Common Climate Misconceptions

Sorting Through George Will’s Analysis; What the Sea Ice Data Actually Tells Us

The recent brouhaha (see related story, this posting) initiated by conservative columnist George Will’s February 15th syndicated column centers in part around his assertion that global sea ice levels now equal those of 1979, belying concerns of melting ice caps.

Will’s statement may be technically true, but it is also broadly misleading, as it cherry-picks two single months 30 years apart and suggests that their similar values imply that no trend in sea ice exists. Few serious climatologists would reach such a conclusion based on that data, and in reality, there is a negative trend in sea ice over the past 30 years, as can be clearly seen by examining actual sea ice data.

In his column, Will argued that “According to the University of Illinois’ Arctic Climate Research Center, global sea ice levels now equal those of 1979.”

Will based this argument on a blog post on the technology news-site Daily Tech entitled “Sea Ice Ends Year at Same Level as 1979″ by Michael Asher. According to sea ice extent data from the National Snow and Ice Data Center, there were 23.95 million kilometers of sea ice worldwide in December 1979 and 24.73 million kilometers in December 2009.

So were Will and Asher correct in their assertion? Technically, yes, but their omission of any examination of the data over the past 30 years apart from those two single months leaves a deeply misleading impression.

Global sea ice shows a strong annual cycle. In the Northern Hemispheric summer there is considerably less arctic ice in the summer than in the winter, and in the winter there is considerably less Antarctic sea ice than in the summer. It can be hard to clearly see a trend over time in sea ice given the sheer magnitude of the annual cycle. The raw global sea ice extent data from the National Snow and Ice Data Center is shown below.

View larger image
Global Monthly Sea Ice Extent data taken from the National Snow and Ice Data Center.

To determine the trend in sea ice extent in light of the magnitude of annual variation, scientists use a simple statistical technique to remove the annual cycle. They calculate the monthly sea ice anomaly; that is, how each January compares to the average January over the period from 1979-2000, each February to the average February, etc.

Plotting these anomalies leads to the following figure:

View larger image
Global Monthly Sea Ice Extent Anomaly in millions of square kilometers calculated based on data from the National Snow and Ice Data Center.

To better understand the step that Will and Asher took, it is useful to highlight the individual months used in their analysis in the context of the overall 30-year trend. The data points used by Will and Asher are shown in the figure highlighted in blue, with the trend that best fits the data shown in red.

It is clear that Will and Asher based their conclusions on a point below the trend (1979) and a point well above the trend (2009). That results in their drawing a misleading conclusion about the intervening period.

View larger image
Global Monthly Sea Ice Extent Anomaly in millions of square kilometers calculated based on data from the National Snow and Ice Data Center. Blue points represent the two used by Will and Asher in their argument.

Additionally, the focus on global sea ice is itself somewhat misleading. The University of Illinois’ Arctic Climate Research Center pointed out the following after Asher’s blog was published as Daily Tech,

In the context of climate change, global sea ice area may not be the most relevant indicator. Almost all global climate models project a decrease in the Northern Hemisphere sea ice area over the next several decades under increasing greenhouse gas scenarios. But, the same model responses of the Southern Hemisphere sea ice are less certain. In fact, there have been some recent studies suggesting the amount of sea ice in the Southern Hemisphere may initially increase as a response to atmospheric warming through increased evaporation and subsequent snowfall onto the sea ice.

While global sea ice appears to be declining, Northern Hemispheric sea ice has unambiguously – and at a statistically significantly rate – declined over the past 30 years, as shown in the figure below.

View larger image
Northern Hemispheric Monthly Sea Ice Extent Anomaly in millions of square kilometers calculated based on data from the National Snow and Ice Data Center.

So what can journalists making their way through the Will/Revkin/et. al contretemps learn from this whole affair?

George Will very likely took the argument from the Daily Tech article in good faith and reported it in his column. Neither he nor other fact checkers sufficiently examined the data itself or sought expert scientific vetting, however. Had they done so, they likely would have realized that cherry picking two months as a basis for a comparison and conclusion is highly misleading when the data exhibit considerable variability.

When reporting on climate trends, journalists – columnists and opinion writers no less than news reporters – should make sure to use arguments and data backed up by robust statistical analysis. Only by doing so can they ensure that the results they are reporting are actually meaningful.

Special thanks to Tamino for providing the idea of performing this analysis.

Zeke Hausfather

Zeke Hausfather, a data scientist with extensive experience with clean technology interests in Silicon Valley, is currently a Senior Researcher with Berkeley Earth. He is a regular contributor to The Yale Forum (E-mail: zeke@yaleclimatemediaforum.org, Twitter: @hausfath).
Bookmark the permalink.

Comments are closed.