Temperature Trends and the Farmer Error

My dad grew up in farm country in Iowa.  He told me stories of the early days of commodity futures when a number of farmers lost a lot of money betting the wrong way.  The error they made is that they would look at their local weather and assume everyone was experiencing the same.  For example, some guy in Iowa would be experiencing a drought and facing a poor corn crop, and would buy corn futures assuming the crop would be bad everywhere.  Unfortunately, this was often not the case.

A few climate sites have monthly contests to predict the next month's average global temperature anomaly.   Apparently, everyone really missed in the January betting.  Since most of the participants were American, they assumed that really cold weather in the US would translate to falling global temperatures.  They were wrong.  The global temperature anomaly in January actually rose a bit.

This is a variation of the same effect I often point out in the opposite direction -- that heat waves in even seemingly large areas do not necessarily mean anything for global temperatures.  The US is only about 2% of the global surface area (land and ocean) and since the cold spell was in the eastern half of the US, it therefore affected perhaps 1% of the globe.  And remember, on average, some area representing 1% of the globe should constantly be seeing a 100-year high or low for that particular day.  It's just how averages work.

No particularly point here, except to emphasize just how facile it is to try to draw conclusions about global temperature trends from regional weather events.

  • mikehaseler

    "And remember, on average, some area representing 1% of the globe should
    constantly be seeing a 100-year high or low for that particular day.
    It's just how averages work."

    Interesting post ... now call me a pedant, but as there are 100 places which could have a record high and 100 that could have a record low ... there are 200 chances of either a record high or low - so on average two places will be having a record 100 year event!

  • obloodyhell

    }}} some area representing 1% of the globe should constantly be seeing a 100-year high or low for that particular day. It's just how averages work.

    A high or low AVERAGE, yes -- but not a high or low, which is not the same thing. Reportage of extremes are not generally the averages but the extremes themselves. These are not subject to a linear distribution, but a normal curve. So in this case, your argument seems defective.

    I do agree that reportage of such extremes is generally blown far out of proportion -- but the reason for the "extremes" is probably because we REALLY don't have that long a base-line for many of these extremes. 25y-50y for any given area, 100y for a more historically populated one, and maybe 200y for some of the longer-term ones, prior to which, the RELIABLE temperature reportage is in the low hundreds of sites going back only another 200y tops. And when it comes to something like climate, even a 400y baseline is not a hell of a lot.

    And many of those areas are affected by other things, too, such as the "Urban Heat Island" effect, so allowing for the affect of those things is contentious at best. That's one reason why the alarmists don't want to release their data, to the extent that they're even DESTROYING the raw data and eliminating that base line.