water was conveyed via siphons to the furrows. For each discharge two cycle times were investigated, namely 16 and 24 minute. For each cycle time three cycle ratios were chosen, i.e. 1/4, 1/2 and 3/4 for 16 minute and 1/3, 1/2, and 2/3 for 24 minute cycle time. The different cycle ratios were applied to study the effects of off-time on the water distribution along the furrow. The water content was recorded by a Profile-probe at three locations, namely at the beginning, middle and end of the furrow. In each location three points were measured in a vertical at a depth from -.1,.1 -.3 and.3 -.7 m-surface. The results show that surge flow irrigation leads to a decrease in advance time compared to continuous flow. The reduction in advance time is more pronounced for high than for low discharges and also more in coarse than in fine textured soils. For both cycle times the advance time reduces compared to continuous flow in both soil types, except for.46 l/s in clay soils; this discharge leads to an increased advance time. For the other cases the reduction was more pronounced for a cycle time of 24 minute than for 16 minute. This reduction was due to the effect of off-time. When the off-time is long enough to infiltrate all the water before the second surge starts, the mechanism of surge flow works effective. The water content along the furrow is also more uniformly distributed than for continuous flow. In conclusion surge flow irrigation under the prevailing conditions in Egypt decreases the advance time, increases the efficiency and uniformity, and hence, it saves water. RESUME L irrigation gravitaire par vague est une méthode d irrigation de surface qui peut être utilisée pour améliorer le rendement de l eau appliquée par rigoles d infiltration. Plusieurs études ont démontré que l irrigation gravitaire par vague permet une distribution plus uniforme augmentant ainsi le rendement de l irrigation de surface. La plupart de ces études ont été entreprises dans des parcelles a grandes dimensions mais les effets de l irrigation gravitaire par vague dans parcelles de dimensions réduites, comme c est souvent le cas en Egypte, sont encore mal connus. Dans le but d étudier les effets de ce type d irrigation dans les parcelles de dimensions réduites, une série d expériences a été réalisée en 1

deux endroits différents en Egypte. Le premier emplacement consiste en un terrain argileux situé dans la station d expériences agricole de l université d Asyut. Le second consiste en en terrain sablonneux de la station expérimentale des zones désertes de l université d Asyut. Les rigoles ont une longueur de 7 m et une largeur de.7 m dans les deux emplacements. Les rigoles sont a extrémité fermée. Trois débits d eau ont été adoptés pour pour chaque type de sol, à savoir.46,.74 et.9 l/s pour le sol argileux et.73, 1. et 1.4 l/s pour le sol sablonneux. L eau est acheminée vers les rigoles moyennant des siphons. Pour chaque débit, deux durées de cycles ont été étudiées, à savoir 16 et 24 minutes. Pour chaque durée, trois rapports de cycles ont été choisis:1/4, 1/2 et 3/4 pour le cycle de 16 minutes et 1/3, 1/2 et 2/3 pour le cycle de 24 minutes. Les différents rapports de cycles ont été appliqués dans le but d étudier les effets de l off-time sur la distribution de l eau le long des rigoles. La teneur en eau a été enregistrée par l intermédiaire d une sonde dans trois endroits, au début, au milieu et à la fin de la rigole. Dans chaque endroit, les mesures ont été relevées en trois points sur une même verticale à des profondeurs de.1,.1.3 et.3.7 m de la surface. Les résultats montrent que l irrigation gravitaire par vague conduit à la réduction du temps d avancement comparé à celui d un débit continu. Cette réduction est plus prononcée pour les forts débits que pour les faibles débits. Elle est aussi plus prononcée pour les sols à texture grossière que pour les sols a texture fine. Pour les deux durées de cycle, le temps d avancement est réduit comparé à celui du débit continu pour les deux types de sol excepté pour le débit.46 l/s et pour le sol argileux; dans ce cas, ce débit conduit à une augmentation du temps d avancement. Pour les autres cas, la réduction est plus prononcée pour un durée de cycle de 24 minutes que pour 16 minutes. Cette réduction est due à l effet d off-time. Lorsque l off-time est suffisemment longue pour laisser infiltrer toute l eau avant le commencement de la seconde vague, le mécanisme d irrigation par vague est efficace. La teneur en eau le long des rigoles est aussi mieux uniformément distribuée que dans le cas du débit continu. En conclusion, l irrigation gravitaire par vagues dans les conditions régnant en Egypte diminue le temps d avancement et augmente l efficacité et l uniformité. Par conséquent, elle économise l eau. INTRODUCTION Surge flow irrigation is the intermittent application of water to furrows or borders in a series of relatively short on- and off-time periods, which usually vary from about 5 minutes to