Improvements to the functionality of the mycanesim® irrigation scheduling advice system for sugarcane.

Master of Science in Bioresources Engineering. University of KwaZulu-Natal, Pietermaritzburg 2016.

MyCanesim® is a web-based crop simulation system that can be used for irrigation scheduling and yield estimation. Two shortcomings of the system identified were that 1) advised irrigation amounts could exceed seasonal water limitations imposed on farmers and 2) simulations are only accurate if farmers follow the recommended irrigation actions and if simulated and actual available soil water content are similar. These can be addressed by incorporating algorithms for optimal scheduling of limited water, and by making use of soil water content measurements in model simulations. The objectives of this study were to 1) evaluate the performance of different optimization algorithms that schedule limited water and 2) determine the accuracy of irrigation scheduling advice and cane yield estimates with and without adjustment of simulations with soil water content records. Four irrigation scheduling algorithms were tested against a baseline algorithm, using 960 hypothetical scenarios consisting of different water supply, climate and cropping situations. These were: (a) Crop stage, which accounts for the yield sensitivity to water deficit as it varies with growth stage; (b) Stress level, which evaluates different soil water depletion levels for determining irrigation dates; (c) Prorata, which reduces irrigation throughout the growing season in proportion to the seasonal allocation shortfall; and (d) Water satisfaction, which iteratively schedules irrigation events on the day with the largest water demand. Algorithms increased simulated yields over the baseline by between 4.7 and 8.6 t/ha on average and operated at computational running times of between 1 and 40 s. The stress level algorithm was recommended for inclusion into MyCanesim®, since it had both a high yield improvement (8.5 t/ha) and quick operational time (2.5 s). Soil water measurements from capacitance probes for thirteen fields in Mpumalanga were integrated through an automated process into the MyCanesim® system. The improvements in the accuracy of irrigation scheduling advice and yield estimates by the integrated system were assessed retrospectively. The integrated system resulted in more accurate irrigation scheduling advice (by 2 days) than weather-based scheduling alone. These two improvements to MyCanesim® should allow sugarcane farmers to achieve higher irrigated water use efficiency and yields because of more accurate irrigation scheduling advice and yield estimates for full and restricted irrigation water supply.