Predicting tomorrow's weather is easy: you look at the sky and at the wind map. A week ahead is harder, but doable. But a month? Forecasters long called this timeframe a 'desert': the atmosphere forgets everything that happened, and the ocean hasn't yet stored enough heat to give hints. Now scientists have found a way to cross that desert.
They gathered 40 years of data on temperature, soil moisture, and ocean conditions, then asked computer programs to find links between these data and the weather a month later. It turned out that two slow 'conductors' work best: moist ground and the El Niño current. Moist soil, like a wet sponge, holds heat longer, while El Niño alters weather across entire continents, but does so gradually. The atmosphere, by contrast, is too fidgety, and its rapid fluctuations only muddy the forecast.
For farmers, this means they could know in advance whether to expect drought or rain, and decide when to sow. For cities, it means better preparation for heat waves or cold snaps. The predictability desert turns out not to be so lifeless after all: it has oases, and computer models are learning to find them.
The maximal information coefficient between the temperature of the past two weeks and the forecast for the following two weeks was below 0.1 at most locations in the United States. In other words, recent weather says almost nothing about what will happen a month out.