Land carbon cycle models are an essential component of Earth system models, and offer an opportunity to not only study climate change but also quantify changes in societally relevant quantities such as food, timber, and energy crop. Simulations of the terrestrial carbon cycle are much variable among models and largely dominated by model uncertainty. Thus, land carbon cycle models are not fully upto the task to inform sound policy that maintains a healthy biosphere and provides the food, energy, and fresh water needed for a growing global population without further exacerbating climate change. We have recently analyzed properties of the land carbon cycle using concepts from dynamical system theory, and found that all land carbon cycle models can be unified by one dynamical equation. The dynamical equation is expressed in a matrix form or called a matrix approach. We have converted a couple dozen of existing models, such as ORCHIDEE, CLM3.5, 4.0, 4.5, and 5.0 as well as some microbial models, to the unified matrix form. Unifying land carbon cycle models offers opportunities to bridge the gap between understanding and complexity of models and their simulation outputs, identify uncertainty sources, and improve existing models and their predictions from multiple streams of observations. The unification can also enable new science, such as ensemble modeling, super-ensemble prediction of future dynamics of land carbon cycle, and model development-evaluation-improvement continuum. In this presentation, we will use examples to explain how we can unify all land carbon models under one overarching framework, how the framework can be used to understand the contribution of system components to overall uncertainty, and how it offers new opportunities to advance a new set of research questions in carbon cycle science.