Slashdot Mirror
NOAA: Earth Smashed A Record For Heat In May 2014, Effects To Worsen
Freshly Exhumed (105597) writes with news that NOAA's latest global climate analysis is showing things are getting hotter. From the article: Driven by exceptionally warm ocean waters, Earth smashed a record for heat in May and is likely to keep on breaking high temperature marks, experts say. The National Oceanic and Atmospheric Administration Monday said May's average temperature on Earth of 15.54 C beat the old record set four years ago. In April, the globe tied the 2010 record for that month. Records go back to 1880. Experts say there's a good chance global heat records will keep falling, especially next year because an El Nino weather event is brewing on top of man-made global warming. An El Nino is a warming of the eastern tropical Pacific Ocean that alters climate worldwide and usually spikes global temperatures.
Water vapor isn't a spoiler - the bands that it absorbs are different from the bands that CO2 absorbs. That's all there is to it really. There's so much water in the atmosphere that water vapor bands are essentially entirely absorbed. We can't reduce the amount of water in the atmosphere, and we wouldn't want to even if we could, so any gains that we make have to be outside the water absorption spectrum.
Such equipment records surely don't exist back to 1880.... So, my argument stands.
Just because you think something couldn't possibly be true doesn't mean it isn't.
Thermometer manufacture was precise enough to produce very high quality instruments by the late 1800s. The issue, as GP points out, isn't random errors popping up (thermometers were about as accurate as rulers could be in the late 1800s), but rather whether we know how the instruments were calibrated and how older scales might match up to instruments that were calibrated against modern international standards.
And, at least in the U.S., standard regular calibration standards had been agreed upon by the first couple decades of the 1900s. Good instruments even decades before then should have very little random error -- the question is only whether anyone bothered to check new equipment calibrated to international standards against the old equipment (or sent the old equipment to be tested once such standards were developed).
So, then you have to ask yourself: people who are bothering to meticulously record scientific data continuously for decades on end -- and they're not going to even bother to check whether their old instruments line up properly with new calibration standards?
Sure, I'm positive there are plenty of places that don't have such records. But we have continuous logbooks going back for centuries in many places. The idea that people taking meticulous records wouldn't even bother to check new equipment against old just seems a little ridiculous... not saying it wouldn't happen, but we have plenty of data points where it did happen to extrapolate estimates for error distributions.
You're acting like temperature measurement in the late 1800s was people guessing random numbers or drawing them out of a hat. But that's not what it was like, and there were lots of places with VERY detailed records.
You must have failed your stats 101 course in college. Let me educate you: any random sampling has limited accuracy. A randomly sampling of real temperatures that is made using low quality thermometers will probably have a very limited accuracy. But as you take more samples, you will either find that the measurements are biased or that they are normally distributed. If they are biased, you should be able to model the bias and apply a correction to achieve a normal distribution. Once you have a normal distribution, you know that an increase in the number of samples allows to say with a greater degree of confidence that the real value is within a certain range of the mean of those samples. So, even if you have shitty thermometers, you may be able to draw useful conclusions from them if you have thousands of measurements. This is the kind of stuff that the big boys talk about when they prepare journal articles for peer review. You know, p 0.05 and all that.
A cursory googling doesn't paint Steve S. Goddard as someone that has a rigourous approach to statistical analysis.
(1.21 gigawatts) / (88 miles per hour) = 30 757 874 newtons