(じんこうちのう)

artificial intelligence

(てんきよほう)

Weather forecast

(かわる)

to change

(つかう)

to use

(けんきゅう)

research, study

(せかいじゅう)

all over the world

(きゅうそく)

rapid; swift

(すすむ)

to advance, to progress

(これまで)

until now / so far

(よほう)

forecast

(たいき)

the atmosphere

(うごき)

movement; trend

(ぶつり)

physics

(しき)

ceremony

(けいさん)

calculation; computation

(やりかた)

way of doing, method

(ちゅうしん)

center; core

(ほうほう)

method; way

(せいど)

accuracy; precision

(たかい)

expensive; high

(おおがた)

large-scale; large-size

(こんぴゅーたー)

computer

(ながい)

long

(けいさんじかん)

calculation time; computing time

(ひつよう)

necessary / needed

(あたらしい)

new

(しゅほう)

technique; approach

(かこ)

the past

(すうじゅうねん)

several decades

(かんそく)

observation (scientific)

(でーた)

data

(がくしゅう)

learning; study

(さき)

ahead, future

(てんき)

weather

(よそく)

prediction; to predict

(かかる)

to take (time/money), cost

(じかん)

time

(じゅうらい)

conventional; up to now

(ずっと)

continuously; all along

(みじかい)

short

(ふつう)

usually; ordinary; normal

(きかい)

machine

(うごく)

to move

(たいふう)

typhoon

(しんろ)

course, path

(うわまわる)

to exceed; to surpass

(けっか)

result

(ほうこく)

report

(ただし)

however, provided that

(げんしょう)

phenomenon

(よわい)

Weak

(いう)

to say

(きろくてき)

record-breaking

(ごうう)

torrential rain; downpour

(ぜんれい)

precedent

(すくない)

few, scarce

(じたい)

situation; state of affairs

(はんだん)

judgement; decision

(あやまる)

to make a mistake; to get wrong

(きしょう)

weather / meteorological

(せんもんか)

expert, specialist

(とうめん)

for the time being

(りょうほう)

both

(くみあわせる)

to combine; to put together

(してき)

pointing out, indication

is doing, has been doing (ongoing state/action)

Expresses a continuing action or resulting state. Formed with て-form + いる. Can describe an action in progress or a state resulting from a past change.

on the other hand

Introduces a contrasting point of view, meaning 'on the other hand'. It pivots from the benefits to the drawbacks.

causative (make/let do)

Causative form. 自由にさせる means 'to let (them) be free'.

than / more than

Indicates comparison. Used to say prices rose on average 2.8% compared to last year.

to become (い-adjective)

Expresses a change of state for い-adjectives. Drop the final い and add くなる. Indicates something has changed or will change to a new state.

Even ~; also at ~

Particle でも after a noun means 'even' or 'also at.' 首相官邸の庭でも means 'even at the Prime Minister's residence garden.'

called/named ~

Used to introduce or name something. In this article, 「鳥貴族」という店 means 'a store called Torikizoku'.

passive form

Passive voice used for the victim: 殺されている (was murdered), 見つかりました (was found).

quotative particle

Marks the content of what someone said or reported. Used with 言う here.

like ~; as if ~

Simile expression comparing something to another thing.

there is a fear/risk that

In 「影響が広がる恐れがある」 it warns of the risk that the impact on citizens' lives could spread.

should ~ / ought to ~

States what is the right thing to do. 組み合わせて使うべきだ is the experts' recommendation that both forecasting methods be used together for now.

人工知能で天気予報が変わる

Artificial intelligence is changing the weather forecast

人工知能を使った天気予報の研究が、世界中で急速に進んでいます。

Research into weather forecasting that uses artificial intelligence is advancing rapidly around the world.

これまでの予報は、大気の動きを物理の式で計算するやり方が中心でした。

Forecasting up to now has centred on a method that calculates the movement of the atmosphere using the equations of physics.

この方法は精度が高い一方で、大型のコンピューターと長い計算時間が必要です。

While this method is highly accurate, it requires a large computer and a long calculation time.

新しい手法は、過去数十年の観測データを人工知能に学習させ、そこから先の天気を予測します。

The new approach has an artificial intelligence learn observation data from the past several decades, and predicts the weather from that point onward.

計算にかかる時間は従来よりずっと短くなり、ふつうの機械でも動くといいます。

The time the calculation takes is said to have become far shorter than before, and the model reportedly runs even on an ordinary machine.

台風の進路予測では、これまでの予報を上回る結果も報告されています。

In predicting the paths of typhoons, results that surpass conventional forecasts have also been reported.

ただし、人工知能はデータにない現象に弱いと言われています。

However, artificial intelligence is said to be weak at phenomena that do not appear in its data.

記録的な豪雨のように前例の少ない事態では、判断を誤る恐れがあります。

In situations with few precedents, such as record-breaking torrential rain, there is a risk that it will judge wrongly.

気象の専門家は、当面は両方の手法を組み合わせて使うべきだと指摘しています。

Weather experts point out that for the time being both approaches should be combined and used together.