several hours. With this technique water is applied intermittently and not continuously as in conventional surface irrigation. The main objective of surge flow irrigation is to improve the efficiency by reducing deep percolation and runoff losses and to obtain a uniform wetting of the root zone. There are two characteristics of surge flow that save irrigation water. When water is admitted to the furrow for a given period of time, and then shut off to allow the furrow to de-water, the intake rate of the furrow is reduced. Thus, when the second surge of water is admitted, less water infiltrates into the soil than would otherwise occur, hence more water is available in the dry parts of the furrow, and the advance is more rapid. The combined effects of reduced infiltration during the advance phase plus a more rapid advance lead to a more uniform distribution of water along the furrow. In some soils, the same quantity of water normally required to get the water to the end of one furrow can be spread out over two furrows with surge flow. Thus, the uniformity of application is significantly improved. Field experiments The experiments were carried out at two different locations. The first location with a clay soil was situated at the Agriculture Experimental Station, Assiut University, Assiut. The second location with sandy soil was situated at the Assiut Univeristy Experimental Station for Desert Lands, El-Wadi El- Assuity, Assiut. The latter is a new place under reclamation and it was the first time that water was applied to this land. A description of the soil characteristics is presented in table (1). The furrow length in the experiments was 7 m and the width was.7 m, for each soil type. The slope was.24 m/m

in clay soil and.4 m/m in sandy soil. The furrows had a blocked end. To monitor the advance and recession time, five points were established along the furrows, namely at L, ¼ L ½ L, ¾ L, and 1 L. The distance between two consecutive points was 17. 5 m. To study the effect of surge flow irrigation on the water distribution along the furrow, the soil moisture content was measured at three locations, namely at the beginning, middle and end of the furrows. In each location three points in a vertical were measured at a depth from -.1 m,.1 -.3 m and.3 -.7 m below surface. To investigate the water content distribution in the cross-sections, one furrow in each soil type was selected for the measurement of the water content at three different verticals in that cross-section, namely at the furrow bottom, at the middle of the side slope and on the outside ridge. The water content in the soil was measured by a probe, which was lowered in fiberglass tubes in the soil. The tubes fitted exactly the profile probe and they were calibrated to measure the water content. During installation care was taken that the tubes were in good contact with the soil particles in order to get reliable values. To improve the reliability of the measurements each test covered three furrows, but the actual measurements were collected only from the middle one. Table 1 Soil physical characteristics for the experimental station at Assiut University, Egypt (Caractéristiques physiques du sol de la station expérimentale de l université d Asyut) Characteristic In the campus Sand Clay % % Silt % In the desert Sand % Clay % Particle size distribution 22.7 53. 24.3 87.2 5.3 7.5 Texture grade Clay Sandy Wilting point % 27.8 1.9 Field capacity % 48.8 15.2 Saturation % 52. 2. Sat. hydraulic cond. m/day.432 1.8 Bulk density kg / m 3 1,2 1,6 Discharge I l/s.46,.74 and.9.73, 1. and 1.4 Cycle time in min 16 and 24 16 and 24 Cycle ratio in % 1/4, 1/2 and 3/4 1/3, 1/2 and 2/3 The profile probe generated a 1 MHz signal, a signal similar to a FM radio. The signal was applied to three pairs of stainless steel rings that produce an electromagnetic field, which extended 1 mm into the soil. The electromagnetic field passed easily through the tube walls, but passed less easily through any air gaps. The water content of the soil surrounding the rings determined its dielectric properties (a measure of a material's response to polarisation in an electromagnetic field. Water has a dielectric constant of about 81, soil of about 4, and air of about 1). If the dielectric properties of the soil were different from the probe, some of the 1 MHz signals were reflected. The reflected part combined with the applied signal formed a standing wave and acted as a simple, sensitive measure of soil moisture content. Moreover, a gravimetric method was used in the sandy soil because of the gravel, which hampered the installation of the fiberglass tubes. The treatments for surge flow irrigation during this investigation were related to the objectives of this research and have been chosen on basis of water management guidelines as presented in the literature. The treatments are presented in table (1) and included the following factors: discharge (Q), cycle time and cycle ratio. Silt %

