A map of China looks like a desert dotted with rare green oases. In those oases, the GloSea5 climate model can forecast weather three months ahead, while in the desert its predictions are no better than random guessing. Scientists spent twenty years testing the model against past data to find where it can be trusted.
The second oasis is summer heat in Beijing. Even after removing the overall climate warming trend, the model still predicts temperature better than a simple "business as usual" approach. That means it picks up a real signal, not just repeats a trend.
But beyond these spots lies desert. Summer winds and winter temperatures are hardly predictable anywhere. In the very South China Sea where the model excels in winter, its accuracy in summer drops to nearly zero. It is like a compass that works in clear weather but is useless in fog.
For China's energy sector, this matters. The country is building many solar and wind farms, and it needs to know in advance whether winters will be windy and summers hot. For now, only the oases can be trusted, but that is already enough to launch pilot services for the South China Sea and Beijing. If the model improves, the green patches will grow.
The correlation of 0.78 for winter winds in the South China Sea is comparable to the skill of El Niño forecasts, which is considered the gold standard of seasonal prediction. Yet in the same region in summer, the correlation drops to 0.13, meaning the forecast loses all skill.