River ice processes were investigated to improve the understanding and modelling of river ice. Manning’s n for open water and ice cover conditions were found to vary with the average discharge of the season. Ice cover thickness was estimated using an improved Stefan’s equation. Through GMDH machine learning Stefan’s equation’s alpha was based on width, slope, and average discharge. A repurposed ice jam stability curve was used for the prediction of initial breakup of an ice cover. Field observations were taken for the Humber River at Bolton for the 2021-2022 and 2022-2023 winter seasons. The implementation of drone photography and 3D modelling using Structure of Motion was done to model ice covers and ice jams. A 3D model of a small ice jam was created and demonstrated new technology for ice jam monitoring, planning, and research by gathering important information on an ice jam from a safe distance.