Beyond Belief
09-02-2005, 06:56 PM
The tragedy currently unfolding in New Orleans is in many ways unprecedented in U.S. history, and it is tempting to think that the misery we are witnessing could have been avoided. I would like to suggest that some level of misery and loss of life was unavoidable. With all of the rhetoric in recent days regarding the possible role of global warming in hurricane activity, it is useful to examine weather disasters in general, and Katrina in particular, from both historical and practical points of view.
Forecast Accuracy and Warnings
Everyone knows that weather forecasts are not totally accurate. For potentially destructive and life-threatening events such as tornadoes, hurricanes, and floods, forecasters necessarily err on the side of caution. This leads to over-warning, which in turn results in some level of complacency on the part of the public. While over-warning leads to a high "probability of detection" (very few events go without warnings), it is at the expense of increased false alarms. But there really is no other acceptable choice. The only alternative would be to issue fewer warnings. But given the imprecision of hurricane forecasts, this would be at the cost of numerous events for which there were no warnings. Many, possibly most, hurricanes that hit land would either have no warnings, or would have insufficient lead time for evacuations and property protection to take place. This would be totally unacceptable to the public. Thus, we are left with the unavoidable situation where some portion of the people will not heed warnings - for example, I personally ignore most tornado warnings -- and so people will die.
Hurricane intensity and track forecasts for Hurricane Katrina were, from a historical perspective, pretty darn accurate. Early forecasts had the hurricane tracking farther east in the Florida panhandle. But as of 11 p.m. Saturday night (48 hours before high winds started reaching the coast of Louisiana) the National Hurricane Center (NHC) was forecasting an "intense hurricane". The forecast track issued at that time was almost dead-center on the eventual landfall location. Katrina ended up intensifying and moving more rapidly than normal, leading to less lead time than would have been desired for the warned areas.
Nevertheless, warnings of a "catastrophic event" were made in time for virtually all of the people who were willing and able to leave New Orleans and coastal areas to do so. Most people did indeed leave the warned areas -- but not all of them. NHC makes it a special point in the case of especially broad hurricanes such as Katrina to tell people to not focus on the exact forecast track of the eye since such a broad area will be impacted anyway.
continued here:
http://www.techcentralstation.com/090205A.html
Forecast Accuracy and Warnings
Everyone knows that weather forecasts are not totally accurate. For potentially destructive and life-threatening events such as tornadoes, hurricanes, and floods, forecasters necessarily err on the side of caution. This leads to over-warning, which in turn results in some level of complacency on the part of the public. While over-warning leads to a high "probability of detection" (very few events go without warnings), it is at the expense of increased false alarms. But there really is no other acceptable choice. The only alternative would be to issue fewer warnings. But given the imprecision of hurricane forecasts, this would be at the cost of numerous events for which there were no warnings. Many, possibly most, hurricanes that hit land would either have no warnings, or would have insufficient lead time for evacuations and property protection to take place. This would be totally unacceptable to the public. Thus, we are left with the unavoidable situation where some portion of the people will not heed warnings - for example, I personally ignore most tornado warnings -- and so people will die.
Hurricane intensity and track forecasts for Hurricane Katrina were, from a historical perspective, pretty darn accurate. Early forecasts had the hurricane tracking farther east in the Florida panhandle. But as of 11 p.m. Saturday night (48 hours before high winds started reaching the coast of Louisiana) the National Hurricane Center (NHC) was forecasting an "intense hurricane". The forecast track issued at that time was almost dead-center on the eventual landfall location. Katrina ended up intensifying and moving more rapidly than normal, leading to less lead time than would have been desired for the warned areas.
Nevertheless, warnings of a "catastrophic event" were made in time for virtually all of the people who were willing and able to leave New Orleans and coastal areas to do so. Most people did indeed leave the warned areas -- but not all of them. NHC makes it a special point in the case of especially broad hurricanes such as Katrina to tell people to not focus on the exact forecast track of the eye since such a broad area will be impacted anyway.
continued here:
http://www.techcentralstation.com/090205A.html