Schmittner is wrong when he says that the statement "global temperatures have not risen during the past 15 years" is not true. Most climate scientists agree that that the earth has not warmed during the past 15 years.Douglass is simply wrong -- the ocean has warmed strongly in that time, according to NOAA data:
This means the top half of the ocean has warmed by about 0.8 × 1022 J/yr, or an average of 250 terawatts (TW). Divided over the entire Earth's surface (viz. assuming all the extra heat goes into the ocean) it comes to an average of 0.5 W/m2. That's a lot of warming.
The two-dimensional surface of a planet with a thin gas above it is about the worst place to look for the kind of energy imbalance that additional greenhouse gases create. The oceans have a heat capacity (mass × specific heat) about 1,000 times that of the atmosphere.
Think of it this way: Imagine you have a children's swimming pool with a strong light shining down on it, with a rock in the middle of the pool that sticks above the surface.
Now you want to know, is the light above getting more intense?
Where are you going to look? Are you going to set your thermometer a few millimeters just above the surface of the rock, and see if the temperature there is changing?
Or are you going to stick your thermometer in the water and see if that's changing.
You will, of course, stick it in the water. The rock is close to the water so swirls in the water near the rock can affect the temperature you measure there, and winds carry heat to and from the oceans, etc. No one would dream of measuring just above the rock (though you'd surely start there, because you don't want to get your feet wet).
Of course, we live on such a rock, and that's ultimately the temperature we care about. But anyone would certainly expect, in the children's swimming pool, that if you measure the water warming up, the thin layer of air just above the rock is eventually going to warm up too.
It's no different. The surface is subject to natural variations, and choosing 15 years is a cherry pick. (Why not choose 12? Or 18? Or 30?)
Not to mention, there has been surface warming in the last 15 years. It's just that it hasn't risen above its statistical uncertainty by the canonical 2-sigma.
HadCRUT4's 15-yr trend is, from the SkS trend calculation, 0.098 ± 0.143 °C/decade (2σ). That's a positive trend, i.e warming, and statistically significant at the 83% confidence level. But note that it would have to be above 0.14 °C/decade to be statistically significant at the 95% confidence level, and that's not always going to be the case for short time intervals. Sometimes it will be even stronger, like the period 1975-1998. But picking that period would be cherry picking, too.
Unless temperature increases monotonically upward, with, say, each month 0.001 °C warmer than the month before, contrarians will always be able to find some period where warming is less than 2σ (95%) significant. If it's not 15 years it will be 12 years or 7 years or something. So this game is about all they have left to rely on.
And you can bet than when the next El Nino comes and surface temperatures shoot up, they will be crying that the increased trend isn't a proper measure because it cherry picks the last few years. Bet on it.