Category Archives: News

Google’s Year of Forgetting

Google has gotten better at forgetting. A year ago, a European court ruled that Google search results in the European Union were subject to European data-protection rules. That meant that while private individuals might not be able to force a newspaper to retract an irrelevant or outdated story about them, they could ask Google to remove links to the story. Despite a slow start, the search giant has now caught up with the requests. In the meantime, Americans, Japanese, Koreans, and others around the world are proposing the adoption of similar privacy-protection policies.

Google—which can claim 93 percent of the European search market, according to StatCounter—began removing certain links from search results on its EU pages in June. The removals were in response to a 13 May 2014 ruling by the Court of Justice of the European Union that enables residents to request the removal of search engine results that point to prickly parts of their past. If it did anything, the ruling proved that personal privacy is popular: Google got over 41,000 requests in the first four days it accepted them. Requests later leveled off at around 1,000 per day.

The off-line analogy might be a guest asking a party host not to bring up an old, reputation-tarnishing story. It is still up to the host to decide whether to comply, and other guests may still whisper—or shout. But thanks to the CJEU ruling, search engine results in the EU are now subject to the same data-protection rules as other company-held personal data. It’s as if party attendees could now ask the party host to keep mum, with the threat of appealing to a national agency.

Yet the ruling left open many questions about how to comply. Soon after removals began in June, for example, Google restored certain links when publishers argued that the stories in question were in the public interest. Individuals unhappy with the decisions of search engine providers—Google has rejected a little over 40 percent of requests so far—can still appeal to their national data-protection agencies.

Source: Google; 860,894 URLs evaluated as of 9 April 2015
Forgotten But Not Gone: Facebook pages lead the list of links Google was asked to forget.

The logic that the ruling follows is one of the things that makes it remarkable. The ruling skips altogether the mechanics of how search results are created from public third-party records. Instead, the court reasoned that the potential richness and accessibility of search results make them more of a threat to a person’s privacy than the individual records themselves. Since Google (for now) is the main controller of that data in Europe, it has the same responsibility as other companies that control personal data in Europe, the court argued.

Yet some groups have questioned the court-imposed, company-implemented method of protecting the public’s right to know: “We argue that it should be handled by a court, not a company,” says Pam Cowburn, communications director for the Open Rights Group in London.

Wikipedia founder Jimmy Wales, who sat on an advisory council that Google convened about the issue, concurred in his comment on the council’s report and added that “the recommendations to Google contained in this report are deeply flawed due to the law itself being deeply flawed.” Google’s advisory council conducted hearings in seven European cities in late 2014.

Another member of the advisory council, German federal justice minister Sabine Leutheusser-Schnarrenberger, argued that Google’s current interpretation of the ruling does not go far enough. “Since EU residents are able to research globally, the EU is authorized to decide that the search engine has to delete all the links globally,” she wrote in her comment on the report.

And globally is how the idea is spreading. A Japanese court ruled in October 2014 that Google must take down some personal information. Yahoo has already announced that it will extend its existing privacy takedown request system to its Japanese sites. Hong Kong’s privacy commissioner for personal data, Allan Chiang Yam Wang, said late last year at the Asia Pacific Privacy Authorities Forum in Vancouver, B.C., Canada, that national privacy officials from many countries were considering how they might enact their own versions.

The question in all cases is whether unfettered search engine results are the same as free expression. There’s good reason to think not. Governments limit the intrusiveness of credit reports, for example, treating them as business tools instead of a form of free expression. Evgeny Morozov asked in The New Republic, “If we don’t find it troubling to impose barriers on the data hunger of banks and insurance companies [for composing credit reports], why should we make an exception for search engines?”

At a March debate presented by Intelligence2, University of Chicago law professor Eric Posner argued that such information is now easier to find than in the past, when public records such as trial proceedings might have ended up in a courthouse basement—available, but not easily accessible to all. “What the ‘right to be forgotten’ does is it raises the cost for strangers to find out information about you. It doesn’t make it impossible,” he said. By making it inconvenient, but not impossible, for someone to find contested information, the right to be forgotten is effectively undoing some of the gains of the information age: re-creating the effect of paper’s archival limitations with today’s technology.

First published in IEEE Spectrum: [html] [pdf].

