1. Confidence is growing

A 2004 paper published in Nature was one of the first widely cited studies in the field of climate attribution. The paper was published a year and a half after Europe suffered its hottest summer in centuries, when crops failed, glaciers in the Alps shrank by 10 per cent and more than 30,000 people died. The study's lead author, Peter A. Stott, A climatologist at the Met Office's Hadley Centre for Climate Science and Services, concluded that the impact of human activity had at least doubled the probability of record-breaking heat waves. Stephanie Herring, a climate scientist at the US National Oceanic and Atmospheric Administration (NOAA), says that the problem was "very difficult" at the time and was only of interest to researchers who truly loved statistical probability, atmospheric physics and aerodynamics. Since 2012, when Herring began publishing annual reports on extreme weather events for the American Meteorological Society, "none of us could have predicted that this area of research would end up getting so much public attention."

In October 2012, Hurricane Sandy hit New York City and New Jersey, bringing public attention to the storm. People are wondering how this storm could have done so much damage so far north. (Hurricanes usually make landfall and cause damage along the mid-to-low latitudes, rarely moving north to the mid-to-high latitudes.) Analysis published a few months later suggested that the melting of large amounts of Arctic sea ice earlier that year, which created large areas of open water that absorbed more solar radiation, may have exacerbated Sandy's intensity - though the link was described only as "a plausible possible mechanism". Kevin Trenberth, a prominent scholar at the Center for Atmospheric Research (NCAR) in the United States, has said that scientists are often too conservative and often underestimate the impact of climate change. Most researchers also remain reluctant to discuss the specific effects of climate change on a single storm.

Advances in analytical methods have made it easier for scientists to isolate the effects of climate change. In the first years of the 21st century, few institutions had high-performance computer systems that could run climate models containing large amounts of data. Today, researchers can do this from a laptop at home via cloud services, and can even combine multiple patterns to run them multiple times, increasing the reliability of the results. The accuracy of the climate models themselves has also generally improved, such as higher resolution, which allows them to provide more precise information about specific locations.

With the development of the attribution field, two research methods have emerged. One of these, called probabilistic event attribution, is used to estimate the extent to which human activity increases the probability of a certain extreme weather event, such as a heat wave. Scientists use numerical models to simulate extreme weather events, give them a scenario in which climate change does not occur, and then compare the results under the climate change scenario to determine whether factors such as increased emissions increase the probability of a particular extreme event. The WWA's first study, for example, compared temperatures in five French cities during the 2015 heat wave with summer temperatures in the first half of the 20th century, proving that climate change did indeed increase the probability of heat waves by four times.

As a contrast, another approach focuses on causation or "story lines." This approach asks questions about specific disasters, such as whether climate change has increased the amount of precipitation in a particular storm. The answer lies in thermodynamic changes, such as warmer air holding more water vapor, which leads to increased precipitation.

Trenberth was one of the early proponents of the "story line" approach. At first, he notes, the scientific debate between the two approaches was intense. In 2014, after a major flood in Boulder, Colorado, NOAA's Martin Herring published a study showing that climate change had not increased the probability of heavy rainfall in the region. Trenberth and his colleagues dispute the results. Trenberth argues that the study did not take into account Mexico's warm sea surface temperatures, where large amounts of water would evaporate into the atmosphere, increasing total precipitation. Herring responded that Trenberth's explanation seemed clear but oversimplified the issue. A news report in the journal Nature details the debate.

Over time, people on both sides have come to realize that the two approaches actually complement each other. "Ideally, you should do both," Mr. Otto said. "They're looking at things from different angles." Elizabeth Lloyd, a member of the American Academy of Arts and Sciences, also said that both methods can provide important information about climate risk. Combining these two approaches can tell policymakers whether roads and Bridges need to cope with more intense precipitation, and emergency managers how often they will need to restrict access to those roads and Bridges in the future because of storms.

In 2017, when Hurricane Harvey stayed in the Houston area for several days, the attribution field took another step forward. At that time, the hurricane brought up to 150 centimeters of rain to the region, far exceeding previous records. Trenberth found that the extremely warm waters of the Gulf of Mexico at the time produced more evaporation than normal, which directly led to excessive precipitation. On the other hand, Otto and his colleagues have also conducted an independent analysis showing that climate change has increased precipitation by 15 percent. Trenberth said governments in hurricane-prone areas need to develop contingency plans for larger floods, including upgrading evacuation routes, adjusting building codes and revamping power grids.

