Search

The maturation process of coffee (Coffea arabica) trees exhibits inherent variability, producing fruits at various physiological maturity stages. This variability affects the resistance between the fruit and its peduncle, posing a challenge in mechanized harvesting: non-selective harvesting. A precise classification of coffee fruit detachment force is essential to address this challenge, ensuring coffee’s quality and producer’s profitability. This study assesses the efficacy of machine learning (ML) models in determining the detachment force across various coffee cultivars under drip-irrigated and rainfed conditions. The dataset included detachment force measurements from 24 cultivars—13 drip-irrigated and 11 rainfed—yielding 1152 data points. Variance analysis compared irrigation methods and three maturity stages: green, cherry, and dry. Detachment force was categorized into four classes based on the dataset’s quartile distribution. The ML models utilized were random forest (RF), support vector machine (SVM), K-nearest neighbors, and artificial neural networks. The SVM model was notably effective in classifying detachment force for rainfed cultivars, with a Matthews correlation coefficient (MCC) of 0.78. In contrast, the RF model was particularly adept for drip-irrigated cultivars, with an MCC of 0.75. The highest classification accuracies were recorded for the extreme force classes I and IV, with precision values of 0.93 and 0.8, respectively, while classes II and III had lower precision at 0.57 and 0.69. Implementing these ML models for detachment force classification has been beneficial, improving decision-making in mechanized harvesting systems. [ABSTRACT FROM AUTHOR]

Background: Climate influences the interaction between pathogens and their hosts significantly. This is particularly evident in the coffee industry, where fungal diseases like Cercospora coffeicola, causing brown‐eye spot, can reduce yields drastically. This study focuses on forecasting coffee brown‐eye spot using various models that incorporate agrometeorological data, allowing for predictions at least 1 week prior to the occurrence of disease. Data were gathered from eight locations across São Paulo and Minas Gerais, encompassing the South and Cerrado regions of Minas Gerais state. In the initial phase, various machine learning (ML) models and topologies were calibrated to forecast brown‐eye spot, identifying one with potential for advanced decision‐making. The top‐performing models were then employed in the next stage to forecast and spatially project the severity of brown‐eye spot across 2681 key Brazilian coffee‐producing municipalities. Meteorological data were sourced from NASA's Prediction of Worldwide Energy Resources platform, and the Penman–Monteith method was used to estimate reference evapotranspiration, leading to a Thornthwaite and Mather water‐balance calculation. Six ML models – K‐nearest neighbors (KNN), artificial neural network multilayer perceptron (MLP), support vector machine (SVM), random forests (RF), extreme gradient boosting (XGBoost), and gradient boosting regression (GradBOOSTING) – were employed, considering disease latency to time define input variables. Results: These models utilized climatic elements such as average air temperature, relative humidity, leaf wetness duration, rainfall, evapotranspiration, water deficit, and surplus. The XGBoost model proved most effective in high‐yielding conditions, demonstrating high precision and accuracy. Conversely, the SVM model excelled in low‐yielding scenarios. The incidence of brown‐eye spot varied noticeably between high‐ and low‐yield conditions, with significant regional differences observed. The accuracy of predicting brown‐eye spot severity in coffee plantations depended on the biennial production cycle. High‐yielding trees showed superior results with the XGBoost model (R2 = 0.77, root mean squared error, RMSE = 10.53), whereas the SVM model performed better under low‐yielding conditions (precision 0.76, RMSE = 12.82). Conclusion: The study's application of agrometeorological variables and ML models successfully predicted the incidence of brown‐eye spot in coffee plantations with a 7 day lead time, illustrating that they were valuable tools for managing this significant agricultural challenge. © 2024 Society of Chemical Industry. [ABSTRACT FROM AUTHOR]

BACKGROUND: We evaluated different machine learning (ML) models for predicting soybean productivity up to 1 month in advance for the Matopiba agricultural frontier (States of Maranhão, Tocantins, Piauí, and Bahia). We collected meteorological data on the NASA‐POWER platform and soybean yield on the SIDRA/IBGE base between 2008 and 2017. The ML models evaluated were random forest (RF), artificial neural networks, radial base support vector machines (SVM_RBF), linear model and polynomial regression. To assess the performance of the models, cross‐validation was used, obtaining the value of precision by R2, accuracy by root mean square error (RMSE), and trend by the mean error of the estimate (EME). RESULTS: The results showed that the RF algorithm achieves the highest precision and accuracy, with R2 of 0.81, RMSE of 176.93 kg ha−1 and trend (EME) of 1.99 kg ha−1. On the other hand, the SVM_RBF algorithm showed the lowest performance, with R2 of 0.74, RMSE of 213.58 kg ha−1 and EME of −15.06 kg ha−1. The average yield values predicted by the models were within the expected range for the region, which has a historical average value of 2.730 kg ha−1. CONCLUSION: All models had acceptable precision, accuracy and trend indices, which makes it possible to use all algorithms to be applied in the prediction of soybean crop yield, observing the particularities of the region to be studied, in addition to being a useful tool for agricultural planning and decision making in soy‐producing regions such as the Brazilian Cerrado. © 2021 Society of Chemical Industry. [ABSTRACT FROM AUTHOR]