Credit: Ian Joughin, CC BY-NC-SA 3.0

Building a Better Glacial Speedometer

Greenland is the land of escaping lakes. In the summer, when soot lands on the ice sheet’s snowy surface and the Sun begins to melt the snow, bright blue lakes form on top of the ice. Just as on land, the water seeks a way down.

Sometimes, instead of carving surface channels, water trickles into the ice sheet through crevasses and vertical shafts called moulins. In the most dramatic cases, a lake can burst through a kilometer-thick ice sheet and rush to the bottom of the glacier in a forceful waterfall. There, under high pressure, water may help the glacier glide a little faster over the rock below.

Just how fast, however, is the subject of an ongoing debate.

More Ice Sheet Lakes, Farther Inland

Geoscientist Kate Briggs of the University of Leeds predicted that such lakes will advance from around 50 to more than 100 kilometers inland on Greenland over the next few decades. She presented evidence for this, based on a study she and her colleagues published earlier this year in Nature Climate Change, during a session at the 2015 European Geosciences Union (EGU) meeting in Vienna, Austria.

If such lakes begin to cover more of the Greenland ice sheet, they could play a growing role in the speed with which it moves, Briggs said.

However, there may be natural brakes built into the system. Physical geographer Jonathan Bamber of the University of Bristol commented during the session that lakes forming on the even thicker ice sheet in the middle of Greenland may not be able to escape all the way to the rock bed below the glacier. The deepest ice is under even more pressure and may resist the hydraulic fracturing that enables surface lakes to break through the ice sheet closer to Greenland’s edge.

Briggs and her colleagues assumed in their model that the water did eventually reach the rock bed and that the additional pressure helped it speed up the ice sheet’s motion. Nonetheless, the net effect of how glacial lakes affect ice flow is “still very much an open question,” Briggs said.

Many Methods, Many Answers

Other scientists installed GPS sensors for short periods deep inside the ice sheet to examine how it moves. They found a caterpillar-like movement in fits and starts, according to research reported earlier this year in the Journal of Geophysical Research: Earth Surface.

It is useful information, but generalizing any local observation about the ice sheet is difficult: Greenland spans 18° of latitude and hosts differing microclimates, and its ice sheet encounters chaotic mountains and sea conditions at its edge. The ice sheet’s interaction with the Earth’s climate is complex, session speakers noted.

Still, tracking its speed is a good place to start, explained ice sheet researcher Twila Moon of the University of Colorado Boulder. “Understanding ice velocity is fundamental to understanding how much ice we’re going to lose,” she told Eos.

During the same EGU session as Briggs’s talk, Moon presented preliminary data from the Landsat 8 series. Some of Moon’s previous work shows that the ice sheet’s velocity changes over the course of the seasons and varies by location. Landsat 8 offers more frequent sampling than previous satellite ice sheet mapping: up to every 16 days. That means she will be able to create maps of the ice sheet’s speed with unprecedented time coverage.

“Her data will add a lot,” said applied mathematician Ian Hewitt of the University of Oxford in the United Kingdom. Hewitt creates computational models of how water, glaciers, and the rock bed interact, as seen in this video:

Perhaps, over the course of the season, Hewitt explained, water carves out channels below the ice sheet, and over time, more of the glacier’s weight settles onto the rock bed between the channels. That might explain why researchers have detected both speedups and slowdowns of the ice sheet after surface lake disappearances. The new density of time sampling will enable him and other modelers to put better limits on their models.

More Data, More Modelling

Briggs and others also examine the ice sheet’s topography using radar from CryoSat-2. She presented some preliminary data on ice sheet thickness during her talk, including more evidence that some of the Greenland ice sheet is thinning in ways not captured by existing regional climate models. Similar to the GPS and optical data, the radar data raise more questions.

It is tempting to argue that the growing density of aerial coverage could mean there is less need for slow, expensive GPS surveys, Moon told Eos. However, some of the finer movement, as shown by the GPS surveys, occurs on time scales too short for satellites to reliably detect.

“We would need subdaily resolution to capture the sudden fits and starts,” Hewitt told Eos. “We won’t get that from satellites.” Instead, it may take a lively combination of more on-the-ground instrumentation, more remote sensing, and more modeling to see into the future of Greenland’s ice sheet and its disappearing lakes.

Hewitt explained, “To really work out what’s going on, we’ll have to integrate subglacial processes into those models and then run them long term.”

First published by Eos: [html] [pdf].