Weather ... or not
Well, there goes another hurricane season.
Forecasters who got it wrong for a second year in a row now have--as Ricky Ricardo would say--some 'splainin' to do. Or maybe not. Priesthoods do hold their work above accountability.
Back in the late spring, weather experts tried their best to scare the hell out of us again with forecasts of a summer and fall chock full of storms and hurricanes. As if the prospect of having your roof ripped off or your home flooded wasn't scary enough already. Having flubbed their pre-season predictions in 2006, they didn't quite get the same breathless reaction from the press in 2007, but there were a fair number of articles in the mainstream media repeating their warnings. I'll be interested to see how they play it next year. I can't believe anyone is going to take them seriously if they decide to cry wolf a third time. At this rate, I will feel more anxious if they predict a mild hurricane season for 2008.
So what's going on? As I see it, there are three forces at work that have rendered pre-season tropical-storm estimates worthless, and none of them are meteorological forces.
First of all, the political maelstrom surrounding the recently popularized theory of anthropogenic global warming has warped the already uncertain science behind large-scale weather prediction. One fine example of global warming irrationality was the attempt by some scientists and media outlets to pin the 2004 tsunami on American SUVs. (Connecting the dots from carbon dioxide emissions to melting permafrost to tectonic plates to earthquakes to massive waves is quite a stretch, but it's sadly conceivable in the new world of weather pseudo-science.) The global warming debate (no, the debate is not over) may be helping dysfunctional third-world governments in their ceaseless search for scapegoats and handouts, but it's not helping science.
Second is the ramped-up media coverage of everything on earth except genocide in Darfur and the gradual Islamist takeover of Europe. Incessant shrieking from the mainstream media about each supposed looming disaster keeps our attention firmly focused right where they want it.
The third force at work is the most complex, and in my opinion the one most deserving of some exploration. The criteria for identifying and analyzing tropical storms and hurricanes have grown dangerously loose and unscientific, allowing for weather prediction driven by agenda rather than by reason. The media have gleefully joined forces with meteorologists in ascribing significance to tallies--how many storms we count--and comparing those tallies to lists from bygone, politically calmer times.
The problem here is obvious: We are no longer measuring how many storms occur, we are measuring how many storms we can count. The only tropical meteorological phenomenon that has shown an anomalous change in recent years is that of meteorologists' willingness to pad the storm tallies they feed to the media. Scientists now draw sweeping historical conclusions from comparison of recent data with that of years past while quietly neglecting to compare the new data-gathering methods with the old. Granted, there is some noise about this problem in the insider-world of weather wonks, but no one seems inclined to bring this dispute to the attention of the public.
This third force can be broken down into three statistical flaws. The first is the most obvious: we didn't have satellites scanning the Atlantic prior to 1960. The 2007 hurricane season's tally includes four named storms (Ingrid, Jerry, Karen, and Melissa) that never threatened land and might have gone unremarked or even unnoticed without satellite imagery. Jerry is the clearest example of this form of tally padding.
Jerry's storm track. That's Newfoundland off to the storm's left. About 700 miles off.
In the 1930s, out of 98 storms, only 12 that never touched land were identified and included in the list. In 2007, out of 14 identified storms, four never touched land. That's a jump from 12% to 28%. Which is the more rational explanation--that a greater percentage of storms are somehow avoiding landfall, or that we are counting more storms?
Which leads us to the second statistical flaw: the criteria for including storms in the annual tally have been gradually modified in ways that increase the number of storms reported and inflate the public's perception of storm activity by naming storms. In 2002, the NOAA altered its policy and began naming not only tropical storms and hurricanes, but sub-tropical cyclones as well. Sub-tropical cyclones are certainly worth noting, since they can develop into tropical storms, but naming them is deceptive, given how we are repeated told that the number of named storms in a season is significant. Since the policy change, NOAA has put three named sub-tropical cyclones on their lists. Two of them (Nicole in 2004 and Andrea in 2007) never became tropical storms, yet their names will forever unbalance the statistics of these years against earlier data.
Although weather experts may see the distinction and not include named sub-tropical storms in their scientific assessments, it is difficult for journalists and activists to see the difference, even when they would like to be honest. The NOAA's own website contains contradictory information on the nature of the first named sub-tropical storm, Nicole. The main webpage on the 2004 season identifies Nicole as sub-tropical in its text, yet the linked map of Nicole's track across a stretch of the open Atlantic identifies her as a tropical storm.
It is also important to note that even after the practice of naming storms was formalized in the 1940s, many storms and hurricanes were judged undeserving and left with just numbers to identify them in the records. The records from the relatively quiet 1960s include five unnamed tropical storms of equal or greater strength and effect than 2007's Jerry. In 1969, two full-blown hurricanes even had to settle for numbers instead of names. Whether or not we name storms wouldn't matter so much if we were not being told that there is some meteorological significance in using up the whole alphabet in a single year.
