Study of the modeling of rice cultivation under different irrigation systems and planning in the district of UsoCoello and analysis of the model applied to the region of Tolima in Colombia

Abstract

The management of irrigation districts is becoming laborious in some regions of Colombia. In the Tolima region, the bimodal climate is present in hot and semi-arid areas that are vulnerable due to the increasing impact of climatic phenomena that cause an intensification of droughts, but also of rainfall. The systems in place sometimes require more water than is needed, in a context of increasing producer density. Agricultural yields, particularly for rice, which is very present in the region, can sometimes decrease in an unprecedented way in the event of insufficient water supply. Moreover, the experiments to be carried out to put an end to these problems are sometimes costly in terms of time and money. In this study, trials of irrigation systems and scheduling were tried, and the use of a calibrated crop simulation model to assess the possibility of predicting the behaviour of these experiments on a variety of Oriza sativa L. from the region. The calibration shows simulations of the evolution of plant parameters, such as canopy cover with an average RMSE of about 2.60 % and correlated for the first cycle and a much poorer simulation for the second cycle, but also a simulated dry yield with an overall MBE of 0.621t/ha that is rather accurate and can differ from the observed evolution of the different treatments to some extent. Further simulations were carried out to assess the importance of the parameters on these outputs. The sensitivity analysis shows 7 parameters explaining 95 % of the total sensitivity of the evolution during a fictitious cycle of different soil water content, biomass, canopy cover and dry yield. These results will allow future evaluation of irrigation schedules in order to optimise yields according to local water consumption. Based on the results of the sensitivity analysis, further analyses can be done in the region to facilitate and plan future calibrations in a local environment. Thanks to the processed and modelled local climate data, it will also be possible to carry out irrigation schedules according to extreme weather events.