《2026年江苏13大市中考名卷优选38套英语》

C
A single researcher with a desktop computer can now make correct weather forecasts using a new AI method that is much faster and uses far less computing power than traditional systems.
At present, weather forecasts are made through many difficult steps, each taking hours. These systems need supercomputers and large teams of experts to develop and keep them working.
However, Aardvark Weather offers a new method by training AI to collect data from weather stations, satellites, weather balloons, ships, and planes around the world to make predictions. This promises improvements in forecast speed, correctness, and cost, according to research published in Nature by the University of Cambridge, the Alan Turing Institute, Microsoft Research, and the European Centre for Medium-Range Weather Forecasts(ECMWF).
Professor Richard Turner from the University of Cambridge explained that this new method can quickly create specialized forecasts for different industries or locations. For example, it could guess temperatures for farms in Africa or wind speeds for wind energy companies in Europe. Traditional weather systems require large teams and supercomputers to deal with data and create forecasting models, taking hours for just one forecast.
"This is a totally different method," Turner said, adding that the model would finally be able to provide correct eight-day forecasts, compared to the present five-day limit, and offer more localized predictions.
Dr Scott Hosking from the Alan Turing Institute said that this breakthrough could make weather forecasting available to more countries, especially developing countries. It will also help governments, emergency teams, and businesses that need weather predictions.
Dr Anna Allen, the lead author of the paper, highlighted that this research could improve forecasts of natural disasters like hurricanes and wildfires, as well as problems like air quality and sea ice predictions.
Aardvark builds from research from Huawei, Google, and Microsoft, replacing a key step in the traditional method with AI, which makes forecasts faster and more correct. It is already in use at the ECMWF.
Researchers found that Aardvark only needs 10% of the data used by present systems, yet in some areas, it does better than the US National GFS forecasting system and matches the correctness of the US Weather Service.
33. What is the main advantage of the new AI weather prediction method?
A. It depends on supercomputers.
B. It requires large teams of experts.
C. It predicts for longer periods of time.
D. It works faster with less computing power.
34. What makes the new weather prediction model different from traditional systems?
A. It mainly predicts weather for developed countries.
B. It can hardly predict natural disasters and air quality.
C. It gathers data from different sources across the world.
D. It only pays attention to specific industries like agriculture.
35. What can we infer about the new weather prediction method?
A. It doesn't need human effort.
B. It is more expensive to use.
C. It requires less data to work.
D. It doesn't develop specialized forecasts.
36. What is the purpose of the article?
A. To describe weather forecasting history.
B. To introduce a new AI weather model.
C. To criticize traditional weather systems.
D. To explain the role of supercomputers.

答案： 33.D 细节理解。根据第一段中much faster and uses far less computing power than traditional systems可知，新方法的主要优势是速度快且计算能力需求低。
34.C 细节理解。根据第三段第一句可知，新系统的差异在于它从全球不同来源收集数据。
35.C 推理判断。根据最后一段可知，研究人员发现Aardvark仅需现有系统10％的数据量，由此可推断，它对数据的需求量更少。
36.B 写作意图。根据首段及全文可知，本文主要是介绍了一种新的AI天气预报方法Aardvark Weather。