The third statistical flaw lies in the way the various data have been parsed to draw specious conclusions--in particular, that we are seeing an unprecedented increase in tropical storm and hurricane activity. We are told that early storms are an ominous sign of things to come, that the power of storms like Katrina is unprecedented, and that our unfortunate generation is witness to an unusual rise in hurricane frequency and strength (rather than just another crest in a multi-year activity cycle).
These fallacies are easily dispelled by spending a little time examining the records (even without consideration for how the old data is not comparable to newer data). Take the 1950s. Hurricane Able arrived on the 15th of May in 1951, well ahead of the season, and reached category 3 strength earlier than any hurricane on record. The following year saw the first tropical storm arrive in February. That's way ahead of 2007's Andrea (May 6, and bear in mind that Andrea never became a true tropical storm). Alice arrived on May 25, 1953. Perhaps we should just say that the season begins in May. Twelve tropical storms struck in 1955, nine of them becoming hurricanes and six of them reaching category 3 or greater. Tropical storm Arlene arrived three days ahead of the season in 1959.
And since hurricanes do not give a damn what decade we think it is, let's look past the end of the 1950s and see if the storms changed their ways. Eleven storms were recorded in 1961, of which eight became hurricanes, six of those category 3 or greater. From my vantage point in the Caribbean, I'm starting to feel lucky I wasn't here for what the media want us to believe were the good old days.
But the weather priesthood doesn't see it that way. Here's Joe Bastardi from AccuWeather on what my reaction should be to his colleagues' miserable batting average over the past two seasons.
Has anyone said, "Ha, ha, ha"? This is the weakest straw-man tactic I've seen in a while. No one in his or her right mind would gloat over being spared the tragedy that we've so recently witnessed in New Orleans. Being pleased that we had a mild storm season does not mean we are pleased that our scientists have proven themselves fallible two years in a row. And don't respond to our justifiable concern with a patronizing repetition of the same old dire threats. We never asked you to tell us how bad each coming season would be. The meteorological establishment came up with that plan on its own. Perhaps it's time to abandon it.
Meteorologists should admit that pre-season forecasting is simply beyond their current abilities, stop playing into the hands of our hysterical mainstream media, and focus on the useful and much-appreciated business of identifying and tracking the storms we care about--the ones that exist.
Forecasters who got it wrong for a second year in a row now have--as Ricky Ricardo would say--some 'splainin' to do. Or maybe not. Priesthoods do hold their work above accountability.
Back in the late spring, weather experts tried their best to scare the hell out of us again with forecasts of a summer and fall chock full of storms and hurricanes. As if the prospect of having your roof ripped off or your home flooded wasn't scary enough already. Having flubbed their pre-season predictions in 2006, they didn't quite get the same breathless reaction from the press in 2007, but there were a fair number of articles in the mainstream media repeating their warnings. I'll be interested to see how they play it next year. I can't believe anyone is going to take them seriously if they decide to cry wolf a third time. At this rate, I will feel more anxious if they predict a mild hurricane season for 2008.
So what's going on? As I see it, there are three forces at work that have rendered pre-season tropical-storm estimates worthless, and none of them are meteorological forces.
First of all, the political maelstrom surrounding the recently popularized theory of anthropogenic global warming has warped the already uncertain science behind large-scale weather prediction. One fine example of global warming irrationality was the attempt by some scientists and media outlets to pin the 2004 tsunami on American SUVs. (Connecting the dots from carbon dioxide emissions to melting permafrost to tectonic plates to earthquakes to massive waves is quite a stretch, but it's sadly conceivable in the new world of weather pseudo-science.) The global warming debate (no, the debate is not over) may be helping dysfunctional third-world governments in their ceaseless search for scapegoats and handouts, but it's not helping science.
Second is the ramped-up media coverage of everything on earth except genocide in Darfur and the gradual Islamist takeover of Europe. Incessant shrieking from the mainstream media about each supposed looming disaster keeps our attention firmly focused right where they want it.
The third force at work is the most complex, and in my opinion the one most deserving of some exploration. The criteria for identifying and analyzing tropical storms and hurricanes have grown dangerously loose and unscientific, allowing for weather prediction driven by agenda rather than by reason. The media have gleefully joined forces with meteorologists in ascribing significance to tallies--how many storms we count--and comparing those tallies to lists from bygone, politically calmer times.
The problem here is obvious: We are no longer measuring how many storms occur, we are measuring how many storms we can count. The only tropical meteorological phenomenon that has shown an anomalous change in recent years is that of meteorologists' willingness to pad the storm tallies they feed to the media. Scientists now draw sweeping historical conclusions from comparison of recent data with that of years past while quietly neglecting to compare the new data-gathering methods with the old. Granted, there is some noise about this problem in the insider-world of weather wonks, but no one seems inclined to bring this dispute to the attention of the public.
