According to a document released by the Department of Homeland Security and posted to Scribd by Huffington Post reporter Andrea Stone, words and phrases ranging from "nuclear" to "ice" are used as search terms by DHS to detect potentially dangerous incidents, such as acts of terrorism, extreme weather, traffic accidents, and hazardous spills. The terms come from a 2011 binder that guides DHS analysts working on the agency's "Media Monitoring Capability Mission" (MMC), DHS's effort to cull clues from open sources -- such as Twitter, media reports, and news stories -- to develop what it calls "situational awareness."
The binder instructs analysts not to include "Personally Indentifiable Information" (PII) -- information that could point to a specific individual source -- in their incident reports: "Before sending out ANY reports ... analysts must ensure that if there is any PII included in a media article, that information must be removed, due to privacy issues!" Exceptions can be made for "extremis situations," which are defined as "when there is an imminent threat of loss of life, serious bodily harm, or damage/destruction to critical facilities." Such PII disclosures require approval from DHS higher-ups.
The binder says that the MMC Mission is in part intended to mitigate reliance on local news reports, which, it says, may be "potentially sensationalizing an incident." It's an interesting footnote to the concerns that Twitter and other social media hasten the spread of misinformation and rumors. Maybe so, but the on-the-ground chatter about an unfolding event can also counter the ways stories are sometimes shaped in more traditional news sources.
Here's the complete list: