The Science of Predicting the Future

While most of the world watched soccer or did laundry on Sunday, a group of scientists at the office of the United Nation’s Environment Program in Kenya hit publish on a report about the future of Planet Earth so comprehensive that it would take three reams of paper—500 pages each—to print.

Its takeaways have been widely publicized and highly forecasted and may boil down to this: The world is on shaky footing. A changing climate over the next several decades will have sweeping impacts over all areas of human civilization. Warmer global temperatures will restrict access to water, reduce food yields, and heighten tensions between countries in search of natural resources. Perhaps the most sobering was the nugget that harsh weather events, such as the ones we’ve already seen with Hurricane Sandy and the California drought, will be more frequent and occur with greater force. No one, the report said, will be spared.

Such an alarmist forecast comes with a rather obvious question: How can anyone know how bad storms will be 30 years from now? If our current crop of meteorologists can’t accurately predict if it’ll rain on Friday, who’s to say what will happen at the end of the century?

Buried deep inside Sunday’s climate tome—between multicolored charts and dizzying spreadsheet—is a very deep explanation of the difference between climate and weather. Climate accounts for long-term trends, plugging numbers into models to see in a very general sense how the system will react. Those types of models can answer whether China will be warmer over the next 10 years than in was over the past 10 years (spoiler: it will). It’s safe then to say that the decade after that, it’ll only be warmer, and so on. Weather, on the other hand, is a matter of specifics. Knowing whether it’ll be cloudy tomorrow at 4:30 p.m. requires so many precise measurements—even one that’s slightly wrong can derail the entire calculation—that it’s impossible to know with 100 percent certainty. Is it likely to rain in the Northern Hemisphere in mid-April? Yes. Will it rain in a specific place at 3 p.m. tomorrow afternoon? With 70 percent certainty, scientists (often trying to avoid attaching their reputations to specific forecasts) say maybe, yes, it sort of might be possible.

An easier way to think of it might be to consider the planet’s seasonal flux. It’s very safe to think that after winter, temperatures will slowly rise. May is likely to be warmer than April and somewhere in July or August, the northern hemisphere will see its highest temperatures of the year. Who knows what’ll happen on, say, August 23? But it’s a reasonable assumption that it won’t snow.

Forecasting often works like that. Long-term visions are often clearer than short term ones. In 20 years, we can all assume the stock market will be higher than it is now, even if you don’t know what the exact balance of your bank account will be. If you follow a linear career path, odds are you’ll have a more senior job title a decade from now, even though you can’t know for sure where you’ll actually work. Sweeping assumptions aren’t meant to be detailed and air-tight, But they do provide a general mindset to get into so that when it does snow on a Monday in June in 2028, you can’t claim to be surprised.