This third force can be broken down into three statistical flaws. The first is the most obvious: we didn't have satellites scanning the Atlantic prior to 1960. The 2007 hurricane season's tally includes four named storms (Ingrid, Jerry, Karen, and Melissa) that never threatened land and might have gone unremarked or even unnoticed without satellite imagery. Jerry is the clearest example of this form of tally padding.
Jerry's storm track. That's Newfoundland off to the storm's left. About 700 miles off.
In the 1930s, out of 98 storms, only 12 that never touched land were identified and included in the list. In 2007, out of 14 identified storms, four never touched land. That's a jump from 12% to 28%. Which is the more rational explanation--that a greater percentage of storms are somehow avoiding landfall, or that we are counting more storms?
Which leads us to the second statistical flaw: the criteria for including storms in the annual tally have been gradually modified in ways that increase the number of storms reported and inflate the public's perception of storm activity by naming storms. In 2002, the NOAA altered its policy and began naming not only tropical storms and hurricanes, but sub-tropical cyclones as well. Sub-tropical cyclones are certainly worth noting, since they can develop into tropical storms, but naming them is deceptive, given how we are repeated told that the number of named storms in a season is significant. Since the policy change, NOAA has put three named sub-tropical cyclones on their lists. Two of them (Nicole in 2004 and Andrea in 2007) never became tropical storms, yet their names will forever unbalance the statistics of these years against earlier data.
Although weather experts may see the distinction and not include named sub-tropical storms in their scientific assessments, it is difficult for journalists and activists to see the difference, even when they would like to be honest. The NOAA's own website contains contradictory information on the nature of the first named sub-tropical storm, Nicole. The main webpage on the 2004 season identifies Nicole as sub-tropical in its text, yet the linked map of Nicole's track across a stretch of the open Atlantic identifies her as a tropical storm.
It is also important to note that even after the practice of naming storms was formalized in the 1940s, many storms and hurricanes were judged undeserving and left with just numbers to identify them in the records. The records from the relatively quiet 1960s include five unnamed tropical storms of equal or greater strength and effect than 2007's Jerry. In 1969, two full-blown hurricanes even had to settle for numbers instead of names. Whether or not we name storms wouldn't matter so much if we were not being told that there is some meteorological significance in using up the whole alphabet in a single year.
The third statistical flaw lies in the way the various data have been parsed to draw specious conclusions--in particular, that we are seeing an unprecedented increase in tropical storm and hurricane activity. We are told that early storms are an ominous sign of things to come, that the power of storms like Katrina is unprecedented, and that our unfortunate generation is witness to an unusual rise in hurricane frequency and strength (rather than just another crest in a multi-year activity cycle).
These fallacies are easily dispelled by spending a little time examining the records (even without consideration for how the old data is not comparable to newer data). Take the 1950s. Hurricane Able arrived on the 15th of May in 1951, well ahead of the season, and reached category 3 strength earlier than any hurricane on record. The following year saw the first tropical storm arrive in February. That's way ahead of 2007's Andrea (May 6, and bear in mind that Andrea never became a true tropical storm). Alice arrived on May 25, 1953. Perhaps we should just say that the season begins in May. Twelve tropical storms struck in 1955, nine of them becoming hurricanes and six of them reaching category 3 or greater. Tropical storm Arlene arrived three days ahead of the season in 1959.
And since hurricanes do not give a damn what decade we think it is, let's look past the end of the 1950s and see if the storms changed their ways. Eleven storms were recorded in 1961, of which eight became hurricanes, six of those category 3 or greater. From my vantage point in the Caribbean, I'm starting to feel lucky I wasn't here for what the media want us to believe were the good old days.
But the weather priesthood doesn't see it that way. Here's Joe Bastardi from AccuWeather on what my reaction should be to his colleagues' miserable batting average over the past two seasons.
"Overall, the nation got off very easy this year and last year ... We are in a time until about 2020 that hurricane threats will be more frequent and more intense on our coastlines. So instead of saying, Ha, ha, ha, there's nothing going on, people should be thankful that there's not as much going on."
Has anyone said, "Ha, ha, ha"? This is the weakest straw-man tactic I've seen in a while. No one in his or her right mind would gloat over being spared the tragedy that we've so recently witnessed in New Orleans. Being pleased that we had a mild storm season does not mean we are pleased that our scientists have proven themselves fallible two years in a row. And don't respond to our justifiable concern with a patronizing repetition of the same old dire threats. We never asked you to tell us how bad each coming season would be. The meteorological establishment came up with that plan on its own. Perhaps it's time to abandon it.
Meteorologists should admit that pre-season forecasting is simply beyond their current abilities, stop playing into the hands of our hysterical mainstream media, and focus on the useful and much-appreciated business of identifying and tracking the storms we care about--the ones that exist.
posted by tompain at 9:40 PM
4 comments
