No one remembers when you're right, but no one forgets when you're wrong. That's a saying we can all probably relate to. But arguably, no one deals with the backlash of getting things wrong as regularly as a weather person.
你做對的時候沒人記得, 但你錯的時候沒人會忘記。 我們所有人可能都有同感。 但可以說,沒有人像天氣專家那樣 經常得面對預報錯誤的強烈反應。
[Am I Normal? with Mona Chalabi]
「我正常嗎? 莫娜·查拉比」
From angry Twitter posts to hate mail, people can get really annoyed when the forecast is nothing like the reality. In 1964, the director of the Taiwan provincial weather bureau was even indicted for failing to correctly forecast the path of a typhoon.
從憤懣推特推文到騷擾信, 天氣預報錯了,人們會非常惱怒。 1964年,台灣省氣象所長 甚至因預測錯颱風路徑而被起訴。
So is our anger justified? I decided to find out just how accurate the weatherman really is. By comparing forecasts from 2017 to the actual temperatures that were recorded, I found that, as you'd expect, the forecast gets more accurate the closer you are to the actual date. So, for instance, when the US National Weather Service issued a forecast seven days in advance, it was off by over six degrees Fahrenheit. At one day in advance, their forecast was only off by about three degrees. But even though better modeling and better technology is expected to bring us better forecasts, we may never be able to predict the weather with 100 percent accuracy. That's because there are more than 100 tredecilion molecules in the atmosphere. That is the number one, followed by 44 zeros. So if we wanted to predict the weather with absolute certainty, we would need to know the position and movement of all of those particles, which is basically impossible for even our best computers.
但我們的不滿有理嗎? 我試著分析氣象員的準確性。 透過將2017年的預報 與紀錄在案的溫度做比較, 我發現,如你所想, 離預警期越近,預報越準確。 例如,美國國家氣象局 七天前發布的預報 和實際溫度相差超過華氏6度。 前一天的預報, 只和實際溫度差3度。 即便更好的建模及科技 有望讓我們更精準預測天氣, 我們仍無法百分之百準確預報天氣, 原因是大氣分子數 超過10的44次方, 也就是1後加44個0。 如果想要精準預測天氣, 我們得知道所有粒子的位置及動向, 即便是超級電腦也根本做不到。
And our drive to know the future isn't limited to the weather. Take elections, for example. Despite what some websites or publications may lead you to believe, the most accurate election polls are the ones that are taken on election day, not the ones that are carried out in advance.
天氣不是我們唯一想預知的未來。 舉選舉為例, 儘管有些網站或刊物會洗腦你, 但選舉日的民意調查才是最準的, 而不是事前的。
So the lesson here isn't a terribly surprising one. Accurate predictions depend on accurate information. And the further out you are, the higher the chances that information can change.
這並非最驚人的事實。 確切資訊才能精準預測。 距事發日越早, 資訊越易改變。
So for better accuracy, you just need to be patient. Try to hold off as close as you can to the actual event. And for now, go easy on your local weather person.
因此,只要有耐心, 就能提高準確度。 盡量貼近事發日在做預測。 現在,寬待你的氣象員。