Tuesday, April 21, 2015

Social Media & Storm Tracking: Silly or Worth While

Social Media Tracking Jonah Engler

Storm Chasers are nothing new and provide a valuable meteorological service toward helping scientists better understand large storms, tornados, and the way they form and move. Some have come under fire for the dangerous risks they take and sometimes decidedly unscientific methodology. Twitter has been hailed time and again as a platform for breaking news even before it hits the wires and is picked-up by larger news agencies. It has also come under fire for any number of things. Most important here is when tweets or news items are falsified.

A group of Purdue students decided to track an oncoming storm's development and damage over Twitter, searching for terms like "damage" and "funnel cloud." A professor there posits that storm tracking can be done using social media not unlike the way radar is used. Unlike meteorological equipment though, social media can also humanize the damage storms bring and the areas they affect. The National Weather Service hopes to develop methods of mining social media news for trends that might help detect severe weather and warn others in its path. The caveat being, again, that not all tweets are accurate or even truthful, leading some to suggest the idea is silly.

Another issue facing the concept is that of demographics. Twitter users are overwhelmingly young, urban, and affluent. A farmer in the rural Midwest who has been through a storm or two is not likely to feel the same about extreme weather as young, urban professionals whose snarky tweets can turn a light drizzle into a monsoon. While the algorithm the NWS is working on has a method of weighing the truthfulness of a tweet, it is unlikely to know how to deal with sarcasm or hyperbole.

This algorithm might be overwhelmed if the students tried to pipe-in a larger section of the Web or social media, in general, but Twitter may not be a reliable enough site for serious study. Furthermore, there is no demonstrable way to show that mining tweets for storm information can help in any way to predict its path. It also seems a poor choice of platforms for issuing weather alerts. As it stands, it may provide valuable information but only after the fact.

Cellphones are ubiquitous and since inclement weather hinders reception, being able to tweet from the eye of a terrible storm might be all someone has one day. It might prove very important.

Jonah Engler is an entrepreneur and a finance expert from New York City.

No comments:

Post a Comment