Aplicação da Análise de Componentes Principais na Obtenção dos Padrões de Clima e Cultura de Mandioca na Mesorregião do Nordeste Paraense

Abstract

Understand in ghow spatial and temporal Weather patterns can influence, or even modulate the yield of agricultural crops, becomes essential in agricultural activities and that in Brazil it has not only economic, but also socio-environmental importance. Cassava, being a crop that in general is without irrigation, depends exclusively on the availability of rainfall to supply its water needs. Thus, the present work seeks to establish spatial patterns of correlation between precipitation, cassava yield and the planted area of the crop, during the period from 2000 to 2019, through the statistical tool Principal Components Multivariate Analysis (PCA), also known as Empirical Orthogonal Functions (FOE) that have been widely applied to climatic, environmental and meteorological studies. The PCA proved to be efficient for this purpose, so that in the experiments carried out, the first three PCs were able to explain up to 70% of the total variance of the data, revealing high correlations between precipitation and yield in the Northern portion of the Northeast of Pará as of 2010. In addition, other patterns of correlations were observed between yield and planted area.