Water content distribution RESULTS AND DISCUSSION Figure (1) shows the average water content for continuous and surge flow irrigation in clay and sandy soil. Based on the figure, three levels of water distribution are presented in table (2). The first level in the table is "homogenous", which means that the distribution of the moisture content at each point along the furrow is the same and gives a horizontal line in the graph. The second one is "slightly homogenous" which means that there is a little difference in water distribution along the furrow, but the differences are not significant. The third level is "not homogenous" which means that the distribution along the furrow has significant differences and it appears in the graph as a curved line. Most of the results exhibit a good distribution for the water content along the furrow. When the discharge increases the number of surges decreases and range from 2 to 3 surges until the water reaches for the first time the end of the furrow. This small number of surges (2-3) gives still a very good water distribution uniformity. These results may be due to the specific conditions, namely short furrows with a blocked end. Figure (2) shows the water distribution in the cross section for the treatment of 1/3 cycle ratio and 24 minute cycle time for a discharge of.9 l/s in clay soils. The figure shows the difference between the distribution before and after irrigation at the beginning, middle and end of the furrow. The figure clearly shows that the change in water content (in volume percent) at the soil surface is 11.5, 1.5, 9.% at the beginning, middle and end of the furrow respectively. Meanwhile, it is 4.5, 3.5 and 4% in the lowest part of the vertical (.3 -.7 m). This result shows that surge flow irrigation under the conditions that prevail in Egypt tends to distribute the water more uniformly than continuous flow. Similar results were found by El-Dine and Hosny (2) and Tabuada et al., (1995). Effect of discharge Figure (3) shows the effect of the discharge on the advance time for continuous and surge flow irrigation in clay and sandy soil. In general, an increase in discharge leads to a decrease in advance time and surge flow treatments show to work good under high discharges in both soil types. For the discharge of.46 l/s the advance time in all surge treatments was longer than for continuous flow in clay soil. Meanwhile, the advance time in surge flow was shorter than for continuous flow for a discharge of.74 and.9 l/s, respectively. The fastest advance time to reach the end of the furrow was observed for the treatment of 1/3 cycle ratio and 24 minute cycle time for.74 l/s and for the treatment of 1/2 cycle ratio and 24 minute cycle time for.9 l/s. In sandy soil all advance times for the surge treatments were shorter than for continuous flow for all discharges, except for the treatment of 3/4 cycle ratio and 16 minute cycle time for a discharge of 1. l/s, which gave a longer advance time than continuous flow. These outcomes indicate that surge flow irrigation under the conditions that prevail in Egypt leads to a shorter advance time in comparison to continuous flow. The results are due to a net reduction in the infiltration rate. The soil bulk density increased as the soil was affected by consolidation during the off-time, which accompanies the intermittent water application. As the soil bulk density increases, the hydraulic conductivity decreases and consequently, the infiltration rate decreases (Samani et al., 1985). Moreover, the development of tension forces in the soil following drainage of water on the surface will consolidate the surface layer. Another mechanism for the improvement, being a shorter advance time, is the surface sealing. As water infiltrates after the first pulse, lubricated particles in the surface may be reoriented horizontally and in a plate fashion that will greatly reduce infiltration in the wetted section of the furrow (Allen, 198). The amount of consolidation and soil sealing of previously wetted soils during the surge off-time is influenced by the soil properties prior to irrigation, such as soil structure and texture, bulk density, degree of saturation and organic matter. The reduction in advance time is more pronounced for high than for low discharges and also more in coarse than in fine textured soils. In a few cases surge flow tends to increase the advance time, for example for a discharge of.46 l/s (figure 3a). The increase in advance time may be either due to the increase in infiltration time

during the off-time or due to the increase of the hydraulic gradient during the off-time which leads to an increased intake rate of the soil. The increase of advance time in the treatment of 3/4 cycle ratio and 16 minute cycle time for the discharge of 1. l/s (figure 3b) in sandy soil may be due to the effect of the gravel in some parts of the furrow which leads to a higher infiltration rate; consequently, it takes a longer time to reach the end of the furrow.

