How well do CMIP5 models simulate the low-level jet in western Colombia?

The Choco jet is an important atmospheric feature of Colombian and northern South America hydro-climatology. This work assesses the ability of 26 coupled and 11 uncoupled (AMIP) global climate models (GCMs) included in the fifth phase of the Coupled Model Intercomparison Project (CMIP5) archive to simulate the climatological basic features (annual cycle, spatial distribution and vertical structure) of this jet. Using factor and cluster analysis, we objectively classify models in Best, Worst, and Intermediate groups. Despite the coarse resolution of the GCMs, this study demonstrates that nearly all models can represent the existence of the Choco low-level jet. AMIP and Best models present a more realistic simulation of jet. Worst models exhibit biases such as an anomalous southward location of the Choco jet during the whole year and a shallower jet. The model skill to represent this jet comes from their ability to reproduce some of its main causes, such as the temperature and pressure differences between particular regions in the eastern Pacific and western Colombian lands, which are non-local features. Conversely, Worst models considerably underestimate temperature and pressure differences between these key regions. We identify a close relationship between the location of the Choco jet and the Inter-tropical Convergence Zone (ITCZ), and CMIP5 models are able to represent such relationship. Errors in Worst models are related with bias in the location of the ITCZ over the eastern tropical Pacific Ocean, as well as the representation of the topography and the horizontal resolution.