Skip to content Skip to main navigation Report an accessibility issue

Social Volume Analysis of Natural Disasters

Radar imagery shows Hurricane Florence making landfall.

Post written by Cassandra Huang

We heard many natural disasters recently all over the world. In the middle of this month, Hurricane Florence, a long-lived powerful storm, caused huge damage in the Carolinas. Last week, Typhoon Mangkhut, the strongest storm in the past two decades in southeast Asia, claimed more than thirty lives in China and the Philippines and nearly crushed Hong Kong. This week, Super Typhoon Trami in the western Pacific Ocean, one of the possible strongest typhoon of this year, is approaching Taiwan or southwestern Japan’s Ryukyu Islands.

In the past few years, social media analysis has been applied to yield insights from people’s real-time perceptions of a natural disaster. For example, Mandel, Culotta, Boulahanis, Stark, Lewis, & Rodrigue (2012, June) examined over 650,000 Twitter posts in August 2011 when Hurricane Irene affected the US. They had an interesting research result that female were more likely to have tweets that concerned the disaster than males.

How can we do this kind of analysis to examine social volume of a disaster? Social Studio can definitely help. As you can see below, Social Studio can analyze the change of the volume on social media platforms such as Twitter, and tell us when people post. For example, there were 1.1 million posts on Twitter related to Hurricane Florence during September 14 to 15, and the peak of posts was 10 am on September 14, which was around one hour before landfall.

Timeline of Twitter posts during Hurricane Florence.

And you can also examine what people mentioned the most in their posts, just like the example shows below: