RESUMO
Knowing the spatial variability of sugarcane biomass in the early stages of development may help growers in their management decision-making. Proximal canopy sensing is a promising technology that can identify this variability but is limited to quantifying plant-specific parameters. In this study, we evaluated whether biometric variables integrated with canopy reflectance data can assist in the generation of models for early-stage sugarcane biomass prediction. To substantiate this assertion, four sugarcane-producing fields were measured with an active crop canopy sensor and 30 sampling plots were selected for manually quantifying chlorophyll content, plant height, stalk number and aboveground biomass. We determined that Random Forest and Multiple Linear Regression models are similarly able to predict biomass, and that associating biometric variables such as number of stalks and plant height with reflectance data can assist model performance, depending on the attributes selected. This indicates that, when estimating biomass in the early stages, sugarcane growers can carry out site-specific management in order to increase yield and reduce the use of inputs.(AU)
RESUMO
Knowing the spatial variability of sugarcane biomass in the early stages of development may help growers in their management decision-making. Proximal canopy sensing is a promising technology that can identify this variability but is limited to quantifying plant-specific parameters. In this study, we evaluated whether biometric variables integrated with canopy reflectance data can assist in the generation of models for early-stage sugarcane biomass prediction. To substantiate this assertion, four sugarcane-producing fields were measured with an active crop canopy sensor and 30 sampling plots were selected for manually quantifying chlorophyll content, plant height, stalk number and aboveground biomass. We determined that Random Forest and Multiple Linear Regression models are similarly able to predict biomass, and that associating biometric variables such as number of stalks and plant height with reflectance data can assist model performance, depending on the attributes selected. This indicates that, when estimating biomass in the early stages, sugarcane growers can carry out site-specific management in order to increase yield and reduce the use of inputs.
RESUMO
The Random Forest algorithm is a data mining technique used for classifying attributes in order of importance to explain the variation in an attribute-target, as soil CO2 flux. This study aimed to identify prediction of soil CO2 flux variables in management systems of sugarcane through the machine-learning algorithm called Random Forest. Two different management areas of sugarcane in the state of São Paulo, Brazil, were selected: burned and green. In each area, we assembled a sampling grid with 81 georeferenced points to assess soil CO2 flux through automated portable soil gas chamber with measuring spectroscopy in the infrared during the dry season of 2011 and the rainy season of 2012. In addition, we sampled the soil to evaluate physical, chemical, and microbiological attributes. For data interpretation, we used the Random Forest algorithm, based on the combination of predicted decision trees (machine learning algorithms) in which every tree depends on the values of a random vector sampled independently with the same distribution to all the trees of the forest. The results indicated that clay content in the soil was the most important attribute to explain the CO2 flux in the areas studied during the evaluated period. The use of the Random Forest algorithm originated a model with a good fit (R2 = 0.80) for predicted and observed values.(AU)