River ice tracking and velocity estimation using remote sensing method and numerical modelling

Due to the severe cold weather in Canada, rivers experience freezing in cold seasons, and ice floes appear on the river's surface, which will change the hydraulic principles of the river. Frozen rivers significantly affect the environment and the economy, for instance, by damaging the hydraulic infrastructures, and they cause financial loss of millions of dollars yearly. A key parameter to determine the role of ice floes in river hydrodynamics is their velocity. Quantifying the movement of these ice floes can provide a unique opportunity to estimate river flow velocity to complete or replace in-situ measurements, which are costly and often challenging, especially in the presence of ice. Therefore, to quantify the movement of ice floes, near-simultaneous satellite images can be used. Satellite imagery data have received much attention in different specialties in recent years. The Remote-Sensing technique has been found to be useful in detecting ice floes on the river's surface. The displacement of ice floes can be provided by tracking them in near-simultaneous satellite images. Here, we analyze co-registered near-simultaneous image pairs from stereo acquisitions of the GEOEYE-1 satellite sensor with a spatial imaging resolution of 0.5 m obtained from the Maxar Company. An image correlation algorithm is needed to track ice floes in two sequential images. We use the Correlation Image Analysis software (CIAS) to quantify the ice floes' displacements in the Lower Athabasca River (LAR) section in Alberta, Canada on 11 November 2009. Here, by assuming that the ice and water velocity fields are similar, we estimate the river velocity for different river cross-sections.