Water content in vol. % 2 15 1 5 Q =.46 l/s A Ct = 16, R = 3/4 Ct = 24, R = 2/3 1 2 3 4 5 6 7 2 15 1 5 Q =.73 l/s 1 2 3 4 5 6 7 B Ct = 16, R = 3/4 Ct = 24, R = 2/3 Water content in vol. % 2 15 1 5 Q =.74 l/s Ct = 16, R = 1/4 Ct = 16, R = 3/4 Ct = 24, R = 1/3 Ct = 24, R = 2/3 1 2 3 4 5 6 7 2 15 1 5 Q = 1. l/s Ct = 16, R = 3/4 Ct = 24, R = 1/3 Ct = 24, R = 2/3 1 2 3 4 5 6 7 Water content in vol. % 2 15 1 5 Q =.9 l/s Ct = 16, R = 1/4 Ct = 24, R = 1/3 1 2 3 4 5 6 7 2 15 1 5 Q = 1.4 l/s Ct = 16, R = 3/4 Ct = 24, R = 1/3 1 2 3 4 5 6 7 Length along furrow in m Length along furrow in m Figure 1 Change in water content in clay and sandy soils (A and B respectively) for continuous and surge flow irrigation with different cycle times and ratios for thee discharges. (Variation de la teneur en eau dans les sols argileux et sablonneux, dans le cas, de l irrigation continue et par vagues pour différentes rapports et durées de cycles et pour trois différentes débits) 6

Beginning of the furrow Middle of the furrow End of the furrow.1.2.3.4.5.6.7.1.2.3.4.5.6.7.1.2.3.4.5.6.7 Figure 2 Water distribution for.9 l/s, cycle time 24 minute and cycle ratio 1/3 at the beginning, middle and end of the furrow in clay soil, (depth and distance in m, moisture content in vol. %). (distribution de l eau pour un débit de.9 l/s, durée du cycle 24 minutes et rapport de cycle 1/3, au milieu et à la fin de la rigole en sol argileux ( profondeur et distance en m, teneur en eau en %)) Effect of cycle time Figure (3) show the advance phase for surge flow with two cycle times and for continuous flow, both in clay and sandy soils. The conclusions of these figures are presented in table (2), which vary for the discharge and soil type. For the clay soil (figure 3a) the alternatives of surge flow with a discharge of.46 l/s led to a longer advance time than continuous flow. The figure shows that the treatment 1/2 cycle ratio and 16 minute cycle time did not reach the end of the furrow. Also, the surge treatments of 3/4 and 2/3 cycle ratio and 16 and 24 minute cycle time, respectively were 1.6 to 1.2 times longer than the advance time for continuous flow. Meanwhile, the advance time of the treatment of 1/2 cycle ratio and 24 minute cycle time was almost the same as for continuous flow. For the discharge of.74 and.9 l/s the surge flow works properly and reduced the advance time in all cases. The best improvement was with the discharge of.74 l/s; for that discharge the advance time for continuous flow was 1.2 to 2. times longer than for surge flow. The fastest treatment was 1/3 cycle ratio and 24 minute cycle time. The discharge of.9 l/s required for continuous flow an advance time that was 1.2 to 1.6 times longer than for the surge flows. The surge flow with 1/2 cycle ratio and 24 minute cycle time was the fastest one to reach the end of the furrow. In sandy soil (figure 3b) the advance time for the surge treatments and for all discharges was shorter than for continuous flow, except for a surge of 3/4 cycle ratio and 16 minute cycle time with a discharge of 1. l/s, which increased the advance time compared to continuous flow. The effect varied for each discharge. For a discharge of.73 l/s the advance time for continuous flow was 1.8 to 1.2 times longer than for surge flow. For the discharge of 1. l/s the advance time was 1.1 to 1.4 time longer than for surge flow; meanwhile, the time was 1.3 to 1.7 times longer for the discharge of 1.4 l/s. For that discharge the treatment 1/2 cycle