The high-tech race to improve weather forecasting, Part 4

????這篇文章有點長，我把它分成了考研閱讀的長度（600-700words）。

????第一部分介紹了歐洲中期天氣預(yù)報中心的功能、極端天氣與天氣預(yù)報的重要性。在結(jié)尾用1960s至今預(yù)測準(zhǔn)確率的提升引出第二部分。

????第二部分介紹了天氣預(yù)測的原理及預(yù)測模型計算問題。

????第三部分介紹私人預(yù)測機構(gòu)的優(yōu)勢與功能。

????最后一部分介紹了AI時代氣象預(yù)測領(lǐng)域的變局。

Sunny with a chance of AI? 天晴，可能有AI

"Sunny with a chance of showers"，即“天晴可能下雨”的變體

1.機器學(xué)習(xí)簡化天氣預(yù)測

Private companies have also been at the forefront of attempts to find new, less computationally onerous?ways of predicting the weather. Many are focusing on?machine learning, a type of artificial intelligence (AI) that looks for patterns in big piles of data.

Salient, an American?startup, uses anAI?trained to recognise patterns in historical data to produce forecasts on a seasonal scale, rather than over days or weeks.?

Its customers include Zurich Insurance Group, which hopes to get early warnings of extreme weather its clients might face.

machine learning?機器學(xué)習(xí)

computationally?計算上

onerous?繁重的

a startup?創(chuàng)業(yè)公司

2.AI可以看出研究者忽略的模式

AI?can?spot?patterns that human researchers may have missed. Ray Schmitt, a researcher at the Woods Hole Oceanographic Institution in Massachusetts, is one of Salient’s founders.?

He had theorised about a link between ocean?salinity?around the east coast of America in spring and rainfall across the Midwest the following summer.?AI?analysis of weather data seems to confirm the connection, though the precise mechanism remains unclear.

spot?看出

salinity?鹽度

3.AI預(yù)測不必了解氣象原理

That illustrates another intriguing feature ofai-based forecasts. Numerical simulations rely on their programmers having a good understanding of the physical processes that drive the weather.?

But using an?AI?to spot recurring patterns can help useful forecasts be produced even before the underlying science is fully understood.

4.現(xiàn)有計算模型的不足及AI的優(yōu)勢

Machine learning has already proved its worth with precipitation “nowcasting”—predicting whether it will rain or snow in a given area over the next few hours.?

But predicting them can be tricky for existing numerical models, partly because, by the time they have finished running, the moment has often passed.AI?pattern recognition requires less?computational grunt,?allowing it to make forecasts more quickly.

computational grunt?計算能力
grunt?野豬的呼嚕聲

5.多方合作AI氣象項目

A 2021 collaboration between DeepMind, a part of Google, and the Met Office in Britain usedAI?to forecast?precipitation?based on observations from rain-detecting radar.?

The?AI?system outperformed existing, numerical forecasting methods nine times out of ten—though it started to?stumble?when asked to forecast beyond about 90 minutes.

precipitation?降雨量

stumble?出錯

6.華為盤古氣象大模型

Other big firms with?AI?expertise are getting involved, too. A paper published in?Nature?on July 5th described Pangu-Weather, an?AI?system built by Huawei, a Chinese firm, and trained on 39 years of weather data.

Huawei claims Pangu-Weather can produce week-ahead predictions comparable in accuracy to forecasts from outfits like?ECMWF, but thousands of times faster.?

Last year Nvidia, an American chipmaker, claimed that FourCastNet, its?AI?weather program, could generate, in two seconds, a forecast that can predict hurricanes and heavy rain?up to a week in advance.

7.政府官員改變看法

Governmental incumbents?are coming around. The?ECMWF?was surprised by the results of Pangu-Weather, says Florence Rabier, the organisation’s?director-general.

“We did see a lot of potential, and they are not exaggerating the claims that it is much cheaper [to run],” she says. The?ECMWF?is now working with Huawei, as well as with Google and Nvidia.

governmental incumbent?政府現(xiàn)任官員

be coming around?改變態(tài)度或想法

director-general?總監(jiān)

8.AI不能取代數(shù)值預(yù)測

That does not mean that?AI?will replace numerical forecasting, though it could help it become more efficient.?AI?relies crucially on high-quality data?onwhich to?train?models.?

Since many parts of the world lack reliable data from weather stations, old-fashioned numerical simulations must be used retrospectively to fill in the gaps.?

And just as computational approaches?face fundamental limits to?their utility, so too do?AI-based ones.?History is a less reliable guide to the future in a world whose weather is being fundamentally altered by climate change.

（1）train the models on high-quality data，這里的on是根據(jù)的意思。除此之外，“train A on B”還可以表示瞄準(zhǔn)。比如：The security guards trained their flashlights on the suspicious figure lurking in the shadows.?保安們用手電筒照向在陰影中潛伏的可疑人物。

（2）face fundamental limits to，介詞to易考完形填空。

（3）“And?just as?computational approaches?face fundamental limits to?their utility,?so too do?AI-based ones.”這是兩個并列的主語從句，表示前后兩者地位相似。可以理解為，計算方法與AI一樣要面臨功能的基礎(chǔ)限制。computational approaches指的是傳統(tǒng)的數(shù)值預(yù)測（numerical forecast）計算模型。

用“just as A, so too B”造句：

①勞逸結(jié)合：“Just as hard work is crucial for achieving success, so too is relaxation.”

②物質(zhì)與精神生活的同樣重要："Just as material comfortplays a significant role in our lives, so too does spiritual well-being."

9.公私合作項目

More public-private collaboration is?on the cards. By 2030, the European Commission hopes to have finished “Destination Earth”, a simulation that can handle both short-term weather patterns and longer-term changes in the climate.?

It hopes that users, with the help ofAI, will be able to visualise how animal?migration patterns?might change as temperatures rise, or?what might happen to?fish stocks?as the oceans warm.?Nvidia, whose chips?power?most of the world’s biggest?AI?models, has said it will participate.

The firm has also signed up to an even more ambitious plan for a network of “Earth?Virtualisation Engines”?proposed?at a meeting this month in Berlin by a group led by Bjorn Stevens, the director of the Max Planck Institute for Meteorology in Hamburg.

be on the cards 有可能發(fā)生

migration patterns 遷徙模式

fish stocks 漁業(yè)資源

“power”是動詞，可以翻譯為驅(qū)動

virtualisation engines?虛擬引擎

10.人工智能技術(shù)給氣象領(lǐng)域帶來新變化

Dr Stevens sees all this?ferment?as part of a shift in how information about the weather?is conceived of, produced and used. Turning?observationsinto something helpful like a forecast?used to require?a lot of expert knowledge, he says.?

That made it the?domain?of?a handful of?big institutions. But recent technological advances, especiallyAI, have made doing that both easier and cheaper.

“That makes [weather] data valuable,” he says. “And that is transforming everything.”

ferment?騷動

be conceived of?設(shè)想

observation?觀察數(shù)據(jù)

domain?領(lǐng)域

a handful of?少量的

used to do?表示過去一直存在的情況，可翻譯為“過去總是……”

歡迎指正，感謝閱讀。