JavaScript is disabled for your browser. Some features of this site may not work without it.
| dc.contributor.author | Tadele, Menwagaw | |
| dc.date.accessioned | 2020-06-12T06:50:56Z | |
| dc.date.available | 2020-06-12T06:50:56Z | |
| dc.date.issued | 2020-02 | |
| dc.identifier.uri | http://hdl.handle.net/123456789/11037 | |
| dc.description.abstract | An accurate flow rate measurement is crucial to improve the performance of irrigation systems by allocating the desired amount of irrigation water to the right irrigation system components. This study was aimed to develop alternative approaches to estimate the water delivered to quaternary canals in data scarce environments. It was conducted at Koga Irrigation Scheme at 6 blocks out of 12 available irrigation blocks namely; at Chihona, Kudmi, Adibera, Tagel, Andinet and Teleta irrigation blocks during the irrigation season of 2019. An optical smartphone application device entitled ‘DischargeApp’ was evaluated on its applicability to measure canal flow rate in comparison to a 90-degree v-notch weir method at selected quaternary canals. Moreover, water delivery performance of quaternary canal outlets was assessed by using three performance indicators: adequacy (Pa), equity (Pe) and reliability (Pr) indicators. Finally, the application of seven Machine learning models to estimate discharge at quaternary canal outlets were evaluated using five input variables: water level per unit width(h), irrigated area ratio of outlets(a), distance of outlets from TC off-take(l), Manning roughness coefficient of TC canal(n) and ranking order of operated outlets along TC(r). The Accuracy of the DischargeApp at field conditions with the flow rates range 15-60 l/s, was improved by changing the surface velocity correction factor. The mean discharge deviation and percent error were reduced from ±3.8 l to ±2.1 l/s and 11.5 to 7.1% respectively. The discharge observations lied within ±15 percent were also increased from 66 to 92.1 percent. The water delivery performance at quaternary canals showed that there was a significant flow variation among quaternary canal outlets in terms of water supply adequacy, equity and reliability. On average, the water supply through head, middle and tail canal outlets were 2.08, 1.84 and 1.74 l/s/ha respectively. At block level, water supply adequacy performance of canal outlets was good (0.9 -1.1) at the two head reach blocks (Kudmi and Chihona) and one tail block (Andinet) while the two middle reach blocks (Adibera and Tagel) showed poor adequacy status (Pa<0.7&>1.1) because of surplus water use. Teleta, which is the most tail reach block scored fair adequacy performance (0.7-0.9). Despite of its adequacy problem, Tagel scored good equity (Pe<0.1) and reliability (Pr<0.1) performances whereas, Teleta block scored poor equity (Pe> 0.25) and good reliability performances. Based on prediction performance and model equation interpretability, Multivariate adaptive regression splines (MARS) was selected to predict discharge at quaternary canal outlets. The performance of MARS at training and testing stages were (R2 = 0.86 & RMSE=3.6) and (R2=0.85& RMSE=3.48) respectively. Since the distance parameter was eliminated due to its zero regression coefficients, the developed MARS equation uses four variables to predict discharge. | en_US |
| dc.language.iso | en | en_US |
| dc.subject | Engineering Hydrology | en_US |
| dc.title | Improving Water Allocation across Canal Outlets Using Irrigation Performance Indicators and Machine Learning Algorithm: Case Study of Koga Irrigation Scheme | en_US |
| dc.type | Thesis | en_US |
