I. Please read the text after the speaker. 

Storms

7. Severe tropical storms cause deaths and affect people in many countries such as Japan, India, Thailand and the U.S. ①For instance, in 2019, Typhoon Hagibis cost Japan more than $10 billion in damages and insured losses.

8. Technology solutions: ②Advances in AI make it possible for a larger quantity of  weather data to be analyzed faster, more accurately and in greater detail, making weather predicting outcomes more accurate. Some experts have developed a framework based on machine learning to study cloud movements and shapes from satellite images. ③It has been found that “comma-shaped clouds” are strongly associated with storm formation. Researchers used computer vision and machine learning techniques to teach computers to automatically recognize and detect comma-shaped clouds in satellite images. Thanks to the technology of machine learning, computers can assist experts to identify the beginning of severe storms in real time, in an ocean of data.

Heat waves

9. In the summer of 2019, several thousand people died during the heat wave in Europe. In Japan, hundreds of people sought medical attention due to extremely high temperatures that rose above 105 degrees Fahrenheit (41 degrees Celsius). ④Heat waves not only cause health issues, increase economic spending and affect agricultural production, but also cause the loss of human lives.

10. Technology solutions: According to the latest studies, predicting heat waves is becoming increasingly important around the world, as their frequency is growing and the losses they bring cover various aspects. Researchers recently developed a new technique to predict heat waves with the power of deep learning. One example is the “CapsNets” system developed by some researchers from an American university. It can be trained to compare current weather patterns to similar ones from the past. ⑤It also learns to identify some patterns of atmospheric circulation which lead to a heat wave so that it can make relatively accurate predictions.

Floods

11. About 1,600 people died in India in 2019 due to flash floods. In March of the same year, torrential rain forced more than 4,000 Indonesians out of their homes. Floods have also heavily affected places such as Nepal, Myanmar, and China.

12. Technology solutions: Floods caused by natural factors, such as increased rainfall or storms, can be predicted with sophisticated technological models. AI can help technologists better interpret a growing amount of data and create forecasting models that can automatically alert authorities. Recently, a U.S. tech company announced it had begun implementing a flood forecasting model in an Indian region. ⑥Partnered with several institutions, it built an inundation model that can predict the water level rise in a particular area, and assess the “water behavior” or what areas can be flooded very quickly.

II. Text analysis

①For instance, in 2019, Typhoon Hagibis cost Japan more than $10 billion in damages and insured losses. (Line 2-3, Para. 7)

翻译：例如，在2019年，台风“海贝思”造成日本超百亿美元的损毁及保险损失。

②Advances in AI make it possible for a larger quantity of  weather data to be analyzed faster, more accurately and in greater detail, making weather predicting outcomes more accurate. (Line 1-3, Para. 8)

翻译：人工智能的进步使得更快、更准确、更精细地分析大量天气数据成为可能，从而使天气预报结果更为精确。

③It has been found that “comma-shaped clouds” are strongly associated with storm formation. (Line 4-5, Para. 8)

翻译：业已发现“逗点云型”与风暴形成紧密相关。

④Heat waves not only cause health issues, increase economic spending and affect agricultural production, but also cause the loss of human lives. (Line 4-5, Para. 9)

翻译：热浪不仅导致健康问题，增加经济消耗，影响农业生产，而且造成生命损失。

⑤It also learns to identify some patterns of atmospheric circulation which lead to a heat wave so that it can make relatively accurate predictions. (Line 6-8, Para. 10)

翻译：它还能识别导致热浪的大气环流模式，因而能够做出相对准确的预测。

⑥Partnered with several institutions, it built an inundation model that can predict the water level rise in a particular area, and assess the “water behavior” or what areas can be flooded very quickly. (Line 5-7, Para. 12)

翻译：它与多家机构合作，建立了一个洪水模型，可以预测特定地区的水位上升，并评估“水行为”或可能很快被淹没的区域。

III. Test 

  Storms

  7. Severe tropical storms cause deaths and affect people in many countries such as Japan, India, Thailand and the U.S. ①For instance, in 2019, Typhoon Hagibis cost Japan more than $10 billion in damages and insured losses.             8. Technology solutions: ②Advances in AI make it possible for a larger quantity of  weather data to be analyzed faster, more accurately and in greater detail, making weather predicting outcomes more accurate. Some experts have developed a framework based on machine learning to study cloud movements and shapes from satellite images. ③It has been found that “comma-shaped clouds” are strongly associated with storm formation.  Researchers used computer vision and machine learning techniques to teach computers to automatically recognize and detect comma-shaped clouds in satellite images. Thanks to the technology of machine learning, computers can assist experts to identify the beginning of severe storms in real time, in an ocean of data.             Heat waves

  9. In the summer of 2019, several thousand people died during the heat wave in Europe. In Japan, hundreds of people sought medical attention due to extremely high temperatures that rose above 105 degrees Fahrenheit (41 degrees Celsius). ④Heat waves not only cause health issues, increase economic spending and affect agricultural production, but also cause the loss of human lives.

10. Technology solutions: According to the latest studies, predicting heat waves is becoming increasingly important around the world, as their frequency is growing and the losses they bring cover various aspects. Researchers recently developed a new technique to predict heat waves with the power of deep learning. One example is the “CapsNets” system developed by some researchers from an American university. It can be trained to compare current weather patterns to similar ones from the past. ⑤It also learns to identify some patterns of atmospheric circulation which lead to a heat wave so that it can make relatively accurate predictions.

  Floods

  11. About 1,600 people died in India in 2019 due to flash floods. In March of the same year, torrential rain forced more than 4,000 Indonesians out of their homes. Floods have also heavily affected places such as Nepal, Myanmar, and China.

12. Technology solutions: Floods caused by natural factors, such as increased rainfall or storms, can be predicted with sophisticated technological models. AI can help technologists better interpret a growing amount of data and create forecasting models that can automatically alert authorities. Recently, a U.S. tech company announced it had begun implementing a flood forecasting model in an Indian region. ⑥Partnered with several institutions, it built an inundation model that can predict the water level rise in a particular area, and assess the “water behavior” or what areas can be flooded very quickly.