I'm afraid you have been led astray by people who were deliberately trying to misinform you. You should probably be angry about that.
>The problem is that the models have all been wrong.
All models are wrong, but some are useful. Climate models have demonstrated skill at modelling important parts of the climate system.
>They can't explain the past behavior of the Earth's climate, and they can't predict the future behavior. They're not even directionally correct, given the fact that global temperatures, albeit amongst the highest they've been on record, haven't increased since the late 1990s[1]. Given the incredible increase in CO2 over the past 16 years, the models would have predicted much higher temperatures, but it's simply not agreeing with their own data.
That is just factually incorrect. There are individual model runs that do match the observed surface temperature since the 90s. The recorded temperatures are still within the expected range of most model projections. Even Hansen's crude 1988 model is surprisingly close to what has happened[1], although that is probably in part down to luck.
The most important thing to know here is that climate models are projections, not predictions. That means they aim to model what happens given a certain set of circumstances. They don't aim to predict what the actual circumstances will be. Those circumstances include things like volcanic activity, which you can appreciate is not something that is easy to predict decades in advance. Also internal variability like ENSO, PDO, NOA , and NAO will play a large part in determining the short term (decadal timescale) surface temperature. As far as I understand most models just model ENSO stochastically, so for any given model run it's going to be down to random chance whether the ENSO state in the model run matches what happened in reality. Over longer periods of time and over many models runs the result is a range of likely climate (for various different emissions scenarios). The aim of climate models is to give an upper and lower bound to what the system will probably do for a given emissions scenario, not to provide a single forecast for the exact state of the climate in x years.
Consider writing a model of a die roll. If you know the die is fair you will be able to write a simple model that gives close results over a large number of throws, but it will have basically zero ability to predict a small number of throws. You could also adjust your model considering real data if it turns out that the die is weighted based on your observations. This isn't a direct analogy to a climate model, but it's a demonstration that models can be useful without the ability to predict the short term state of the system.
Having said all that, there are some areas where the models do pretty badly. Arctic sea ice is one, but in that case the problem has been a big underestimation of the actual rate of melting.
Finally, I'm not sure you read that link you posted because it discuses why 'no warming in 16 years' is a cherry picked figure[2]. It's a bit of a statistics game as well. If your null hypothesis is that the trend has continued unchanged then the data don't give any reason to think that the trend has changed[3]. Of course it's important to remember that surface temperature is only showing a small part of the total energy in the climate system, and our surface temperature record has gaps (it misses out most of the poles). Short term variability in the trend is to be expected; you can also cherry pick time periods when the short term trend was much greater, for example 1992-2006 [4].
[1] http://tamino.wordpress.com/2014/03/21/hansens-1988-predicti... this also shows some of the techniques that have probably been used to trick you into thinking models have been wildly wrong, which is handy if you want to know what to look for in the future
[3] http://tamino.wordpress.com/2013/02/12/2012-updates-to-trend... for a look at how recent years fit into the longer term trend. It uses 2000 as the cutoff - somebody redid it using 1998 but I can't seem to find that now, the story wasn't much different though
I disagree with you. I don't think that what you presented actually answers the points I've made except to rationalize why the models are completely wrong.
Let me is simplify things. Is there a model based on CO2 that explains the variations in global surface temperature of the last 1200-5000 years, as mentioned in the OP's article? That should be easier to project given that we are not talking about a period of 20 years, correct? I genuinely don't know, which is why I'm asking and I would like to learn more. Is there a model that projects the various ice ages and subsequent warmings over a longer period of time, including the medieval warming period, which you claim should be easier?
And a question for you: what evidence would you personally require in order to no longer in the greenhouse gas theory? I know what I need to see in order to start to believe in the greenhouse gas theory: I need to understand how CO2 is responsible for the variability of the Earth's climate pre-industrial revolution. If CO2 wasn't responsible for the Earth's climate change pre-industrial revolution, then what was it and could that same thing be responsible what for we see today?
