Signal to Noise, part II

We keep hearing that the climate today is warmer than it's ever, ever, ever been. You keep getting me yammering on and on (and on and on) about how we're trying to measure a very small signal (a degree per century) with instruments that measure a huge amount of noise (plus or minus twenty degrees a day).

In other words, the noise is much, much bigger than the signal.

But let's forget about the "trick" to "hide the decline" and all that. Forget all the mysterious "adjustments" to the measured data that make up over 80% of the warming in the 20th Century.

Ignore all that. Let's assume, for the sake of argument, that none of that matters. Here's the question: how accurate are the sensors themselves? Well for 90% of the ones in the US, the answer is "not very". You'll no doubt be shocked to learn that this 90% have a common error that makes them all read high (i.e. report a temperature that is more than 1 degree higher than the actual temperature). You'll also be no doubt shocked to hear that this is a recent occurrence, only since 1970 or so.

Anthony Watts (from Watts Up With That) explains why:





But if you lose sleep at night worrying about Global Warming, and are concerned that this make be just more ammunition for those skeptics to shoot holes in your pet theory, don't despair. The Fed.Gov assures us that they have a brand spanking new sensor grid in place to measure the temperature, and it's much, much more accurate. And in fact, it is more accurate.

And it's been online now for 18 months. And so, we can say with 100% certainty that 2009 was the warmest year since, well, 2009.

Hey you Deniers! Get off my lawn!