The following article was written by George Morrison (member of Citizens Climate Lobby)
First of all, we know that the earth system is accumulating heat.
We know this because satellites can measure the amount of incoming solar radiation versus the outgoing longwave radiation at the very top of the atmosphere. And more is coming in than leaving.
So we know that, in aggregrate, the oceans, ice and land are absorbing that net heat anomaly. But between those heat sinks, there is a lot of sloshing around (e.g. phenomenon like El Ninos and La Ninas), and different heat sinks have different capacities in terms of their ability to quickly absorb or spill heat. But if we look at the accumulation of heat in total - especially the relatively massive amount that goes to heating the oceans, there has been no pause:
And, just for fun, a short experiment in this video to demonstrate the relative heat capacities of air and water (I was quite surprised myself at this!). ( http://www.youtube.com/watch?v=SytY0qI6GMw )
Ok, but why does the "16 years of no warming" talking point seem to have caught on?
Well, to begin with they are only looking at global surface (air) temperatures. That's not the whole picture, but it's also not a total cherry pick, because (a) we live in the air, at the surface, it's what we are most concerned with and (b) the physical mechanism by which global warming occurs is initially via warming in the atmosphere.
And, if you simply look at a specific temperature data-set (from the Hadley Climate Research Unit, HadCRU...) from mid-1997 (a cherry picked date, see below) to present, and analyze it just by eyeballing it (i.e. drawing a line through it that fits what your eye/intuition tells you, and using no statistical tests)... well, you can see why some people are persuaded by this (this is the actual graph that "skeptics" are referring to when they make this claim):
But if you look at the data over a longer period, and with more robust statistical analysis, the picture changes a great deal.
Here is what the prior ~ 20 years looked like (in blue), with a linear regression applied. The mid-1997 to present data is in red.
Now one needs to squint a lot harder to see the mid-1997 to present data looking quite as flat. In fact, if you simply continue to extend that prior regression, it looks like this.
What would have happened if that trend continued unabated? This:
What would have happened is pretty much what did happen. Imagine that.
And a last point, for now.
I mentioned that there is a fair bit of natural variability in the surface temperature record - cycles of ocean circulation that either warms or cools the atmosphere (El Nino, La Nina), volcanoes, etc.
In this article, "About the lack of warming...", the author distinguishes amongst El Nino, La Nina and "Neutral" years, to correct for some of the noisiness in the series. And it starts to look more like this, which more clearly shows the signal amongst the noise.