Iceland Volcano Dust Cloud Stops Air Traffic

Iceland Volcano Dust Cloud Stops Air Traffic

While there was no disaster in terms of loss of life, this event captured worldwide attention, forcefully, for several days. Under the headline above, a lead article from a South Pacific newspaper on April 20 says: "Airlines have lost over $1.7 billion in revenues since a volcano in southern Iceland erupted, causing cancellations of flights bound for Europe, the International Air Transport Association (IATA) said in a statement. IATA noted that from April 17 to 19, when travel disruptions were greatest, airlines lost $400 million per day."

From Telegraph.co.uk, a few points in the timeline:

2am Wednesday, April 14 -- After weeks of reduced activity, Eyjafjallajokull volcano in Iceland erupted beneath its glacier ice cap, which by midmorning was spewing out a substantial stream of steam and ash. By midday the surrounding rivers had risen by 10 ft and a plume of ash had risen through the air to four miles in the sky.

9am Thursday, April 15 -- NATS announces the whole of British airspace will shutdown between midday and 6pm, an unprecedented move. After the September 11 terror attacks in New York, flights continued on non-transatlantic routes. By the evening Ryanair would announce it was cancelling all flights for the next four days – a decision eventually forced upon all other airlines.

8.40pm Sunday, April 18 -- BA safely lands a test flight in Cardiff, with chief executive Willie Walsh on board, flying three hours out to the Atlantic. The next day he declared the "current blanket restrictions on airspace are unnecessary" after tests came back saying its engines were unaffected

7am, Tuesday, April 20 -- British airspace due to reopen, starting in Scotland, NATS has announced. British Airways said it would start flying out of and into British airports from 7pm on Tuesday.

We could not include this event in the formal series of hypothesis tests because of its length and because the beginning of the event literally overlaps with two others, previously set. However, there is no question that it meets many of the criteria we use for selecting formal events. It was deeply engaging for very large numbers of people; it was the focus of most news sources for several days; it directly affected large numbers of travelers.

We set an exploratory analysis for the Volcano Dust Cloud using the same procedures as for formal events. The event extends from April 14 to April 19, six full UTC days. The first day overlaps with two formal events, the Kumbh Mela in India, and the big earthquake in Qinghai China, the Tibetan plateau. Since this is an exploration, statistical interpretation is not warranted, but the graph of accumulating deviations shows a persistent excess of correlation in the network, especially on the 14th, and 16th through 18th.

It is important to keep in mind that we have only a tiny statistical effect, so that it is always hard to distinguish signal from noise. This means that every "success" might be largely driven by chance, and every "null" might include a real signal overwhelmed by noise. In the long run, a real effect can be identified only by patiently accumulating replications of similar analyses.