Background: The recent studies of the variations in the atmospheric column-averaged CO2 concentration ( X CO 2 ) above croplands and forests show a negative correlation between X CO 2 and Sun Induced Chlorophyll Fluorescence (SIF) and confirmed that photosynthesis is the main regulator of the terrestrial uptake for atmospheric CO2. The remote sensing techniques in this context are very important to observe this relation, however, there is still a time gap in orbital data, since the observation is not daily. Here we analyzed the effects of several variables related to the photosynthetic capacity of vegetation on X CO 2 above São Paulo state during the period from 2015 to 2019 and propose a daily model to estimate the natural changes in atmospheric CO2. Results: The data retrieved from the Orbiting Carbon Observatory-2 (OCO-2), NASA-POWER and Application for Extracting and Exploring Analysis Ready Samples (AppEEARS) show that Global Radiation (Qg), Sun Induced Chlorophyll Fluorescence (SIF) and, Relative Humidity (RH) are the most significant factors for predicting the annual X CO 2 cycle. The daily model of X CO 2 estimated from Qg and RH predicts daily X CO 2 with root mean squared error of 0.47 ppm (the coefficient of determination is equal to 0.44, p < 0.01). Conclusion: The obtained results imply that a significant part of daily X CO 2 variations could be explained by meteorological factors and that further research should be done to quantify the effects of the atmospheric transport and anthropogenic emissions. [ABSTRACT FROM AUTHOR]

With the advancement of remote sensing, it is now possible to identify and characterize greenhouse gas emissions under deferment land uses. Given the above, this study aimed to characterize the spatial–temporal variability and the main factors controlling the average atmospheric CO2 column (Xco2) in the macroregion of Ribeirão Preto (MRP), São Paulo, a significant sugarcane producer in Brazil. We obtained remote sensing data from January 2015 to December 2018. The variables used were Xco2 and sun-induced fluorescence of chlorophyll (SIF) by NASA's Orbiting Carbon Observatory-2 satellite (OCO-2), relative humidity (RH), global radiation (Qg), and the average temperature at 2 m (T2m) by the NASA-POWER platform, and leaf area index (LAI) and evapotranspiration by Penman–Monteith (ET) by MODIS sensor. We evaluated the data in trimester's averages, where descriptive statistics, Pearson correlation and linear regression have been applied. The spatial distribution was made by the inverse distance weighted (IDW). The minimum (390.40 ± 0.41 ppm) and maximum (394.75 ± 0.34 ppm) mean of Xco2 was observed in the first quarter of 2015 and third quarter of 2017. The Xco2 obtained negative correlations with the SIF (−0.81), LAI (−0.81), RH (−0.74), ET (−0.84), and Qg (−0.51). Hotspots and coldspots of Xco2 tend to vary over the years. We conclude that the temporal variation of Xco2 above sugarcane areas in southern Brazil is well represented by a periodic function. Our results indicate photosynthesis and soil exposure after harvest are factors that could act as source and sink of CO2. [ABSTRACT FROM AUTHOR]

BACKGROUND: The loss of coffee leaves caused by the attack of pests and diseases significantly reduces its production and bean quality. Thus this study aimed to estimate foliation for regions with the highest production of arabica coffee in Brazil using nonlinear models as a function of climate. A 25‐year historical series (1995–2019) of Coffea arabica foliation (%) data was obtained by the Procafé Foundation in cultivations with no phytosanitary treatment. The climate data were obtained on a daily scale by NASA/POWER platform with a temporal resolution of 33 years (1987–2019) and a spatial resolution of approximately 106 km, thus allowing the calculation of the reference evapotranspiration (PET). Foliation estimation models were adjusted through regression analysis using four‐parameter sigmoidal logistic models. The analysis of the foliation trend of coffee plantations was carried out from degrees‐day for 70 locations. RESULTS: The general model calibrated to estimate the arabica coffee foliation was accurate (mean absolute percentage error = 2.19%) and precise (R2adj = 0.99) and can be used to assist decision‐making by coffee growers. The model had a sigmoidal trend of reduction, with parameters ymax = 97.63%, ymin = 9%, Xo = 3517.41 DD, and p = 6.27%, showing that foliation could reach 0.009% if the necessary phytosanitary controls are not carried out. CONCLUSION: Locations with high air temperatures over the year had low arabica coffee foliation, as shown by the correlation of −0.94. Therefore, coffee foliation can be estimated using degree days with accuracy and precision through the air temperature. This represents great convenience because crop foliation can be obtained using only a thermometer. © 2021 Society of Chemical Industry. [ABSTRACT FROM AUTHOR]