The sugarcane crop is of great economic relevance to Brazil, and the precise productivity estimation is a major challenge in production. Therefore, the aim of this study was to estimate the productivity of sugarcane cultivars in different regions, using multispectral sensors embedded in RPAs and biometric variables sampled in the field. The study was conducted in two experimental areas, located in the municipalities of Itirapina-SP and Iracemápolis-SP, with 16 cultivars in a randomized block design. The images were acquired using the multispectral sensor MicaSense Altum, allowing the extraction of spectral bands and vegetation indices. In parallel, biometric variables were collected at 149 and 295 days after planting (DAP). The machine learning models Random Forest (RF) and Extreme Gradient Boosting (XGBoost) were calibrated using different sets of variables, and, despite the similar performance, it was decided to use the model derived from XGBoost in the analyses, since it deals more effectively with overfitting. The results indicated a good performance of the model (R² = 0.83 and 0.66; RMSE = 18.7 t ha⁻¹ and 25.3 t ha⁻¹; MAE = 15.7 and 20.2; RPIQ = 3.22 and 2.61) for the validations K-fold and Leave-one-out cross-validation (LOOCV). The correlations between biometric variables, spectral bands, and vegetation indices varied according to crop development stage. The leaf insertion angle presented a strong correlation with near-infrared (NIR) (r = 0.76) and the indices ExG and VARI (r = 0.70 and r = 0.69, respectively). The present work demonstrated that the integration between multispectral and biometric data represents a promising approach for estimating sugarcane productivity.

Remotely Piloted Aircrafts (RPAs) equipped with multispectral sensors have emerged as promising tools for estimating foliar nitrogen content (TFN). So, this study applied a methodological approach aimed at simulating UAV multispectral data using hyperspectral leaf data obtained in a controlled environment, with the objective of evaluating its predictive potential and its transferability to field data collected by UAVs for TFN estimation. To this end, spectral bands and spectral indices (SIs) equivalent to those of UAV-mounted sensors were simulated based on hyperspectral data acquired by a benchtop sensor, and subsequently used in modeling via Partial Least Squares Regression (PLSR) and Random Forest (RF). The results showed similar performance across the levels, with R² values of 0.75 and 0.76 for PLSR and RF on the UAV data, and 0.75 and 0.74 for PLSR and RF on the simulated data, respectively. The RF model also performed well in cross-domain validation, with R² = 0.70 when calibrated with simulated data and applied to UAV data. Furthermore, the simulated data maintained high predictive power even with a reduced sample size. It is concluded that spectral simulation constitutes a viable strategy for expanding the applicability of nutritional monitoring using multispectral sensors.

Remotely Piloted Aircraft (RPAs) equipped with multispectral sensors have emerged as promising tools for estimating foliar nitrogen content (FNC). In this context, this study applied a methodological approach aimed at simulating UAV multispectral data using hyperspectral leaf data obtained in a controlled environment, with the objective of evaluating its predictive potential and its transferability to field data collected by UAVs for FNN estimation. To this end, spectral bands and indices equivalent to those of UAV-mounted sensors were simulated based on hyperspectral data acquired by a benchtop sensor, and subsequently used in modeling via Partial Least Squares Regres-sion (PLSR) and Random Forest (RF). The results showed similar performance across the levels, with R² values of 0.75 and 0.76 for PLSR and RF on the UAV data, and 0.75 and 0.74 for PLSR and RF on the simulated data, respectively. The RF model also performed well in cross-domain validation, with R² = 0.70 when calibrated with simulated data and ap-plied to UAV data. Furthermore, the simulated data maintained high predictive power even with a reduced sample size. It is concluded that spectral simulation constitutes a viable strategy for expanding the applicability of nutritional monitoring using multi-spectral sensors.

Proper nitrogen (N) management is essential for increasing the productivity of sugarcane (Saccharum spp.) and reducing the economic and environmental impacts associated with excessive fertilizer use. This study compared the performance of two portable spectroradiometers, FieldSpec 3 and HandHeld 2, in estimating foliar nitrogen content based on hyperspectral data in the visible and near-infrared regions, obtained throughout the crop cycle. The experiment was conducted in Piracicaba, São Paulo, Brazil, under four N rates: 0, 60, 120, and 180 kg ha⁻¹. Spectral measurements were taken at the foliar and canopy levels at eight evaluation times, accompanied by laboratory determination of N content. Partial Least Squares Regression (PLSR) and Random Forest (RF) models were fitted using the spectral data and days after cutting (DAC), included as a categorical factor and evaluated using 10-fold internal cross-validation, based on the metrics R², RMSE, MAE, and Willmott’s refined agreement index (dr). The foliar data performed better with PLSR (R² = 0.727; RMSE = 1.381 g kg⁻¹; MAE = 1.109; dr = 0.917) than canopy data (R² = 0.591; RMSE = 1.489 g kg⁻¹; MAE = 1.157; dr = 0.866). PLSR also outperformed RF at both acquisition levels. The green (~550 nm) and red edge (~740 nm) regions were the most relevant for N estimation. Under the evaluated conditions, model performance was associated with the spectral acquisition level and conditions, the instrumental configuration, and the modeling strategy employed.

This study evaluated the potential of VIS–NIR–SWIR leaf spectroscopy to predict N, P, and K contents in sugarcane, considering spectral variability across two growing seasons (2023/2024 and 2024/2025), phenological stages, and five varieties. Spectral signatures (350–2500 nm) were acquired using a FieldSpec 3 spectroradiometer and processed with MSC, smoothing, and the first Savitzky–Golay derivative. Diagnostic bands were identified by Spearman correlation, and prediction was performed using partial least squares regression (PLSR). PCA showed that spectral variability was driven mainly by seasonal and phenological factors, whereas varietal effects were secondary. In 2023/2024, greater spectral homogeneity was associated with lower predictive performance. In 2024/2025, greater spectral heterogeneity was associated with improved prediction for N (R² = 0.83; RMSE = 0.78 g kg⁻¹) and P (R² = 0.80; RMSE = 0.10 g kg⁻¹). Potassium remained the most challenging nutrient to predict (maximum R² = 0.25), mainly due to its ionic nature and the resulting lack of significant correlation with specific VIS, NIR, and SWIR spectral features. These results indicate strong potential for predicting N and P in fresh sugarcane leaves, although model robustness depends on the extent of spectral variability in the dataset.

This study used machine learning techniques to predict leaf nitrogen content from canopy hyperspectral data of sugarcane (Saccharum officinarum L.), thereby providing a rapid and non-invasive alternative to traditional laboratory analysis. The experiment was conducted in Piracicaba, São Paulo, Brazil, using ground-based spectroradiometry to obtain the canopy spectral signatures at different vegetative stages and under varying nitrogen doses. Spectral data were acquired at 67, 99, 144, 164, 200, 228, 255, and 313 days after cutting followed by leaf sampling and laboratory quantification of the leaf nitrogen content (LNC). Partial least squares regression, random forest, and extreme gradient boosting models were applied to predict the LNC. The relationship between the spectral response and leaf nitrogen was primarily associated with the visible and RedEdge regions, which are sensitive to chlorophyll absorption, pigment concentration, and crop nutritional status. Among the evaluated models, PLSR achieved the best performance, indicating that canopy hyperspectral sensing may be used as a complementary approach for nitrogen monitoring under field conditions; however, its predictive accuracy remained moderate. Overall, the results suggest that VIS-near-infrared region (NIR) hyperspectral data, combined with machine learning techniques, can support nitrogen monitoring in sugarcane. However, further validation under expanded field conditions is required before its implementation in operational nutrient management.