What I've seen from people who believe this theory is that they will believe it no matter what. For example, if climate gets warmer, that's because of CO2. If the climate gets colder, that's because of CO2. If the climate goes wild, like during the time of hurricane Katrina, that's because of CO2. If it gets calmer, like it did after Katrina, that's because of CO2. I can't think of a single climate outcome that will disprove the theory, since all outcomes appear to prove it.
In my experience, a theory that can't be prove wrong despite different outcomes isn't a theory, it's a religious statement, which is not science. I believe that believers in the theory have moved over into the religious territory, which makes it about as fruitful as talking to a creationist.
> Let me is simplify things. Is there a model based on CO2 that explains the variations in global surface temperature of the last 1200-5000 years, as mentioned in the OP's article?
Classic crank tactic: Demand outcomes far beyond anything that's claimed by working scientists, let alone what's necessary to validate the hypothesis in question. Then declare victory for yourself when it turns out that everyone's got better things to do than spend time knocking down straw men.
Demonstrating which end of the Dunning-Kruger spectrum you're at isn't necessarily a great way to win a scientific debate.
>The problem is that the models have all been wrong.
All models are wrong, but some are useful. Climate models have demonstrated skill at modelling important parts of the climate system.
>They can't explain the past behavior of the Earth's climate, and they can't predict the future behavior. They're not even directionally correct, given the fact that global temperatures, albeit amongst the highest they've been on record, haven't increased since the late 1990s[1]. Given the incredible increase in CO2 over the past 16 years, the models would have predicted much higher temperatures, but it's simply not agreeing with their own data.
That is just factually incorrect. There are individual model runs that do match the observed surface temperature since the 90s. The recorded temperatures are still within the expected range of most model projections. Even Hansen's crude 1988 model is surprisingly close to what has happened[1], although that is probably in part down to luck.
The most important thing to know here is that climate models are projections, not predictions. That means they aim to model what happens given a certain set of circumstances. They don't aim to predict what the actual circumstances will be. Those circumstances include things like volcanic activity, which you can appreciate is not something that is easy to predict decades in advance. Also internal variability like ENSO, PDO, NOA , and NAO will play a large part in determining the short term (decadal timescale) surface temperature. As far as I understand most models just model ENSO stochastically, so for any given model run it's going to be down to random chance whether the ENSO state in the model run matches what happened in reality. Over longer periods of time and over many models runs the result is a range of likely climate (for various different emissions scenarios). The aim of climate models is to give an upper and lower bound to what the system will probably do for a given emissions scenario, not to provide a single forecast for the exact state of the climate in x years.
Consider writing a model of a die roll. If you know the die is fair you will be able to write a simple model that gives close results over a large number of throws, but it will have basically zero ability to predict a small number of throws. You could also adjust your model considering real data if it turns out that the die is weighted based on your observations. This isn't a direct analogy to a climate model, but it's a demonstration that models can be useful without the ability to predict the short term state of the system.
Having said all that, there are some areas where the models do pretty badly. Arctic sea ice is one, but in that case the problem has been a big underestimation of the actual rate of melting.
Finally, I'm not sure you read that link you posted because it discuses why 'no warming in 16 years' is a cherry picked figure[2]. It's a bit of a statistics game as well. If your null hypothesis is that the trend has continued unchanged then the data don't give any reason to think that the trend has changed[3]. Of course it's important to remember that surface temperature is only showing a small part of the total energy in the climate system, and our surface temperature record has gaps (it misses out most of the poles). Short term variability in the trend is to be expected; you can also cherry pick time periods when the short term trend was much greater, for example 1992-2006 [4].
[1] http://tamino.wordpress.com/2014/03/21/hansens-1988-predicti... this also shows some of the techniques that have probably been used to trick you into thinking models have been wildly wrong, which is handy if you want to know what to look for in the future
[2] http://tamino.wordpress.com/2014/02/09/cherry-p/ some relevant stats on cherry picking
[3] http://tamino.wordpress.com/2013/02/12/2012-updates-to-trend... for a look at how recent years fit into the longer term trend. It uses 2000 as the cutoff - somebody redid it using 1998 but I can't seem to find that now, the story wasn't much different though
[4] http://tamino.wordpress.com/2013/09/21/double-standard/