ratio and 16 minute cycle time resulted in the same advance time as continuous flow. In general surge flow irrigation under the conditions that prevail in Egypt tends to decrease the advance time as shown by most of the experiments. The most effective surge during the advance phase was for the discharge of.74 l/s and a 1/3 cycle ratio and 24 minute cycle time in clay soil (figure 3a). Increasing the discharge to.9 l/s for the same soil led to the same, but smaller effects than for the discharge of.74 l/s. The fastest surge for this discharge was 1/2 cycle ratio and 24 minute cycle time. The reduction in the advance time was due to one or more of the mechanisms by which surge irrigation reduces the advance time as reported by Humpherys, (1989a), Kemper et al., (1988) and Samani et al., (1985). The only discharge, which increased the advance time compared to continuous flow, was the discharge of.46 l/s in clay soil (figure 3a). 7

1 A 1 B Time in min 8 6 4 Ct = 16, R = 3/4 Ct = 24, R = 2/3 8 6 4 Ct = 16, R = 3/4 Ct = 24, R = 2/3 2 Q =.46 l/s 17.5 35 52.5 7 2 Q =.73 l/s 17.5 35 52.5 7 1 1 Time in min 8 6 4 Ct = 16, R = 1/4 Ct = 16, R = 3/4 Ct = 24, R = 1/3 Ct = 24, R = 2/3 8 6 4 Ct = 16, R = 3/4 Ct = 24, R = 1/3 Ct = 24, R = 2/3 2 Q =.74 l/s 2 Q = 1. l/s 17.5 35 52.5 7 17.5 35 52.5 7 1 1 Time in min 8 6 4 2 Ct = 16, R = 1/4 Ct = 24, R = 1/3 17.5 35 52.5 7 Length along furrow in m Q =.9 l/s Figure 3 Advance phase for different discharges in clay and sandy soils (A and B respectively), for continuous and for surge flow irrigation for two cycle time and different cycle ratios. (phase d avancement pour différents débits dons des sol argileux et sablonneux (A et B respectivement) dans les cas d irrigation continus et par vagues pour deux durées de cycles et différent rapports de cycles) 8 6 4 2 Ct = 16, R = 3/4 Ct = 24, R = 1/3 17.5 35 52.5 7 Length along furrow in m Q = 1.4 l/s

Table 2 Results of clay and sandy soils treatments related to continuous and surge flow irrigation. (résultats des traitements des sols argileux et sablonneux pour les cas de l irrigation continue et par vagues) Soil type Clay soil Sandy soil Discharge l/s.46.74.9.73 1. 1.4 Cycle time in min 16 24 16 24 16 24 16 24 16 24 16 24 Cycle ratio Initial water content in vol. % Number of surge Advance time in min Total inflow time Difference in inflow time between cont. and surge flow in min Difference in water supply Between cont. and surge flow in m 3 Change in water content in vol. % Cont. 3. 1 96 96 11.6 SH 1/2 33. 8 3/4 35.1 8 12 96 1.2 H 1/2 36.2 7 9 84-12.33 9.3 H 2/3 33.9 7 116 112 +16 -.44 1.7 H Cont. 29.7 1 75 75 11.6 NH 1/4 35.8 13 62 52-23 1. 9.8 SH 1/2 3.8 6 56 48-27 1.2 9.6 H 3/4 34. 4 54 48-27 1.2 9.8 H 1/3 35.8 4 37 32-43 1.9 5.2 H 1/2 37. 5 62 6-15.66 1.7 H 2/3 38.7 3 56 48-23 1. 9.3 H Cont. 34.9 1 45.7 45.7 1.2 H 1/4 35.6 9 44 36-9.7.52 8.2 SH 1/2 34.3 4 38 32-13.7.72 7.8 H 1/3 36.5 4 33.5 32-13.7.72 7.4 H 1/2 35.1 2 28 24-21.7 1.2 4.8 H Cont..3 1 78 78 11.7 SH 1/2.3 8 7 64-14.61 1.5 NH 3/4.3 6 72 72-6.26 11.6 SH 1/2.3 5 66 6-18.79 1 SH 2/3.3 4 65 64-14.61 1.1 NH Cont..3 1 55 55 11.3 SH 1/2.3 5 5 4-15.9 9.9 NH 3/4.3 6 7 72 +17-1. 14.3 SH 1/3.3 5 44 4-15.9 9.7 SH 1/2.3 3 39 36-19 1.1 7.5 NH 2/3.3 3 48 48-7.42 1.5 SH Cont..3 1 45 4.5 11.9 H 1/2.3 4 4 32-8.5.71 9.8 NH 3/4.3 2 24 24-16.5 1.4 8.4 NH 1/3.3 4 36 32 -.5.4 11.8 H 1/2.3 2 3 24-16.5 1.4 7.8 NH H = Homogeneous; SH = Slightly Homogeneous; NH = Not Homogeneous In clay soils, low discharges led to more infiltration and by introducing surge treatments to low discharges the available infiltration time was increased, especially during off-time; hence, more water Water content distribution along the furrow 2