1. Confidence is growing

A 2004 paper published in Nature was one of the first widely cited studies in the field of climate attribution. The paper was published a year and a half after Europe suffered its hottest summer in centuries, when crops failed, glaciers in the Alps shrank by 10 per cent and more than 30,000 people died. The study's lead author, Peter A. Stott, A climatologist at the Met Office's Hadley Centre for Climate Science and Services, concluded that the impact of human activity had at least doubled the probability of record-breaking heat waves. Stephanie Herring, a climate scientist at the US National Oceanic and Atmospheric Administration (NOAA), says that the problem was "very difficult" at the time and was only of interest to researchers who truly loved statistical probability, atmospheric physics and aerodynamics. Since 2012, when Herring began publishing annual reports on extreme weather events for the American Meteorological Society, "none of us could have predicted that this area of research would end up getting so much public attention."

In October 2012, Hurricane Sandy hit New York City and New Jersey, bringing public attention to the storm. People are wondering how this storm could have done so much damage so far north. (Hurricanes usually make landfall and cause damage along the mid-to-low latitudes, rarely moving north to the mid-to-high latitudes.) Analysis published a few months later suggested that the melting of large amounts of Arctic sea ice earlier that year, which created large areas of open water that absorbed more solar radiation, may have exacerbated Sandy's intensity - though the link was described only as "a plausible possible mechanism". Kevin Trenberth, a prominent scholar at the Center for Atmospheric Research (NCAR) in the United States, has said that scientists are often too conservative and often underestimate the impact of climate change. Most researchers also remain reluctant to discuss the specific effects of climate change on a single storm.

Advances in analytical methods have made it easier for scientists to isolate the effects of climate change. In the first years of the 21st century, few institutions had high-performance computer systems that could run climate models containing large amounts of data. Today, researchers can do this from a laptop at home via cloud services, and can even combine multiple patterns to run them multiple times, increasing the reliability of the results. The accuracy of the climate models themselves has also generally improved, such as higher resolution, which allows them to provide more precise information about specific locations.

With the development of the attribution field, two research methods have emerged. One of these, called probabilistic event attribution, is used to estimate the extent to which human activity increases the probability of a certain extreme weather event, such as a heat wave. Scientists use numerical models to simulate extreme weather events, give them a scenario in which climate change does not occur, and then compare the results under the climate change scenario to determine whether factors such as increased emissions increase the probability of a particular extreme event. The WWA's first study, for example, compared temperatures in five French cities during the 2015 heat wave with summer temperatures in the first half of the 20th century, proving that climate change did indeed increase the probability of heat waves by four times.

As a contrast, another approach focuses on causation or "story lines." This approach asks questions about specific disasters, such as whether climate change has increased the amount of precipitation in a particular storm. The answer lies in thermodynamic changes, such as warmer air holding more water vapor, which leads to increased precipitation.

Trenberth was one of the early proponents of the "story line" approach. At first, he notes, the scientific debate between the two approaches was intense. In 2014, after a major flood in Boulder, Colorado, NOAA's Martin Herring published a study showing that climate change had not increased the probability of heavy rainfall in the region. Trenberth and his colleagues dispute the results. Trenberth argues that the study did not take into account Mexico's warm sea surface temperatures, where large amounts of water would evaporate into the atmosphere, increasing total precipitation. Herring responded that Trenberth's explanation seemed clear but oversimplified the issue. A news report in the journal Nature details the debate.

Over time, people on both sides have come to realize that the two approaches actually complement each other. "Ideally, you should do both," Mr. Otto said. "They're looking at things from different angles." Elizabeth Lloyd, a member of the American Academy of Arts and Sciences, also said that both methods can provide important information about climate risk. Combining these two approaches can tell policymakers whether roads and Bridges need to cope with more intense precipitation, and emergency managers how often they will need to restrict access to those roads and Bridges in the future because of storms.

In 2017, when Hurricane Harvey stayed in the Houston area for several days, the attribution field took another step forward. At that time, the hurricane brought up to 150 centimeters of rain to the region, far exceeding previous records. Trenberth found that the extremely warm waters of the Gulf of Mexico at the time produced more evaporation than normal, which directly led to excessive precipitation. On the other hand, Otto and his colleagues have also conducted an independent analysis showing that climate change has increased precipitation by 15 percent. Trenberth said governments in hurricane-prone areas need to develop contingency plans for larger floods, including upgrading evacuation routes, adjusting building codes and revamping power grids.