During the day you might start off sunny and then feel cooler when the wind goes up or a cloud goes past. But as the sun rises, the trend is for temperatures to increase.
During the year you might be in the UK in March and get some cold weather, so temperatures drop for a few weeks. That doesn't prove that summer isn't warmer than winter.
If you sit on the beach as the tide comes in, you can watch the waves lapping on the shore. Sometimes a wave will not reach as far as one before so the water level has dropped. That doesn't prove that the tide coming in means that water levels are going down.
I think MTRStudent gives a good answer. If the temperature at the beginning of April is higher than the temperature at the end, is that proof that summer will not follow spring? It would be to the denier mindset.
Ottawa Mike claims "All noisy systems have underlying trends." I have no idea why he says that, but it is certainly not true, Then he posits that engineers have doubts about AGW because of their experience with noisy systems--when that would instead be evidence that they should discount any "pause" in warming.
Kano once again claims "Once you get past 150ppm of CO2 it has little to no warming effect." even though he knows that is wrong.
EDIT: Rio, I actually used to work on systems where the SNR for an individual experiment was much less than 1. Nevertheless the signal was quite real--we'd just run the experiment 10,000 times or so and add up the results and the signal would become clear.
On the other hand, there is no guarantee that a noisy system will have any real signal in it, or "underlying trend". Sometimes noise is just noise.
All noisy systems have underlying trends (i.e. the signal).
I think engineers are more likely skeptical of AGW based on the idea of signal to noise ratio. In systems where the noise is much larger than the signal, there is little to be gained. For example, the change in global temperature not only has short term noise (variations) but the signal is generally smaller than even the noise of the instruments used to measure it.
Extracting a signal buried in noise is generally a very difficult problem in engineering. That's why engineers question the confidence of most climate data let alone future predictions of it.
Mount Everest. It's still a mountain even though every step you take may not always be higher than the last.
Kano: You can put CO2 into a tube and pass white light through it. When you increase the concentration of CO2, you increase the temperature of the gas. And people did that 100 years ago. So how, exactly, can there be 'zero empirical evidence' when we've more than a century's worth of experimental data?
Kano: The research you cite, although popular on denialist webpages, was wrong. This was because the conclusions were expanded to the planet based on the then poorly understood properties of radiative transfer in a temperature-stratified atmosphere. Similarly, it was also believed at the time that the overlap in CO2 and water vapour bands meant water vapour would effectively nullify the effect of CO2, and it was believed that the huge carbon content of the oceans would overwhelm any effect of human-produced CO2. These ideas were shown to be wrong. In the case of ocean CO2, this was shown to be incorrect in the 1930s. In the other cases, these were disproven in the 1950s, particularly by Charles Keeling of Caltech. A good reference can be found in Principles of Planetary Climate by Raymond T. Pierrehumbert (Cambridge University Press).
As fundamentalists, alarmist often get confused with long term system dynamics and short term weather statistics.
It would be hugely gratifying, if they made up their minds.
Anything greater than a SNR of 1:1 is considered a signal. Peg is so FOS.
Edit: probably one of the better examples of how noise and signaling can be confused.
http://www.pbs.org/wgbh/aso/databank/ent...
The overall trend in climate is not only noisy, but it is also rather slow. Over the space of a few years, the trend is just a few tenths of a degree, but people often see changes of several degrees from year to year.
One system that has a similar pattern is the seasonal temperature change vs. temperature change between day and night. Temperature change between day and night tends to be bigger than the temperature change from day to day or even month to month. But, using denialist logic, if we get a warm spring day and frost the next morning, we should expect to be under a mile of ice by July.
Every system, both human-made and natural, is "noisy" (i.e., there is some [however small] part of the varianece that cannot be explained). The speed of light has a defined solution; but, there is more than one way to calculate the speed of light, and depending on the method used (and based on assumptions about the properties of the cosmic vacuum) the solutions are not exactly the same. Newton's "Universal" Law Of Gravitation (and Einstein's Theory of Gravity) both contain a "constant" variabile whose calculation is equivocal. Pi is an irrational number.
