Métodos da máxima verossimilhança e máxima verossimilhança restrita na estimação de parâmetros do modelo de variabilidade espacial

Abstract

The objective of this paper is the construction of thematic maps of the soil chemical properties and soybean productivity of an area located in the West of Parana State, in the city of Cascavel, where 83 samples of 12 parameters of chemical properties of soil and productivity were collected. It was also performed the exploratory analysis of the collected data and spatial analysis to estimate parameters through the maximum likelihood and restricted maximum likelihood methods, and, subsequently, the construction of thematic maps of these properties took place, using the two methods. The methods are validated by the method of cross validation, so that one can choose the most suitable for constructing the map of the variables studied. Through the maps, we have the history of the property where the data were collected and allows analysis and should help the making of decisions by the producer.