Assessing the Accuracy of Severe Weather Warning Systems

There are two separate but related issues to consider in assessing severe weather alert systems. The first is the accuracy of the observations and forecasting of the weather, and the second is the effectiveness of disseminating that information to the authorities and the public so that safety measures can be taken. In terms of accuracy, the system is extremely effective and reliable. But in terms of how well that reliability is applied in actually getting the news to people who will be affected by the weather, the system suffers from a number of inconsistencies.

In the U.S., weather monitoring relies on a number of observation sources, the centerpiece of which is the NEXRAD radar network. NEXRAD stands for “Next-Generation Radar,” and is a network of 158 radar sites covering all but a very few small areas of the 48 contiguous states, with additional stations located in Alaska, Hawaii, Puerto Rico, the Azores, Guam, Okinawa, and South Korea. The type of radar used is known as the WSR-88D, or Weather Service Radar 1988 Doppler; installation of the network began in 1990 with the first station at the National Weather Service’s Radar Operations Center in Norman, Oklahoma and was completed in 1997.

This radar produces the familiar multi-colored images of weather formations seen on TV newscasts and online weather maps, and is capable of a wide variety of scanning programs. The benefit of the WSR-88D is that it can detect not only rain, but also differences in wind velocity and air pressures, depending on the detection algorithms being used. And it can do it quickly, having a scan repeat time of four to ten minutes. This allows forecasters to observe weather conditions in nearly real-time, and when this system is used in conjunction with field observers, traditional instrument stations, and other systems such as lightning detection networks, the information gathered is nearly foolproof for short-range forecasting. And the system is constantly being improved, with exotic-sounding new technologies such as polarimetric and phased-array radar systems now used by the military planned for introduction into the weather monitoring system over the next decade.

The trouble with weather observation and forecasting is that if the message can not be effectively delivered to those in harm’s way in a timely manner, the best equipment in the world is virtually useless. In 2005, as Hurricane Katrina approached New Orleans and the Gulf Coast, the National Weather Service issued a hurricane warning for the area a full 32 hours before the storm made landfall in southern Louisiana. Twelve hours later, they followed it up with another strongly-worded advisory that read in part, “Most of the area will be uninhabitable for weeks … perhaps longer. … ONCE TROPICAL STORM AND HURRICANE FORCE WINDS ONSET…DO NOT VENTURE OUTSIDE!” And yet despite the rather blunt assessment, the emergency preparedness and response to the Katrina disaster, resulting as it did in over 1,000 deaths, is generally considered to have been an utter failure. State and local officials are responsible for passing the alert to the people in their areas, and while they can count on assistance from the Federal government after the fact, they serve as the front-line defense against the onslaught of destructive weather. How well they accomplish that task is determined not only by their own local resources and commitment, but also by how well the people respond to the warnings they are given. All too often, the victims of violent weather turn out to be the stubborn few who value their property above their own safety and think they can “ride it out,” or the adventurers who deliberately put themselves in harm’s way to get some really great video to post on YouTube. No system is infallible, and sometimes people get hurt despite all the best intentions and efforts of those whose job it is to maintain the system. When that happens, it’s a tragedy; when it happens because someone willfully ignored the warnings, that’s just stupidity being appropriately rewarded.

National Weather Service Radar Operations Center:

Stumpf, Gregory J., Travis M. Smith, and Claire Thomas,”The National Severe Storms Laboratory’s Contribution to Severe Weather Warning Improvement: Multiple-Sensor Severe Weather Applications,” Atmospheric Research, Volume 67-68, July-September 2003, Pages 657-669.

National Weather Service Homepage: