Sensoriamento remoto multiespectral passivo e ativo para o monitoramento do café arábica

Abstract

Monitoring the development of an agricultural area, with the help of images obtained from Radio Detection and Ranging (RADAR) active sensors, is an alternative to overcome such barriers in orbital monitoring scenario. Due to the characteristics of this type of sensor, it is possible to record properties of the terrestrial target, regardless of the occurrence of clouds between the target in question and the orbital sensor. This clouds occurrence is considerable common in coffee areas, due to the recommended edaphoclimatic conditions to perform its cultivation. Verifying that active orbital sensors are underused for agricultural monitoring, and especially for coffee growing, the present study aimed to evaluate the use of passive and active orbital sensors data for coffee areas monitoring, correlating their spectral records with field data. For this study, the database was prepared with information extracted from the images of passive and active sensors, as well as field data, in order to identify spectral and temporal patterns of coffee crops, evaluating them through geoprocessing techniques combined with univariate and multivariate statistics analyses. From the approaches adopted in each study focus, it was possible to generate models and validate them in order to introduce a new growth and development monitoring methodology for coffee parks, regardless of the climatic conditions presented in the study regions. It was also possible to verify the RADAR imaging application to record the spectrum-temporal behavior of coffee. Such result promoted a new possibility of coffee crops mapping and monitoring without significant interference of local cloudiness, eliminating the main difficulties that currently exist as a bottleneck in the advance of remote sensing of coffee growing.