Non-invasive crop analysis through image-based methods holds great promise for applications in plant research, yet accurate and robust trait inference from images remains a critical challenge. Our study investigates the potential of AI model ensembling and hybridization approaches to infer sorghum crop traits from RGB images generated via unmanned aerial vehicle (UAV). In our study, we cultivated 21 sorghum cultivars in two independent seasons (2021 and 2022) with a gradient of fertilizer and water inputs. We collected 470 ground-truth N measurements and captured corresponding RGB images with a drone-mounted camera. We computed five RGB vegetation indices, employed several ML models such as MLR, MLP, and various CNN architectures (season 2021), and compared their prediction accuracy for N-inference on the independent test set (season 2022). We assessed strategies that leveraged both deep and handcrafted features, namely hybridized and ensembled AI architectures. Our approach considered two different datasets collected during the two seasons (2021 and 2022), with the training set from the first season only. This allowed for testing of the models’ robustness, particularly their sensitivity to concept drifts, in the independent season (2022), which is fundamental for practical agriculture applications. Our findings underscore the superiority of hybrid and ensembled AI algorithms in these experiments. The MLP + CNN-VGG16 combination achieved the best accuracy (R² = 0.733, MAE = 0.264 N% on an independent dataset). This study emphasized that carefully crafted AI-based models applied to RGB images can achieve robust trait prediction with accuracies comparable to the similar phenotyping tasks using more complex (multi- and hyper-spectral) sensors presented in the current literature.

This study integrates open-source satellite-based shoreline extraction and statistical shoreline-change analysis to evaluate tidal correction and prediction ability in two contrasting coastal settings. Shoreline change is assessed using CoastSat-derived shorelines and AMBUR (Analyzing Moving Boundary Using R)-based rate modeling at Texas, United States of America and Ishikawa, Japan. Publicly available Sentinel‑2, Landsat, and high-resolution elevation datasets (NOAA (National Oceanic and Atmospheric Administration) lidar, GSI (Geospatial Information Authority of Japan) DEM, and SDB (Satellite Derived Bathymetry)-derived bathymetry) are combined with hourly or model tide records to extract, tidally correct, and analyze multi-decadal shoreline positions. Validation at San Jose Island compares tidally corrected Sentinel‑2 shorelines with 1 m NOAA lidar shorelines in terms of baseline distance, beach slope, and shoreline elevation. The CoastSat shoreline reproduces the lidar-derived planform with a distance RMSE

Accurate estimates of reservoir storage and long-term capacity loss are essential for reliable operation, sediment-management planning, and risk-informed decision making. However, bathymetric point surveys are irregularly distributed, and the reconstruction of the submerged surface introduces interpolation uncertainty that propagates into stage–storage curves and capacity-loss estimates. This study investigates the influence of spatial interpolation uncertainty on reservoir capacity-loss estimates and evaluates long-term sedimentation by comparing reservoir-bed conditions in 1987 and 2023 using Kamashi Reservoir (Uzbekistan) as a case study. Historical reservoir-bed conditions were reconstructed from digitized and georeferenced engineering topographic maps prepared in 1987, while the 2023 reservoir bed was modelled using bathymetric survey data. Five surface modelling approaches Inverse Distance Weighting (IDW), Radial Basis Functions (RBF), Ordinary Kriging (OK), Universal Kriging (UK), and Empirical Bayesia