Authors: Harusha Abeynayake, Heli Yu, Epari Ritesh Patro & Ali Torabi Haghighi
Publication type: A1
Abstract
River ice strongly influences hydrology, hydrodynamics, and ecosystems in cold regions, with an impact on flooding, hydropower generation, navigation, and aquatic habitats. Modeling river ice processes is therefore essential for both scientific understanding and operational management. Over the past several decades, a variety of hydraulic models have been developed to simulate river ice processes, aiming to improve our knowledge of river ice formation and evolution and to better predict ice phenomena, especially ice-related risks on both short-term and long-term scales. This review aims to synthesize recent progress in the numerical modeling of river ice processes covering formation, transport and accumulation, thermal growth and decay, and breakup. This study primarily assessed 13 models, considering their ability to simulate key thermodynamic and mechanical river ice processes. We summarized theoretical developments that underpin process representation, including thermodynamics, hydraulics, and ice mechanics, and reviewed a suite of widely used models, ranging from one-dimensional (1D) models such as RIVICE, JJT-model, River1D, RICE, RICEN, and CRISSP1D, to the two-dimensional (2D) models DynaRICE, CRISSP2D, and KHIONE. This review clarifies the trade-offs between model dimensionality, process completeness, and operational practicality by integrating model-specific data requirements, capability matrices, and global application evidence. This review shows that 1D models derived from the RICE–RICEN lineage remain foundational, and CRISSP1D emerges as the most comprehensive due to its extensive process coverage and strong validation across large rivers. CRISSP2D is emerging as the most integrated and physically detailed model for representing spatially complex thermodynamics and ice processes among 2D modeling frameworks.