What was done
Based on real-world data pertaining to the onset, end and total rainfall of the South American Monsoon System (SAMS) -- as characterized by precipitation data for the period 1979-2006, which they derived from the Global Precipitation Climatology Project -- the authors evaluated the ability of ten IPCC global coupled climate models (with distinct physics and resolutions) to simulate real-world SAMS characteristics.

What was learned
Bombardi and Carvalho report that over northern South America the annual precipitation cycle "is poorly represented by most models," and, more specifically, that "most models tend to underestimate precipitation during the peak of the rainy season." In addition, they say that "the misrepresentation of the Inter-Tropical Convergence Zone and its seasonal cycle seems to be one of the main reasons for the unrealistic out-of-phase annual cycles simulated near the equator by many GCMs," and that "poor representation of the total monsoonal precipitation over the Amazon and northeast Brazil is observed in a large majority of the models. As a consequence, therefore, they note that "simulations of the total seasonal precipitation, onset and end of the rainy season diverge among models and are notoriously unrealistic over [the] north and northwest Amazon for most models."

What it means
Once again, we have a demonstration of the fact that when computer-model output is compared with real-world data of the past, the comparison often does not look good, which gives us little confidence in the models' ability to correctly simulate the future.