2026年2月9日 / 美国东部时间下午5:30 / 哥伦比亚广播公司新闻
西班牙两所大学的研究人员发现，在过去约70年里，美国本土84%的州都出现了变暖迹象，这一比例高于此前的研究结论。
上周发表在《PLOS Climate》期刊上的报告显示，自上世纪中期以来，美国41个州的气温持续上升。该研究未涵盖阿拉斯加和夏威夷州。
马德里卡洛斯三世大学的经济学家兼教授赫苏斯·贡萨洛（Jesús Gonzalo）用一个谜语解释说，这些州的变暖模式并不统一。在西班牙，这个讽刺性的统计笑话是：”有两个饥饿的人争夺一只鸡。一个人吃掉了整只鸡，让另一个人挨饿。所以，平均来说，每个人吃了半只鸡。”
在这种情况下，人均平均食物消耗量未能传达出一个重要事实：在计算食物摄入量的两个人中，有50%的人根本没有吃到任何东西。
“平均数字说明不了太多问题，对吧？”贡萨洛说。
这个例子体现了贡萨洛与其合著者报告的主要结论，即敦促研究人员和政策制定者在评估温度趋势与气候变化的关系时，不要只看数学平均值。
贡萨洛和他的合著者、西班牙萨拉戈萨大学的玛丽亚·多洛雷斯·加德亚·里瓦斯（María Dolores Gadea Rivas）在报告中写道，尽管”气候系统是全球性运作的，但其影响是局部性的，导致显著的区域差异。”
为了检测这些变化，他们分析了美国本土数万份每日气温记录，其中一些可以追溯到1950年。
数据显示，气候变化对气温的影响在美国比之前认为的范围更广，且可测量。因为相当多的气候研究使用平均值来量化这些影响。平均值（也称为算术”均值”）是通过将某一批数值（例如加利福尼亚州记录的气温）相加，然后除以该总和中的数值数量计算得出。
贡萨洛表示，平均值本应总体反映所计算数值的趋势，但并不总是能全面地说明情况。他指出，在某些情况下，依赖平均值可能会产生误导。
“美国可能是气候最多样化的国家之一。主要信息是，如果你想衡量多样性，不要使用平均值，”贡萨洛说，”如果你只关注平均值，可能会犯错误。”
贡萨洛的新气候报告认为，只关注平均气温及其随时间的变化或不变，会给气候科学家和所有参与气候政策的人带来严重风险。
研究发现，1950年至2021年间，美国27个州的平均气温上升，这与早期研究结果一致。但他们还发现，另外14个州的气温也显著上升。贡萨洛指出，这一巨大差异可能会改变美国应对这一问题的方式。
“当你试图设计更有效的缓解和适应政策时，这一点很重要，”他说。
该报告没有依赖平均值，而是研究了特定州的全部温度范围（从最低到最高）并追踪其随时间的变化——就像人们会考虑特定地区的全部收入范围以准确了解其财富分布和进展一样。报告揭示了其作者所说的”此前隐藏的变暖模式”，这些模式遍布美国各地。
研究人员发现，包括加利福尼亚州、爱达荷州、内华达州、俄勒冈州、华盛顿州和怀俄明州在内的西部各州，最高气温上升幅度最大；而爱荷华州、明尼苏达州、蒙大拿州、内布拉斯加州和达科他州等中部各州，多年来最低气温的上升幅度相对高于最高气温。
北部各州（包括中西部、东北部和西北部）的整个温度范围内都出现了类似的上升，这意味着较冷的温度和较暖的温度上升速度相当。
贡萨洛说，所有这些趋势似乎都在加速，针对实际情况调整公共应对策略将特别有益。
异常的”变暖空洞”
报告称，美国只有德克萨斯州、俄克拉荷马州、密西西比州、阿拉巴马州、佐治亚州、堪萨斯州和阿肯色州没有显示出具有统计学意义的变暖迹象。
这个位于美国中部和南部的州列表与一个已知现象——”变暖空洞”（warming hole）——相符。它描述的是尽管气候变化但气温并未上升的地区。科学家长期以来一直认为这是一种异常现象。
美国国家海洋和大气管理局（NOAA）称，研究表明，该地区反直觉的降温可能是由气溶胶、土地利用变化或水循环变化引起的，这些因素在某些地方可能因温室气体排放而加剧。
2023年，NOAA表示需要开展更多研究来回答其中一些问题。
41 U.S. states are getting warmer, each in its own way, report says
February 9, 2026 / 5:30 PM EST / CBS News
Researchers at two Spanish universities have found that 84% of the contiguous U.S. states have shown signs of warming over the last 70 or so years, which is more than previously suggested.
