Meteorologist Mike Smith, CEO of WeatherData and author of “Warnings: The True Story of How Science Tamed the Weather” believes the power outages were indeed the primary cause of the high death toll because “[tornado] warnings were not received or not effectively received.”
He prepared an analysis with his rationale and provided it to the Capital Weather Gang. I am re-printing some key excerpts of it with his permission [bold text indicates my added emphasis]:
Since the tornado warning program began in 1957, death tolls – even with a growing population – have been steadily decreasing. As an example, just five days prior to the April 27 tornado outbreak in the South, the Good Friday Tornado in St. Louis reached rare F-4 intensity (upper 2% of all tornadoes) and struck a densely populated area yet there were no deaths and no serious injuries. This was especially remarkable given that historically more people have been killed by tornadoes in the St. Louis metropolitan area than any other.
Even with the large tornado outbreak that occurred in the South and Carolinas on Friday and Saturday April 15-16, 2011, the death toll was about one-tenth of April 27th’s and 86% of those (31 of 36) were associated with mobile homes, known to be especially vulnerable. So, the current death toll of 236 in Alabama and the total death toll of 337 is simply stunning in the 21st Century given the state-of-the-art in tornado warning science and warning communications technology
. . .
[For the tornado outbreak on April 27 in Alabama] when the morning thunderstorms had departed, media and other reports indicate that at least 262,000 electric “customers” (individual homes and businesses) were without power. The electric utility industry assumes 3-4 people per “customer” which translates to 786,000 to 1,048,000 people.
A second line of thunderstorms with high winds moved across far northern Alabama at midday which caused some additional power failures and hampered the efforts to repair the morning outages.
While less certain than the morning power outage reports, it appears that “thousands” of additional customers lost power in southern Tennessee and northern Alabama due to the midday storms.
. . .
Without electricity, televisions and the internet do not work. Given enough time, the backup batteries at cellular telephone towers exhaust themselves of power as do the batteries on individual cell phones. So, the usual systems for receiving the warnings would be, at best, disrupted. In some areas, especially in rural regions without power, the warnings may not have been available through any routine source due to local radio stations signing off at sundown.
If the warnings were not received, it would account for the “pre-warning era” level of fatalities.
It is impossible, at this time, to know exactly the role the premature (i.e., before the thunderstorms bearing the tornadoes) power failures played in this event.
Smith closes by “urging” the National Weather Service to examine the roll of power failures in the death toll. He told me he agrees the abundance of mobile homes and insufficient building stock played a role in the fatalities, but said if the people living in these structures had 24 minutes of lead time (the average lead time for the storms, according to the National Weather Service), it would have afforded them with the opportunity to move to a sounder structure.