Modelagem do parâmetro de irrigação: um estudo de caso para o capim sudão BRS estribo

Abstract

Water availability is one of the key factors for the success of agricultural production. Under natural conditions, the crops are supplied with water through precipitation, and during drought periods irrigation management tasks are required. Irrigation is an artificial technique used to supply water to crops, and in order for this task to be well managed, the irrigation parameter is needed. The irrigation parameter is the main indicator when doing irrigation tasks, usually expressed in millimeters. In summer, weather conditions can restrict water availability, which causes problems in crops and productive chains. To tackle this problem, in partnership between Embrapa Pecuária Sul and Sulpasto, the capim-sudão BRS Estribo was developed and launched, which is an annual summer forage, which has been gaining preference among producers. This forage has been adopted because, if well managed, it has a high potential of production. In this study we present an empirical model, which based on a sowing date, meteorological data and forage parameters, calculates the irrigation parameter. It is also shown that the CART, SVM and ANN models can predict the irrigation parameter with an R2 equal to 0,986; 0,985 and 0,988 respectively, from the empirical model. The results presented in this study indicate that the empirical model and artificial intelligence models can be used in the practical application of planning and management of water resources.