=====
Kano --
>>There is zero empirical evidence pointing towards CO2 having caused any warming.<<
You could not be more wrong. AGW was discovered empirically and it was defined based on empirical evidence (empirical evidence that overwhelmed the skepticism of almost every climate scientiist in the world during the late 1990s and early 2000s). AGW was identifed specifically because the empirical evidence (global temperature data) over the last century or so began to significantly (measuraqbly, empirically) deviate from its historic pattern and from the patterns of every natural force that is known to influence global temperature.
In searching for an explaination of the new and changing empirical observations, the laws of physics dictated that one place to look was the earth's atmosphere. By collecting and analyzing empirical evidence (the spectral properties of gases in the atmosphere), it was discovered that the trapped energy matched the wavelength (spectral) properties of CO2. Based on this information, scientists in numerous scientific fields developed and field-tested hypotheses to determine if a CO2 signal was present in their empirical data. Subsequent backcasting vaslidation experiments from computer models provided important supporting evidence of the anthropogenic CO2 effect; however, scientific acceptance of AGW theory did not develop from - and is not dependent on - climate models.
AGW exists because of empirical evidence - and without empirical evidence no one would have ever mentioned AGW. Once again, just because you are unaware of the empirical evidence and the scientific reseach that has been conducted, especially over the last several decades - it does not mean that there is no empirical evidence or that the thousands of scientific studies do not exist.
=====
OM ---
>>I think engineers are more likely skeptical of AGW based on the idea of signal to noise ratio. <<
You're right; but, on balance, that's not the right way to look at it. Signal detection really is an engineering tool. Almost everything is based on and derived from the need to identify our own induced signals. I've commented in professional papers that trying to "discover" signals in natural systems makes little sense if you don't already have an expectation of what signal(s) could be in a series and an explanation of why they might be there.
That is the reason why scientists who deal with large natural systems are so obsessed with their data. We really almost need to brute-force our way to solutions by accumulating massive and overwhelming data. And that is precisely the reason that the scientific consensus on AGW was so late (late 1990s - early 2000s) in forming. Scientists and teams of scientists were hard-line skeptics until they actually saw evidence of AGW in own specific their data.
It is not just a coincidence that the timing of the broad acceptance of AGW in the scientific community follows immediately after the 1997-1998 El Nino event. That signal was so strong and so visible in multiple types of data that everyone saw it in their data. Prior to that event, it was a struggle to identify the ENSO signal - and without that (or something) there was no way to really sort out any other signals. For example, at that time the scientific community was split about 50-50 on whether the PDO was even a real thing (or just some spurious artifact or function of some other oscillating system).
Stature of Japanese adults in the 50 years following WW2. Over a foot of variation in the population, yearly trend due to diet improvements less than 1/2 inch but clearly discernible with decent statistical technique.
Populations of snowy owls and lemmings, and I don't believe in an underlyling trend, sorry.
Once you get past 150ppm of CO2 it has little to no warming effect.
There is zero empirical evidence pointing towards CO2 having caused any warming.
Elizabeth. The glass tube experiment was carried out by Knut Anstrom who determined that CO2 is saturated at 50ppm which he pointed out to Svante Arrhenius.
So climate scientists are unable to model noisy systems then. Gotcha.
One of the problems with explaining climate change to laymen is that climate is a "noisy" system. The changes (especially those in surface temperature) from day to day, and from year to year, and to some extent even from decade to decade, are far larger than the underlying signal from CO2 warming. This is where we get confusion ranging from "I just shoveled 3 feet of global warming off my driveway" to "no warming for 17 years".
Does anyone know of any popularly understood system that has a similar pattern? That is, one where there are very large, essentially random or cyclical fluctuations, but also an underlying factor, smaller than those fluctuations, that will tend to cause the overall set to increase or decrease. If not, what is your best science-illiterate-level analogy or simplified explanation for the phenomenon? (please describe the analogy or explanation yourself, don't just link to it, though if you have a link, feel free to add it)