infiltrates and less water advances, consequently the advance time in surge flow will be longer than with continuous flow. In sandy soil (figure 3b) all the discharges (.73, 1., and 1.4 l/s) for surge flow significantly reduced the advance time. The fastest treatment was 1/2 cycle ratio and 24 minute cycle time for the discharge.73 and 1. l/s, which can be explained by the off-time of 12 minutes. This long off-time gave the maximum effect by one or more of the previously mentioned mechanisms. For the discharge of 1.4 l/s the effect was less than that with the other discharges. The 3/4 cycle ratio and 16 minute cycle time was the fastest treatment to reach the end of the furrow. The results of sandy soil clearly show that surge flow worked better in coarse than in fine textured soils. Similar results have been reported by Yonts et al., (1996). In coarse textured soils (sandy soil) all surge treatments for all discharges reduce the advance time compared to continuous flow, except for the treatment with a discharge of 1. l/s and 3/4 cycle ratio and 16 minute cycle time, which increased the advance time compared to continuous flow. This increase may be due to two factors. Firstly, the soil in the newly reclaimed land contains 2% gravel and maybe the furrow in the experiment has a gravel layer, which increased the infiltration and consequently the advance time. Secondly, holes made by animals, which live in dry, especially in newly reclaimed soils, may have led to an increased infiltration in the subsurface layers. In this case the flow did not reach the end of the furrow as all the water was infiltrated and no water advanced. This situation clearly appeared for the lowest discharge of.46 l/s with 1/2 cycle ratio and 16 minute cycle time in clay soil, where the water did not reach the end of the furrow due to the phenomena called dead storage requirement. During the on-time, water infiltrated, filled the dead storage areas and cross-section of the furrow during the on-time. During the off-time, water infiltrated into the soil and the furrow was emptied. When the next pulse entered the furrow, it must refill the zones of surface ponding caused by furrow irregularities and roughness, before the flow can advance into the dry parts of the furrow. This dead storage requirement depends on the geometry, roughness, slope and length of the furrow. During the 8 minute pulses, a large percentage of the pulse is required for surface storage and infiltration, and little remains for the advance. This phenomenon is very important to give the maximum furrow length in relation to furrow stream size, cycle time and cycle ratio, Coolidge et al., (1982). Effect of cycle ratio The effect of cycle ratios on surge flow irrigation has been studied by comparing the 1/2 and 3/4 cycle ratio for 16 minute cycle time with the 1/3 and 1/2 cycle ratio for 24 minute cycle time. The 1/2 cycle ratio has been compared with 1/3 and the 3/4 cycle ratio has been compared with 1/2 respectively, because the mentioned pairs of cycle ratio have equal in on-time (8 and 12 minute respectively) but different off- time (8 and 16 minute for the first pair and 4 and 12 minute for the second pair). Table (3) presents the effect of off-time on the advance time in clay and sandy soils, respectively.

