Here's something fairly recent (a guest post) on Watts Up With That.
Things I've recently learned from this great site:
1. Antarctic ice isn't shrinking, and it has not done so in recent decades. Some of it is splitting and cracking, which might be consistent with cooling and becoming more brittle. In any case, there are lots of photographic records of Antarctic ice splitting, then new ice forming quickly due to the frigid temperatures.
2. Arctic ice shrank in 2007, but it is basically back where it was. There is tremendous change from season to season, and from year to year, in Arctic ice: much of it usually melts in the summer, then forms again in the winter. Most Arctic ice is over water, whereas most Antarctic ice is over land. The Franklin expedition was trapped in Arctic ice in some freakishly cold winters. The North Pole has been ice free in the summer at times--even early in the twentieth century (before man-made CO2 could have been a factor). Even when the ice shrank in 2007, this may have resulted from the ice being pushed by the wind to warmer latitudes where it could melt, as opposed to melting in its usual cold spot.
3. There was a long-term trend of rising levels in the world's oceans, but weirdly, this trend has slowed or stopped just when the IPCC model of global warming would have it accelerating. The Maldives are no more under water today than they were decades ago.
4. Mountain glaciers are retreating in various parts of the world, often widely separated from each other. Surely this is a global trend? But: Kilimanjaro and others have been retreating since before man-made CO2 became a factor. A recent study indicates that there are cycles, rather than one clear trend, and northern hemisphere glaciers are on a different cycle than southern hemisphere ones. There have been times of retreating glaciers in both hemispheres, but there have also been times of retreating glaciers in one hemisphere, advancing glaciers in the other.
5. Surface temperature, for what it is worth, has not gone up since 2001--exactly the period when the IPCC model should have been proving itself through noticeable temperature increases. Man-made CO2 is definitely up, but it may simply not be a good match with any other specific factor in regional or global climate. The twentieth century was warm, and it saw a dramatic increase in man-made CO2, but even within that century the match was far from perfect, there have been warmer periods in history when there was virtually no man-made CO2, etc. Anthony Watts has a lot of fun with the fact that surface temperature readings depend on weather stations, many of which are ridiculously unreliable or biased.
The heat and drought in Australia's recent summer did not set records, and they seem to be cyclical; by contrast, the cold that is part of the fall now underway in Australia is setting records. North America has had two cold winters in a row, the UK has not had a 30 degree day in summer since 2006, etc.
6. This latest: heat in the oceans should be a better test of the IPCC model than mere temperatures there, or on the surface. The IPCC model would predict a steady increase in ocean heat; any departure from that into some kind of cyclical pattern would require at least a substantial revision of the model, and a period of prolonged continuity in ocean heat would probably prove the model wrong. There has been a period of five years (2003-2008) of no increase in ocean heat.
Heat is not temperature. "Two liters of boiling water contain twice as much heat as one liter of boiling water even though the water in both vessels is the same temperature. The larger container has more thermal mass which means it takes longer to heat and cool." I think this bears on the whole quest for specific local temperature readings that are unusually high. A specific spot on the surface may be unusually hot without affecting in any noticeable way the total heat of the earth.