Effects of parametric uncertainty on Intertropical Convergence Zone (ITCZ) precipitation: Understanding the role of interactions between parameters in reducing the double <scp>ITCZ</scp> bias

Abstract Reducing the double Intertropical Convergence Zone (ITCZ) bias has long been the subject of intensive research among the climate modeling community; however, the role of uncertain moist physical parameters from the atmospheric component model in causing the double ITCZ bias has not been analyzed systematically. This study investigated the potential for reducing the double ITCZ bias through perturbing nine moist physical parameters, using uniform sampling and Latin hypercube sampling methods, and quantified the parametric uncertainty and effects of two‐way interactions and overall nonlinear interaction between parameters on ITCZ precipitation in the Grid‐Point Atmospheric Model of the IAP/LASG Version 2. Results showed that the double ITCZ bias with excessive precipitation over the central and eastern Pacific south of the equator can be significantly reduced by suppressing deep convection intensity via multiple‐parameter perturbation. The uncertainty ranges of the double ITCZ bias associated with multiple‐parameter perturbation are apparently larger than those reconstructed from linear superposition of the single‐parameter perturbations, thereby highlighting the profound influences of interactions between parameters on ITCZ precipitation. The two‐way interaction terms are the primary contributors to the total variance of ITCZ precipitation. The effects of efficiency of precipitation and evaporation for deep convection are highly dependent on the threshold value for relative humidity for deep convection. The overall nonlinear effect among all the parameters tends to sharply exacerbate the bias of ITCZ heavy precipitation and reduce the bias of light precipitation. This strong overall nonlinear interaction among parameters in extreme precipitation simulation plays an important role in causing the double ITCZ bias, and it is closely associated with its impact on the intensity of the total moist processes. These findings validate the feasibility of reducing the double ITCZ bias by tuning moist parameters, and highlight the impact of interactions between parameters on simulation of ITCZ precipitation.