Monitoramento de características morfofisiológicas de cafeeiros a partir de imagens de aeronave remotamente pilotada

Abstract

Remotely Piloted Aircraft (RPA) are used as platforms for remote sensing (SR) to carry out constant monitoring of the crop, as well as the identification of anomalies in them, in time and space, which makes it possible to assist in the management of the culture. Given this scenario, this work analyzed the potential of high-resolution images generated by multispectral SR data obtained from RPA in the characterization of morphophysiological parameters of coffee trees. The first study aimed to evaluate the relationship between yield and defoliation measured in the field and obtained through RPA images. With the processing of images from the year 2020, the crop showed a reduction of 17.3% and 18.4% in leaf area and volume, respectively, after harvest. And, in 2021, the crop showed a reduction of 12.8% and 9.8% of leaf area and volume, respectively, after harvest. In this way, it was possible to quantify the area and volume of a coffee crop after harvesting through the image obtained by RPA and also to analyze the interactions between field data with the production of the same harvest year, which are directly proportional, and the interaction of one-year image data with the previous production, being inversely proportional. The second study aimed to identify which Vegetation Index (VI) is suitable to explain the Chlorophyll inversion (Chl) methodology and to evaluate the relationships between the IVs obtained from the RPA image and leaf chlorophyll (Chl leaf) indices ) and canopy chlorophyll (Chl canopy ) in coffee trees during the rainy and dry seasons. The IVs that best explained Chl in the rainy season were the IVs MCARI2 RPA , MSR RPA and SR RPA . Both drought periods evaluated did not find a pattern in the relationships between Chl leaf , Chl canopy, and IVs. Finally, the third study aimed to characterize the temperature obtained by through RPA and evaluate its relationship with the water potential (PH) and stomatal conductance (gs) of an experimental coffee plantation through geostatistical techniques. With the data of gs, PH and the temperature maps, it can be observed that with the reduction of the PH, there was a stomatal closure and a reduction of the gs, favoring the increase in temperature due to water deficit in the studied periods. In addition, it was possible to observe the spatial distribution of temperature obtained through a thermal camera embedded in the RPA. The temperature distribution maps allowed to visualize the heterogeneous spatial distribution, which allowed to identify the areas where the plants were exposed to climatic conditions, which could be indicative of water deficit.