Matrix Approach to Accelerate Spin-Up of CLM5

The spin-up problem hinders our ability to study some key issues in land carbon cycle modeling, such as sensitivity analysis. This work applies a new semi-analytical spin-up (SASU) framework to accelerate spin-up of Community Land Model matrix version 5. We evaluate the computational efficiency and steady-state consistency of ND, AD and SASU approaches using CLM5 at both one-site (Brazil) and the global scale. In addition, we test SASU in Parameter Perturbation Experiment (PPE) in CLM5.