|
چکیده
|
A major step in designing a subsurface drainage system is calculation of drainpipe spacing. Nearly, all unsteady-state drainpipe-spacing equations use ratio of hydraulic conductivity (K_s) to drainable porosity (f) in the form of κ=K_s/f. This work developed a novel inverse model based on Bees Algorithm to estimate the κ. We examined the efficiency of the developed inverse model to appraise the κ using the existing data sets at laboratory and field scales. For each data set, we compared the performance of the developed inverse model with those of a dimensionless curve generated by the authors (generalized dimensionless curve), Skaggs’s (1976) dimensionless curve being proportional to the data set (Skaggs’s dimensionless curve), a drainpipe discharge method, and a direct method. The results showed that the best statistical indicator values were achieved for each data set when the drainage equations utilized the κ deduced from the developed inverse model. All the studied techniques performed the worst for Skaggs et al.’s (1973) data at 57.5 hrs. after drainage start. For this data set, the studied techniques were, respectively, ranked as the developed inverse model, Skaggs’s dimensionless curve, the generalized dimensionless curve, and the direct method. Even in this worst case, the global performance index was improved almost 93% when the drainage equation utilized the κ appraised by the developed inverse model, instead of that appraised by Skaggs’s dimensionless curve. The overall performance comparison of the three inverse techniques evaluated in this study revealed that the root mean square error and coefficient of determination related to the developed inverse model are the best ones, followed by those related to Skaggs’s dimensionless curve, and the developed inverse model, respectively. In summary, the results revealed that, compared to the other studied techniques, the developed inverse model is the most efficient technique to estimate the κ, which is promising for practical purposes at the laboratory and field scales.
|