One of the problems with these kinds of analyses is that the arbitrary choice of the window length runs the risk of mis-identifying an upswing in an oscillatory behavior as a 'trend'. There is a periodicity evident in the data sets with a wavelength of about sixty years. Any choice of data which utilizes a time period less than 60 years, or utilizes a period longer than but not an integer multiple of 60 years, runs the risk of 'finding' a trend that does not exist, or grossly exaggerating one that is present.
This is a fundamental problem in the way the CAGW advocates present data. The (approximately) 60 year oscillation hit a trough about 1979 and was on an upswing from then until perhaps 2002. Any 30 year (or shorter) analysis of recent history will lie within that upswing, and will tend to 'find' a false upwards trend.
If you remove the 60 year cycles from the data, there is still an upward trend (0.059degC/decade), but it is less than the rate at which temperatures have recovered since the little ice age.
In other words, there is nothing happening now that is outside the range of natural variability.
You sir are cheating. With any window of 30 years or more, even 20 years or more you know the empirical data will not support large warming trends. Please do not confuse us with the facts.
Basic point is the most recent trend is slightly negative. With the Grand solar minimum coming on, the weak cycle 24 and projected near nonexistent cycle 25 unless CO2 is capable of doing everything the alarmists claim it can we are in for some significant cooling.
Likely hood of a 2C global temp increase by 2050? Very near zero.
EDIT: @ Mike, the 1970's were an abnormally cold period. To say global warming started after a "cold" period is cherry picking your data at best and disingenuous to blatantly dishonest at worst.
EDIT 2: @gcpn8 if you reduce the current CO2 concentration by 80% then all higher forms of life on earth will die. (Yes, this includes humans) This is because there will NOT be enough CO2 in the atmosphere to support photosynthesis.
It is generally acknowledged that global warming started in the late 70s. So going back centuries will not help, as you then only have cycles of warming and cooling thrown together. But yes, I have found some trends more than .2 per decade. GISTEMP 1984-1999 comes in at .233/decade.
What this tells us is that the extreme warming scenarios are unlikely, and warming is likely to be less than 2C.
This is part challenge and part question.
Challenge:
Using any modern global temperature record set (e.g. HadCRUT, GISS, UAH, etc.), pick any time period 30 years or more for the largest temperature increase trend you can find (e.g. x.xxC/decade). As a matter of fact, I'll make it a little easier. You can use any time scale of 20 years on more. I'm betting that you can't find one greater than 0.2C/decade.
Question:
For limiting future warming, this is a generally agreed international target: "The G8 countries have committed to limiting global warming to 2°C and reducing their greenhouse gas emissions 80% by 2050." http://ec.europa.eu/news/external_relations/090710_en.htm
Given that we have already recorded a 0.8C increase since industrial times, that means we have 1.2C of "headroom" left between now and 2050. That works out to a required trend of 1.2C/3.7decades = 0.32C/decade.
We have seen virtually no CO2 reductions to date and are generally following a "business as usual" scenario. Using the maximum trend you found in the challenge, describe how likely it will be that the 2C limit will be reached in 2050 even if we don't reduce CO2 at all?