In a report published last week in the journal PLOS Climate, the researchers found that 41 U.S. states have gotten warmer since the middle of the last century. Alaska and Hawaii weren’t included in the study.
The states’ warming patterns weren’t uniform, Jesús Gonzalo, an economist and professor at the University of Charles III in Madrid, explained using a riddle. In Spain, the satirical statistics joke goes: “There are two hungry people vying over a single chicken. One eats the entire thing, leaving the other to starve. So, on average, each person has eaten half of a chicken.”
In that scenario, the average amount of food consumed per person failed to convey the important fact that 50% of the pair whose chicken intake was being calculated had nothing to eat at all.
“The average doesn’t say much, no?” Gonzalo said.
Its premise exemplified the main conclusion of the report that Gonzalo co-authored, which urged researchers and policymakers to look beyond mathematical averages when evaluating how temperature trends relate to climate change.
In the report, Gonzalo and his co-author, María Dolores Gadea Rivas of Spain’s University of Zaragoza, wrote that although “the climate system functions on a global scale, its effects are experienced locally, resulting in significant regional variability.”
To detect variations, they analyzed tens of thousands of daily temperature readings collected across the contiguous U.S., some of which dated back to 1950.
The data showed that the effects of climate change on temperature are existent and measurable in a larger swath of the U.S. than previously thought, since a fair amount of climate research quantifies those effects using averages. Also referred to as the arithmetic “mean,” an average is calculated by adding together a certain batch of values — temperatures recorded in the state of California, for instance — and dividing that by the number of values included in that sum.
Averages are meant to generally reflect the tendencies of values involved in their calculations. But they don’t always provide a complete picture, according to Gonzalo, who says relying on them can be misleading in some instances.
“The U.S. is maybe one of the countries with the most heterogeneous climate. And the main message is, if you want to measure heterogeneity, don’t use the average,” Gonzalo said. “If you pay attention to the average, you can make a mistake.”
Gonzalo’s new climate report argued that focusing exclusively on average temperatures, and how they have or haven’t increased over time, can have serious pitfalls for climate scientists and anyone involved in climate policy.
An aerial view of construction of new ski trails and a ski lift on February 8, 2026 in Park City, Utah. Mario Tama / Getty Images
The study found that average temperatures rose in 27 U.S. states between 1950 and 2021, corroborating earlier research. But they also increased significantly in 14 more, which, Gonzalo noted, is a huge difference that could potentially alter the country’s approach to addressing the issue.
“This is consequential when you try to design more efficient mitigation and adaptation policies,” he said.
Instead of averages, the report examined the full range of temperatures in a given state, from lowest to highest, and tracked them over time — the same way one might consider the full range of incomes in a particular place to accurately understand its wealth distribution and progress. The report brought to light what its authors described as “previously hidden patterns” of warming in every corner of the nation.
Western states, including California, Idaho, Nevada, Oregon, Washington and Wyoming, saw the largest increases in their highest temperatures, while Central states, like Iowa, Minnesota, Montana, Nebraska and the Dakotas saw relatively greater jumps in lower temperatures than higher ones over the years, the researchers found.
Northern states, across the upper Midwest, Northeast and Northwest, saw similar increases across the whole of their temperature ranges, meaning that cooler temperatures and warmer temperatures rose at comparable paces.
All of those trends appear to be accelerating, Gonzalo said, and tailoring public response strategies to realistically fit the scope of the situation at hand could be especially beneficial.
The anomalous “warming hole”
The only U.S. states that haven’t shown statistically significant signs of warming are Texas, Oklahoma, Mississippi, Alabama, Georgia, Kansas and Arkansas, the report said.
That list of states in the central and southern U.S. is consistent with a known phenomenon called a “warming hole.” It describes a region that hasn’t experienced rising temperatures despite climate change. Scientists have long considered it an anomaly.
Research has suggested that counterintuitive cooling in that area could be induced by aerosols, changes in land use or the water cycle, which may have intensified in certain places as a response to greenhouse gas emissions, according to the National Oceanic and Atmospheric Administration.
The NOAA said in 2023 that additional studies should be done to answer some of those questions.