Here an article I'd like you to read. The money quote:
A personal note: I've met Nick and he is a dedicated public servant who wants to do the right thing. But, in this case, was sounding the sirens the right thing?
Here is the radar with the National Weather Service's tornado warning plotted in red. You'll notice only a small part (southwest) of Johnson Co. was in the tornado warning.
Sounding the sirens in Edgerton, Gardner, and Spring Hill was appropriate. I'd also say to sound them in Olathe because they are right on the edge of the warning. But the rest of the county? No. It was a false alarm -- the tornado dissipated before it reached Johnson Co.
Earlier in the evening, the sirens went off in Lawrence even thought it was never in a tornado warning (see below):
I'm sure that emergency manager thought he or she was doing the right thing and erring on the side of safety, as well.
Here's the problem: We are training people to ignore the sirens because of all of these false activations.
The National Weather Service builds a "margin of safety" into the (relatively new, since 2005) "path-based" warnings. The emergency manager (who is not a meteorologist) building an additional margin of safety over and above the NWS's leads to far too many siren activations.
When the sirens were pressed into tornado warning service in the 1950's and 60's, the "countywide" and "err on the side of safety" philosophy made sense because meteorology was not very good at tornado warnings.
Now, meteorologists can actually predict the path of tornadoes. I believe it is time for selective siren activation to become the norm.