Ehsan Tavakoli; Behzad Ghorbani; m r; h s; b gh
Abstract
Increasing conveyance efficiency and reducing water loss in distribution canals are very important. The current study aimed at investigating and comparing empirical equations and SEEP/W model for estimating seepage of earth channels. For this study, the earth canal of Boldaji, constructed in loamy soil ...
Read More
Increasing conveyance efficiency and reducing water loss in distribution canals are very important. The current study aimed at investigating and comparing empirical equations and SEEP/W model for estimating seepage of earth channels. For this study, the earth canal of Boldaji, constructed in loamy soil and located downstream of modern irrigation and drainage network of Gandoman and Boldaji, Chaharmahal and Bakhtiari province, was chosen. Using dimensional similitude equations and considering 0.13 as scale ratio, the dimensions and discharges of the mentioned channel were applied to the laboratory model. In this study, 9 discharges (40-161 L/s) were converted to the applicable discharges of the model. Experiments carried out for 4 different water table depths and trapezoidal and triangular cross sections with 3 replications. Furthermore, by employing modified empirical models of Moritz, Ingham, India, Molesworth and Yennidumia, Offengenden, and Davis-Wilson, besides SEEP/W model, the amounts of seepage in the lab model were estimated. The results showed that for all investigated conditions and considering R2, r, RMSE, and MAE, Moritz presented the closest and the most accurate estimates of seepage (R2=0.992, r=0.996, RMSE=0.48 and MAE=0.44 lit/m2/min). Moreover, due to poor results of Molesworth and Yennidumia and Indian equations, they are not suggested for the study area.Although SEEP/W demonstrated appropriate performance for trapezoidal cross section, it did not show promising results for the triangular one. Considering the lower costs and shorter time as the results of utilizing dimensional analysis, it is recommended for controlled laboratory conditions in other similar regions.
a f; b gh
Abstract
Distribution uniformity of water is one of the most important parameters for evaluation of irrigation systems. Numerous equations have been developed to calculate distribution uniformity coefficient in sprinkler irrigation systems. The provided equations do not necessarily yield the same results in calculation ...
Read More
Distribution uniformity of water is one of the most important parameters for evaluation of irrigation systems. Numerous equations have been developed to calculate distribution uniformity coefficient in sprinkler irrigation systems. The provided equations do not necessarily yield the same results in calculation of uniformity coefficient for one specific farm in a particular condition. The aims of this study were the evaluation of various equations proposed by different researchers, investigation of the effects of different field conditions on the results, investigation of the existing relationships among outcomes of the equations, and feasibility of using these equations. For this purpose, distribution uniformity coefficients were calculated using the equations presented by Christiansen, Hawaiian Cane Society Specialists Hart and Reynolds, Wilcox and Swailes, Karmeli, Criddle et al, Benami and Hore, and Beale and Howell, for 10 sprinkler irrigation systems in Shahrekord fields. Data analysis was performed using Statistical Analysis System (SAS) Software in a randomized complete block design. The results indicated that there were significant differences (P< 0.05) between the aforesaid coefficients and some of these equations estimated a negative number or a number larger than 100% for coefficient of uniformity in some specific conditions of sprinkler irrigation. Furthermore, in all cases, beta distribution could estimate the water distribution uniformity coefficient (CU) better than the normal and uniform distributions. In all three beta, normal, and uniform distributions, minimum agreement between predicted and measured values was related to equations by Karmeli, Beall and Howell, and Benami and Hore; and the most agreement, particularly in the beta and normal distribution, was related to the Christiansen equation. The results conclusively indicated that a number of coefficients of uniformity such as Benami and Hore, Karmeli, and, to some extent, Beal and Howell coefficients are strongly dependent on specific field conditions and are not applicable under other field conditions.