Automated River Ice Freeze-up Monitoring and Characterization

Rivers in cold regions such as Canada are covered by ice during a significant portion of the year. River ice processes such as ice cover formation, progression, recession, and break-up can affect river hydraulics, sediment transport, and its morphology. Ice can also drastically impact physical, chemical and biological properties of rivers. Freeze-up is one of the most significant stages of ice cover formation in cold seasons. River ice cover is a complex and poorly understood process driven by a multitude of factors, among with local climatic condition, river geomorphology and hydraulic characteristics. Given the lack of reliable predictive model, In situ field investigations, although difficult and potentially dangerous, are crucial for understanding ice cover formation dynamic. Remote sensing and the application of digital cameras in river ice processes monitoring have recently been used by several researchers. The acquired images and video files have been used in several studies for qualitative assessment; however, accurate quantified data acquisition is still very demanding. One of the most challenging tasks of shore-based monitoring of river ice processes is river ice detection. Different thresholding methods have been used previously to address this need; however, none of the methods have demonstrated promising results for a robust algorithm. The main goal of this study was to present a novel approach for river ice detection in an automated pipeline for shore-based monitoring of river ice. Images of a series of trail cameras mounted at different locations along Dauphin River, Manitoba, Canada were used in this study to train Convolutional Neural Network (ConvNN) models to detect river ice. Results of the trained ConvNN models were then used to detect river ice cover. The presented automated ice detection algorithm is faster and more accurate compared to the other examined common detection methods. Results of such a study can potentially be used for better estimation of river ice concentration along the monitored sections of the river that can be utilized in numerical models. Moreover, the ice detection algorithm is fast enough to be employed for marine transportation at high latitudes.