Quantifying the Zonal-Mean Structure of Tropical Precipitation

The tropical zonal-mean precipitation in climate models is well known to have substantial biases such as an erroneous double intertropical convergence zone in the Pacific, but a comprehensive quantification of these biases is currently missing. Therefore, we introduce a set of nine indicators that fully characterize the position and magnitude of the tropical extrema in zonal-mean precipitation. An analysis of the Coupled Model Intercomparison Project (CMIP) historical and Atmospheric Model Intercomparison Project (AMIP) simulations reveals large biases in the position and, especially, in the magnitude of the zonal-mean precipitation extrema in both sets of simulations relative to observations. We find some of the nine indicators to be correlated, and that the structure of tropical precipitation can be well represented using four indicators, though these indicators are different in AMIP and CMIP. Previously defined indicators can only partly explain the biases, and so the more comprehensive terminology introduced here is a useful tool for characterizing tropical precipitation.