Análise espectro-temporal de produtos do sensor MODIS como diagnóstico para a cafeicultura de precisão

Abstract

The coffee is an agricultural and economic activity of great importance for the Brazilian market. Despite the social and economic importance to the country, the coffee needs additional information from your production system. The remote sensing through satellite images and products available free, like MODIS sensor, can be a key tool for mapping and monitoring for the coffee. Therefore, the objective of this work was to analyze the application of MODIS sensor products as diagnostic data for precision coffee. The study area is located in the city of Campos Gerais/MG, Brazil with area of 461.28 hectares of coffee plantations, divided into experimental plots. Two times series were used and acquired from INPE, (a) time series EVI2 MODIS sensor; (b) number of rainfall (mm/month), the TRMM sensor. Another series used, acquired from NTSG, was to evapotranspiration (mm/month) of MOD16A2/A3 MODIS product and was also collected "in situ" property data in the study, such as, area, growing, productivity, among others. The collected data were selected and organized into a database. After creating the database, the spectral curves and performed correlation and regression for the variables (data) were generated. The results showed that EVI is influenced by moisture content and evapotranspiration, and then the EVI influence on crop yields. Therefore, for diagnostic accuracy in coffee growing, the use of products with better spatial resolutions for the products of the MODIS sensor is required.