Current research in ecosystem science emphasizes the monitoring, assessment, prediction, and sustainable management of regional ecological conditions. Terrestrial ecosystem models serve as critical tools for understanding and simulating ecosystem structures, processes, and services. Their development reflects an increasing focus on quantitative descriptions and numerical simulations, progressing from investigations of ecosystem components and their interactions toward the integration of multi-source data, multi-process simulations, multi-scale applications, and the construction of comprehensive simulation systems. This paper reviews the classification and evolution of terrestrial ecosystem models, systematically analyzing trends in model fusion and functional integration while identifying key challenges in scale adaptability, parameter transferability, and human disturbance simulation. Future advancements in ecosystem modeling should focus on elucidating the mechanistic processes of ecosystem evolution, incorporating multi-source data and intelligent technologies, and developing integrated natural-social-economic simulation systems. Such models must account for multi-factor and multi-scale dynamics, quantitatively describing ecological processes, their temporal changes, response mechanisms, and interactions at macro scales. These improvements will enhance the capacity for dynamic ecosystem management amid complex environmental shifts and anthropogenic pressures.