Extrapolating From A Single Data Point: Climate and Sandy

I have a new article up at Forbes on how crazy it is to extrapolate conclusions about the speed and direction of climate change from a single data point.

Positing a trend from a single database without any supporting historical information has become a common media practice in discussing climate.  As I wrote several months ago, the media did the same thing with the hot summer, arguing frequently that this recent hot dry summer proved a trend for extreme temperatures, drought, and forest fires.  In fact, none of these are the case — this summer was not unprecedented on any of these dimensions and no upward trend is detectable in long-term drought or fire data.   Despite a pretty clear history of warming over the last century, it is even hard to establish any trend in high temperature extremes  (in large part because much of the warming has been in warmer night-time lows rather than in daytime highs).  See here for the data.

As I said in that earlier article, when the media posits a trend, demand a trendline, not just a single data point.

To this end, I try to bring so actual trend data to the trend discussion.

  • GoneWithTheWind

    This is the worst storm in this area since 1938. Which I guess means climate change was worse in 1938 then it is today. Or another way to put this is this is the first time in history we ever had a hurricane come up the east coast that met a artic cold front coming from the Northwest. Well... actually not the first time in fact it isn't uncommon, but this is the firt time in recent history it was so bad. Well except that if you look at the East coast and the sandy hooks and barrier islands they were in fact caused by much great storms in much greater numbers. The entire East coast was shaped by hurricanes and Nor'easters. So this is neither the worst storm in our brief recorded history nor the worst storm based on prehistory data. What is different about this storm is how long it has been between these storms compared to historical data. We have been in a 100-150 year mild human friendly climate phase. A cyclical phase. The previous cycle of global warming, the medeval warming, was in fact much warmer and encouraged the Vikings to colonize Greenland. It is a foregone conclusion that a global cooling will follow this global warming cycle. In fact there is some evidence that the next global cooling cycle may have already begun. We can expect more and worse storms when that happens. New York and New Jersey would be wise to take actions to mitigate this. I find it incredible that in New York City you can find an entrance to the subway that is just a few feet above sea level and it has no way to seal the entrance. Who designed that? Or major tunnels under rivers and bays with no way to both seal them from surface water and pump them out effectively. Again the engineers were asleep or incompetent. Then there are those hundreds perhaps thousands of homes built on the sand that get washed away in every storm. How can that be approved by the city permitting authority??

  • SamWah

    Where I worked 30 years ago. we said "one data point IS a trend". Which means only that I can remember some stuff. Media people can't seem to do that, and won't look it up.

  • Dale

    GoneWithThe Wind:
    "Then there are those hundreds perhaps thousands of homes built on the sand that get
    washed away in every storm. How can that be approved by the city permitting
    authority??"

    You wouldn’t have to worry about the city permitting authority if the Government
    would stay out the home loan business, and the flood insurance business. If people had to pay the real - cost to risk –
    of their flood insurance, homes would stop being built in flood areas.

  • http://www.facebook.com/profile.php?id=1791540090 Bruce Hall

    "Climate Change" is a catch-all phrase that means "any severe weather occurrence is caused by human activity... especially within industrialized areas." It is comparable to attributions of witchcraft.

  • JW

    Never let a crisis go to waste.

  • Rocky

    Rest assured the AGW cult members will be out in force preaching their message of doom and gloom and Sandy is their "I told ya' so". They have been predicting such storms for years but one of this magnitude hasn't materialized until now. At some point in time a storm of this type was bound to hit since this patter is not unusual or unprecedented on the East coast. Storms of similar pattern were quite typical during the 50's and the only aberration was the severity of this storm. Time will tell if the general population will get sucked back into the cult.

  • Seb Tombs

    For presumed rational humans you demonstrate an amazing inability to think: when 100-year storms happen every year, when polar ice caps & glaciers everywhere are melting & overpopulation has led to a huge increase in hydrocarbon consumption, you can't rationally ignore the correlation. But at this point it doesn't matter, since it's probably to late to reverse the trend (thanks to geniuses like you who ignore facts) & after a few wars & famines, human population will decrease back to what the planet can bear.

  • sabre_springs_mark

    100 years storms happen every year - proof please. Before this year a hurricane has not hit the USA in 7 years. The longest lull in recorded history.

  • sabre_springs_mark

    The storm wasn't even all that severe compared to the 1950's when accounting for inflation and changes in living patterns this storm is the 17th most destructive in US history.

    I think the bigger issue is the government subsidizing flood insurance making it much easier for folks to build where they shouldn't

  • sabre_springs_mark

    Exactly - I don't even think the government permit process is the issue. It is the financing of homes in inappropriate areas, as you said, and the subsidized flood insurance, encouraging more homes in inappropriate areas, as you said.

  • sabre_springs_mark

    I am guessing if they really wanted to do something to save Manhattan, it would be very easy to install a flood gate at the mouth of the Hudson. It isn't that wide. Of course, it wouldn't be maintained, and so would fail in a storm, and the eco - freaks would prevent its building for some reason or other.

  • Broccoli

    Seb Tombs obviously doesn't understand the details of 100 year events. There is no such thing as a 100 year storm. There is 100 year wind speed, or 100 year storm surge, or 100 year rainfall in a 24 hour period, or 100 year drought over 3 months, etc. etc. for a specific location on the map. 100 year storm surge doesn't mean once in 100 years on average a storm will have that level of storm surge on the entire east coast. It means for longitude X and latitude Yon average there will be a 100 year event at that exact location.
    This is why 100 year events happen pretty much continuously if you count the whole US, and for your town a 100 year event is likely to happen every few years.

  • sabre_springs_mark

    I was hoping Seb would figure this out for himself. With news today, able to instantly gather info from the whole world in seconds we hear about 100 year events from all over the world, and it scares us, because we don't realize, the only reason we didn't hear about them in the 1970's is because the technology to instantly spread news wasn't there, so the newscasters had to only broadcast what was important locally.

  • seb tombs

    So you think last year's hurricane Irene, that devastated Vermont, among other places, was just a little rainstorm?
    Keep looking for excuses to deny facts, & nit-pick away ... no hurry...