In recent years, drastic fluctuations in sea ice extent have underscored the critical need to deepen our understanding of its role in the global climate system. Beyond its climatic relevance, variations in sea ice extent and microstructure significantly influence ecological systems, particularly affecting microbial communities like ice algae that inhabit the porous structures within the ice. While large-scale models of sea ice dynamics have seen ongoing development and refinement, the intricate small-scale interactions involving temperature, salinity, and phase change processes remain inadequately captured. This study introduces a comprehensive multiscale modeling framework that captures thermodynamically consistent phase transitions between ocean water and sea ice. At the microscale, a phase-field model is utilized to simulate the formation and evolution of pores in the ice. These microscale pore characteristics are subsequently upscaled and incorporated into a macroscale formulation via the extended Theory of Porous Media (eTPM), enabling the prediction of sea ice growth rates. The framework effectively couples multiple physical fields, phases, and spatial scales, thus bridging microscale ice morphology with macroscale ice behavior. Simulation outcomes highlight the robustness and relevance of the proposed approach for modeling sea ice formation processes.