- 精华
- 1
- 积分
- 2882
- 经验
- 2882 点
- 威望
- 258 点
- 金钱
- 760 ¥
- 魅力
- 440
|
近似原文:(已确定哦!!激动哈!感谢vivian0311 MM)
60-Year-Old Bias in Data on Sea Temperatures
Every scientist knows about measurement bias, the systematic errors that can creep into data. A simplistic example would be a metal ruler that gives inaccurate readings when it expands in warmer temperatures.
Avoiding measurement bias can be impossible. Nobody’s perfect, after all, and neither is equipment. What’s important is recognizing bias and taking it into account when working with the data.
But it is not always easy to detect bias. A case in point is being reported in Nature, where researchers have uncovered measurement bias in six-decades-old data on global surface temperatures. The problem arises from how British and American ships measured the temperature of surface water.
David W. J. Thompson of Colorado State University and colleagues analyzed global mean surface temperatures from the 1880s to the present, tweaking(拧,扭,作调整) the data to remove the effects of El Niño and other “weather noise,” as Dr. Thompson described it. They discovered a sudden drop of 0.5 degree Fahrenheit in 1945, but just in data collected at sea, not on land — a hint that the drop might have something to do with the measurements.
At that time, British and American ships did much of the logging of sea temperatures worldwide. On British ships, crews measured the temperature of seawater collected in a bucket(水桶). But since about 1939, most American ships had switched to measuring the temperature of seawater as it was drawn through an intake pipe for use as an engine coolant(冷冻剂,冷却液,散热剂). Because of heat from the engine room, American measurements were generally higher.
After looking more closely at the data, Dr. Thompson said, they realized what had happened. Most of the wartime(战时) data came from American ships, with just 20 percent of the readings from British ones. But starting in August 1945, there was an abrupt switch. Nearly half the readings came from British ships. Because those readings are generally colder, Dr. Thompson said, that accounts for the sudden temperature drop.
Now that the bias has been recognized, climate researchers will need to take it into account in their models and simulations. But Dr. Thompson said the long-term trend, toward higher temperatures, would not be affected.
文章不是完全一样但是差不多 |
|