Table 3 Advance time in minutes for two cycle on-time and different cycle off-times for both soil types. (durée d avancement en minutes pour deux cycles on-time et différent cycles off-time pour les deux types de sols) Soil type clay sand Discharge l/s Cycle time in min 16 24 on-time=8 on-time=12 on-time=8 on-time=12 off-time 8 off-time 4 off-time 16 off-time 12 continuous.46 96 12 9.74 62 56 54 37 62.9 46 38 33.73 78 72 66 1. 55 5 7 44 39 1.4 41 4 24 36 3 The results indicate that the water for 8 minute on-time advanced faster for an off-time of 16 minute than for 8 minute off-time, in all soil types and for all discharges. That means for a 1/3 cycle ratio with a 24 minute cycle time the water advanced faster than for a 1/2 cycle ratio with 16 minute cycle time. When comparing 12 minute off-time with 4 minute off-time in all soil types under all discharges, the effect was not clear. In some cases the advance time was shorter under 12 minute off-time compared with 4 minute off-time such as: in the discharge of.46 l/s in clay soil and for the discharges of.73 and 1. l/s in sandy soil. A reverse behavior can be observed in some cases. For example the advance time decreased for 4 minute off-time compared to 12 minute off-time under all discharges in both soil types; such as in the discharge of.74 l/s in clay soil and the discharge of 1.4 l/s in sandy soil. For some conditions and within limits an increase of off-time decreased the infiltration rate and the advance time, which may be due to two mechanisms: a) the redistribution and development of the negative pressure in the soil slows down as the off-time continues; b) the consolidation rate of the soil decreases as the negative pressure increases. The intake rate decreased as off-time increased like for 16 minute off-time and some of the 12 minute off-time. Similar results were reported by Brown et al., (1988) and Samani et al., (1985). This does not mean that increasing the off-time will result in a faster advance of water in surge flow irrigation. Since the advance does not only depend on the reduction in intake rate of the wetted section, but also depends on the length of the dry advance for each surge, Samani et al., (1985). A suitable off-time is at least the time required to infiltrate the water completely before the next surge starts. CONCLUSIONS Surge flow irrigation under the conditions that prevail in Egypt leads a more uniformly water distribution than continuous flow. This uniformity is more pronounced in clay soil than in sandy soil. In surge flow irrigation the water advances in a shorter time to the end of the field than with continuous flow. The best results were obtained with the discharges of.74 l/s in clay soil and 1. l/s in sandy soil. The reduction in advance time for these flows was more pronounced than for the other discharges. The 24 minute cycle time led to more reduction in advance time than the 16 minute cycle time. A 16 minute off-time reduced the advance time more than the other cases of off-time (4, 8 and 12 minute), but the most suitable off time depends on the time required to infiltrate all the water before the next surge start. This off-time of 16 minute gave the maximum effect for the surge mechanism to work. Surge flow irrigation can work under short field conditions and gives acceptable results regarding to water saving. It is comparable to the results obtained in long field conditions. The evaluation of the field experiments and selection of the best treatments in view of application efficiency, water saving and crop production will be further investigated.

REFERENCES Allen, N.L., 198. Advance rates in furrow irrigation to cycled flow. Ms.c thesis, presented at Utah State Univ. Logan, U.S.A. Brown, M.J., Kemper, W.D., Trout, T. J. and Humpherys, A.S., 1988. Sediment, erosion and water intake in furrows. Irrig. Science, Vol. 9, pp. 45-55. Coolidge, P.S., Walker, W.R. and Bishop, A.A., 1982. Advance and runoff surge flow furrow irrigation. Journal of Irrigation and Drainage Division, ASCE. Vol. 18, No. IR1, PP. 35 42. El-Dine, T.G. and Hosny, M.M., 2. Field evaluation of surge and continuous flow in furrow irrigation systems. Water Resources Management, Vol. 14, PP. 77-87. Humpherys, A.S., 1989a. Surge irrigation: II An overview. ICID Bulletin. Vol. 38, No. 2, pp. 35-48. Kemper, W.D., Trout, T.J., Humpherys, A.S. and Bullock, M.S., 1988. Mechanisms by which surge flow irrigation reduce furrow infiltration rates in a silty loam soil. Transactions, ASAE. Vol. 31, No. 3, PP. 821 828. Samani, Z.A., Walker, W.R. and Willardson, L.S., 1985. Infiltration under surge flow irrigation. Transaction of ASAE, Vol. 28, No. 5, PP. 1539 1542. Tabuada, M.A., Rego, Z.J.C., Vachaud, G. and Pereira, L.S., 1995. Modelling of furrow irrigation. Advance with two-dimensional infiltration. Agricultural Water Management. Vol. 28, PP. 21-221. Yonts, C.D., Eisenhauer, D.E. and Fekersillassie, D., 1996. Impact of surge flow irrigation on furrow water advance. Transactions of ASAE Vol. 39 No. 3, pp. 973-979.