Parameter modulation of Madden-Julian Oscillation behaviors in BCC_CSM1.2: The key role of moisture-shallow convection feedback

To reveal key parameter-related physical mechanisms in simulating Madden-Julian Oscillation (MJO), seven physical parameters in the convection and cloud parameterization schemes of Beijing Climate Center Climate System Model (BCC_CSM1.2) are perturbed with Latin hypercube sampling method. A new strategy is proposed to select runs with good and poor MJO simulations among 85 generated ones. Outputs and parameter values from good and poor simulations are composited separately for comparison. Among the seven chosen parameters, a decreased value of precipitation efficiency for shallow convection, higher values of relative humidity threshold for low stable clouds and evaporation efficiency for deep convective precipitation are crucial to simulate a better MJO. Changes of the three parameters act together to suppress heavy precipitation and increase the frequency of light rainfall over the Indo-Pacific region, supplying more moisture in low and middle troposphere. As a result of a wetter lower troposphere ahead of the MJO main convection, the low-level moisture preconditioning along with the leading shallow convection tends to be enhanced, favorable for MJO's further development and eastward propagation. The MJO's further propagation across the Maritime Continent (MC) in good simulations is accompanied with more land precipitation dominated by shallow convection. Therefore, the above-mentioned three parameters are found to be crucial parameters out of the seven ones for MJO simulation, providing an inspiration for better MJO simulation and prediction with this model. This work is valuable as it highlights the key role of moisture-shallow convection feedback in the MJO dynamics.