Modelo baseado em dados meteorológicos para estimar o volume de água aplicado em uma plantação de dendê

Abstract

One of the biggest challenges of the agriculture is to estimate the water volume need to be applied on a plantation. Irrigate efficiently, besides saving water, increases plants productivity and makes them healthier. However, determining the quantity of water depends on various weather conditions, characteristics of the soil, the type of irrigation system used and the plant to be irrigated. In this research, is presented a knowledge model for estimating the volume of water to irrigate a plantation Elaeis guineensis, also known as oil palm. This plantation is located at the Farm School in São Luís, State University of Maranhão. This model was developed following the flow of process steps of KDD (Knowledge Discovery in Databases), using the technique of multiple linear regression on a database provided by the Geoenvironmental Center from the State University of Maranhão. The result of this procedure was a linear equation that calculates the volume of water to irrigate that plantation, helping the irrigator in the decision making. This equation showed good analytical results, indicating a strong positive correlation between the variables and low error ratesRO