ارزیابی عملکرد روش های برآورد ضرایب معادله نفوذ کوستیاکف در آبیاری جویچه ای با جریان موجی

به منظور افزایش راندمان کاربرد و یکنواختی توزیع آب در روش‌های آبیاری سطحی، تکنیک‌هایی نظیر کاربرد جریان موجی توسعه داده شده است. عملکرد مناسب آ؛بیاری موجی مستلزم تعیین دقیق پارامترهای معادله نفوذ آب در خاک می‌باشد. هدف از تحقیق حاضر بررسی عملکرد سه روش دو نقطه‌ای الیوت و واکر، یک نقطه‌ای شپارد و روش رگرسیون غیرخطی در تخمین نفوذ آب در آبیاری جویچه‌ای با جریان موجی می‌باشد. به همین منظور از داده‌های ارزیابی مزرعه‌ای سه جویچه آزمایشی شامل دو جویچه با جریان موجی و یک جویچه با جریان پیوسته استفاده شد. طول جویچه‌ها 150 متر، فواصل آن‌ها 0.75 متر و بافت خاک (لوم رسی) بود. نتایج نشان داد که روش یک نقطه‌ای ضعیف‌ترین عملکرد را در برآورد نفوذ در جویچه‌های با جریان موجی و همچنین جریان پیوسته داشت. از سوی دیگر نتایج دو روش رگرسیون غیر خطی و دونقطه‌ای قابل قبول و نزدیک به هم بدست آمد. مقادیر شاخص جذر میانگین مربعات خطا روش‌های یک نقطه‌ای، دونقطه‌ای و رگرسیون غیرخطی در پیش‌بینی زمان پیشروی جویچه‌های با جریان موجی به ترتیب 10.85، 2.52و 2.46دقیقه و در جریان پیوسته به ترتیب 8.78، 6.57و 4.77دقیقه بدست آمد. همچنین مقادیر میانگین خطای نسبی روش‌های مذکور در برآورد حجم آب نفوذ یافته در جویچه‌‌های با جریان موجی به ترتیب42.12، 3.51 و 4.79 درصد و در جریان پیوسته به ترتیب 7.70، 6.33 و 2.53 درصد برآورد شد. نتایج نشان داد، روش رگرسیون غیرخطی عملکرد بهتری در برآورد نفوذ در آبیاری جویچه‌ای با جریان موجی و پیوسته داشت.

evaluation of the performance of methods for estimating the coefficients of kostiakov infiltration equation in furrow irrigation with surge flow

background and objectives surface irrigation is known as the most common method of irrigation for reasons such as low initial cost and low energy consumption. the main problems of surface irrigation methods are low water application efficiency and distribution uniformity. in order to remove these restrictions, some techniques such as the application of surge flow have been developed. however, these new techniques will also be successful if they are designed correctly. one of the requirements for proper and efficient design of surge irrigation is to accurately determine the infiltration equation parameters. due to continuous changes in the infiltration process during on-off cycles in surge irrigation, determining the empirical equation of infiltration in this method is complex. no study has been performed to evaluate the proposed methods to estimate the empirical coefficients of the kostiakov infiltration equation in furrow irrigation under surge flow. the purpose of the present study was to investigate the performance of three methods: elliott and walker two-point, shepard one-point and nonlinear regression methods for estimating water infiltration in furrow irrigation with surge flow. methodology the field evaluation data of three experimental furrow included two furrows with surge flow and a furrow with continuous flow were used. the experiments were performed in two irrigation events. the length and spacing of the furrows was 150 m and 0.75 m, respectively. the soil texture was clay loam. in all experiments, furrows were free-drainage without a return runoff system. surge flow was applied during the advance phase and was converted to continuous flow when the water flow reached the end of the furrows. the longitudinal slope of the farm was variable; the slope of the initial 60 m was about 0.0088 m m-1 and the rest of the furrows' length was approximately 0.0034 m m-1. in each experiments, for the first surge (dry phase), the empirical coefficients of the kostiakov equation were determined using elliott and walker two-point, shepard one-point and nonlinear regression methods. for second and later surges, these parameters were estimated by crtm model (proposed by blair and smerdon). to evaluate the performance of the methods, simulations were performed using winsrfr software and the results including advance times, the volume of infiltrated water and runoff losses were compared with the observed values. findings the mean values of root mean square error index for the one-point, two-point and nonlinear regression methods in estimating the advance times in furrows with surge flow were calculated to be 10.85, 2.52 and 2.46 min, respectively, and in furrow with continuous flow were 8.87, 6.57 and 4.77 min, respectively. also, the values of mean relative error for estimating the volume of infiltrated water in the furrows with surge flow were 42.12, 3.51 and 4.79%, respectively, and for continuous flow were 7.70, 6.33 and 2.53%, respectively. the values of mean relative error for prediction of runoff losses in experiments with surge flow were calculated 100, 12.80 and 17.98% (first irrigation event) and 93.44, 4.95 and 4.85% (second irrigation event), respectively. in experiments with continuous flow, the values of this index were obtained 11.11, 14.81 and 22.22% in the first irrigation event and 16.16, 27.45 and 88.5% in the second irrigation event, respectively. results showed that the prediction error of the advance time and volume of infiltrated water increased in the second surge and later one due to the surge effect phenomenon. on the other hand, comparing the results of two irrigation events revealed that the accuracy of methods in the second irrigation event was higher than the first. this may be due to hydraulic stabilization of the furrows bed. the surge flow reduced the total water applications compared to continuous flow, so that the ratio of water application depth in