The barometer in the port of Papeete on Tahiti and the one in Darwin, Australia, are separated by 7,400 kilometers. When one shows high pressure, the other often shows low, and this difference, called the Southern Oscillation, drives rainfall from India to Brazil. Predicting it even a month ahead is a problem with a century-long history. Classic linear models are like a ruler laid along a winding river: they capture the general slope but miss every bend. The oscillation is nonlinear in nature, so forecasts often miss the mark.

The new approach resembles a council of a hundred short experts. Each expert is a small decision tree that knows only a small part of the picture: for example, "if pressure in May dropped by 2 hectopascals and rose by 1 in June, expect a value of 5 in October." One such expert is often wrong, but when a hundred weigh in and their opinions are averaged, the forecast becomes accurate. The method is called Bayesian additive regression trees, or BART. It has been used for other tasks, but no one had tried it on time series where each value depends on previous ones.

The author took monthly values of the oscillation from January 1876 to August 2009, a total of 1,603 observations. The trees were built randomly using the Markov chain Monte Carlo (MCMC) method, which can be imagined as a drunkard wandering across a map of probabilities and gradually finding the best spots. To keep the model from overloading, the number of trees was reduced to 20, and it was forced to select only useful past values, called lags. For predicting the October value from previous months, the best model used just five lags, from May to September. On the training set, the correlation reached 0.90; on the test set, it was 0.81, with a mean absolute error of 4.89 and a root mean square error (RMSE) of 5.90. If September is excluded to forecast two months ahead, the correlation drops to 0.66.

For forecasts 1 to 12 months ahead, BART with five lags gave a correlation of 0.647 for one month and 0.218 for twelve. The linear model AR(5) on the same data showed 0.620 for one month and 0.112 for twelve. The difference seems small at short horizons, but at long ones BART is twice as accurate. However, after five months, BART forecasts degenerate into an almost constant value with small variance, as if the council of experts, seeing no new data, starts repeating the same thing. This is the method's weakness: it honestly estimates uncertainty one step ahead, but for distant horizons it requires distributed computing, which is still expensive.

The work shows for the first time that a council of a hundred short experts is applicable to climate series and outperforms the linear model in the accuracy of point forecasts. Unlike neural networks, which learn from huge datasets but do not provide simple intervals, BART gives the full uncertainty distribution. This opens the door to nonlinear modeling of climate indices without complex tuning. Hydrologists, agrometeorologists, and climate risk specialists, who need nonlinear forecasts with uncertainty estimates, can use the method right away. The barometers in Papeete and Darwin will keep measuring pressure, and the council of a hundred experts will now listen to them more attentively.