订阅

多平台阅读

微信订阅

杂志

申请纸刊赠阅

订阅每日电邮

移动应用

商业 - 科技

大数据的最大挑战来自气候变化

Katherine Noyes 2014年07月07日

研究人员正在利用大数据技术来模拟、解读和演示气候变化对环境的影响。

    你知道吗,今天的全球海平面要比1880年的时候高出8英寸,而就在本世纪内,全球海平面预计还将上涨2到7英尺。另外,美国沿海地区有260万户家庭的500余万人口的住宅,在海水满潮时,只高出海平面不到4英尺。

    毫无疑问,气候变化是个大问题,不管导致它的原因是什么。

    那么如何计算气候对环境的影响呢?事实上,要进行这些复杂的计算,是一个极具挑战性的课题。要想了解气候变化对一国一地的影响水平,绝对不是在一张餐巾纸上写写画画就能算得出来的。

    这时你就需要大数据技术了。

    “上升的海平面”(Surging Seas)是由非盈利组织“气候中心”(Climate Central)开发的一款互动式地图工具,它用图形的形式详细描绘了海平面上升和风暴潮给美国大陆沿海3000多个城市、城镇和农村造成的威胁。它的细节可以精确到每一个街区——你可以搜索一个特定的地理位置,或是按照需要继续缩小目标范围。这个工具会与存在洪泛风险的地区进行匹配,并且提供相关实时报道、数据下载、行动计划、内嵌小工具和其它相关事项的链接。

    这种数据处理方式仅仅在几年前还是不可能实现的。

    能力有多大,困难就多大

    气候中心的战略沟通副总裁兼研究主任理查德•怀尔斯表示:“我们的战略是以人们能够理解的方式告诉他们当地的气候情况,唯一能实现这个目标的方法就是通过大数据分析。大数据让你能够简单、清晰地表达。”

    怀尔斯指出,目前主要有两种大数据形式可以用来帮助人们了解和应对气候变化。第一类是某些在近期才收集到的数据,但它们往往数据量极大且非常复杂,搁在以前很难对其进行有效分析,比如美国国家航空航天局(NASA)对各大城市的热成像绘图。怀尔斯表示,这种数据“一直到不久之前,还因为数据量过大而基本上没法处理,但是现在你已经可以在一台普通的电脑上处理它们了。”

    第二类大数据是一些相对较老但可能不那么可靠的数据。怀尔斯表示,这些数据“基本上一直都在那儿”,比如美国的历史气温趋势。这种数据一般不太复杂,但有可能存在不少缺口和误差。比如怀尔斯就指出:“1936年,俄克拉荷马州的某个负责量气温的家伙有可能不小心把温度计弄坏了。”这样的话,当年可能就有两个月根本没有气温记录。

    怀尔斯表示,要解决这些问题,现有的数据可以说“能力有多大,困难就有多大。但是大数据技术使得揭示一城一地的气候变化成为可能。”

    气候中心从政府的历史记录中获取原始数据,然后为美国各地的150余家地方电视台的天气预报节目制作高度本地化的气候图形,以阐释该地区的气候变化。比如怀尔斯指出:“今年六月,托雷多市变热了。我们一直利用这些数据试图让当地人了解气候变化趋势。”

    100万小时的计算

    气候中心的地图是阐释海平面上升情况的一个非常有效的工具。此外,大数据技术还能帮助研究人员模拟、分析和预测气候变化的影响。

    Global sea levels are about eight inches higher today than they were in 1880, and they are expected to rise another two to seven feet during this century. At the same time, some 5 million people in the U.S. live in 2.6 million coastal homes situated less than 4 feet above high tide.

    Do the math: Climate change is a problem, whatever its cause.

    The problem? Actually making those complex calculations is an extremely challenging proposition. To understand the impact of climate change at the local level, you’ll need more than back-of-the-napkin mathematics.

    You’ll need big data technology.

    Surging Seas is an interactive map and tool developed by the nonprofit Climate Central that shows in graphic detail the threats from sea-level rise and storm surges to all of the 3,000-plus coastal towns, cities, counties and states in the continental United States. With detail down to neighborhood scale—search for a specific location or zoom down as necessary—the tool matches areas with flooding risk timelines and provides links to fact sheets, data downloads, action plans, embeddable widgets, and other items.

    It’s the kind of number-crunching that was all but impossible only a few years ago.

    ‘Just as powerful, just as big’

    “Our strategy is to tell people about their climate locally in ways they can understand, and the only way to do that is with big data analysis,” said Richard Wiles, vice president for strategic communications and director of research with Climate Central. “Big data allows you to say simple, clear things.”

    There are actually two types of big data in use today to help understand and deal with climate change, Wiles said. The first is relatively recently collected data that is so voluminous and complex that it couldn’t be effectively manipulated before, such as NASA images of heat over cities, Wiles said. This kind of data “literally was too big to handle not that long ago,” he said, “but now you can handle it on a regular computer.”

    The second type of big data is older datasets that may be less-than-reliable. This data “was always kind of there,” Wiles said, such as historic temperature trends in the United States. That kind of dataset is not overly complex, but it can be fraught with gaps and errors. “A guy in Oklahoma may have broken his thermometer back in 1936,” Wiles said, meaning that there could be no measurements at all for two months of that year.

    Address those issues, and existing data can be “just as powerful, just as big,” Wiles said. “It makes it possible to make the story very local.”

    Climate Central imports data from historical government records to produce highly localized graphics for about 150 local TV weather forecasters across the U.S., illustrating climate change in each station’s particular area. For example, “Junes in Toledo are getting hotter,” Wiles said. “We use these data all the time to try to localize the climate change story so people can understand it.”

    ‘One million hours of computation’

    Though the Climate Central map is an effective tool for illustrating the problem of rising sea levels, big data technology is also helping researchers model, analyze, and predict the effects of climate change.

1 2 3 下一页

我来点评

  最新文章

最新文章:

500强情报中心

财富专栏