Most people spending much time on the blogosphere are well aware of claims that “Global Warming stopped in 1998” or similar-style remarks. Even though the 1998-2008 period contains most of the warmest years on the instrumental record (something that is very unusual), and all the years are well above the traditional 1951-1980 (or 1961-90) climatologies, a key focus for skeptics has been the lack of upward slope in a linear regression over the 1998-2008 period.
It has also been emphasized by many that a traditionally defined climatology involves roughly 30 years of data, and so at least a few decades are needed to say much about the underlying trends in climate. However, there is not much in the peer-reviewed literature regarding the probability (or significance) of decadal flatlines, or coolings, when the climate regime is superimposed on a long-term warming trend due to radiative forcing. This is the subject of an upcoming paper in Geophysical Research Letters by David Easterling and Michael Wehner in Is the Climate Warming or Cooling? (subscription required, abstract not available since the paper has not been formally published yet). Their conclusion is that these kind of decadal time-frames can yield slopes of either warming or cooling in a warming world, even in the later 21st century, and nothing is odd about the 1998-2008 trend.
We can get a glimpse of how flimsy short-term trends are for climate change analysis. Here is the full NCDC (base period 1961-1990) record (see Smith et al 2005 which the authors take after).
Sorry, not very good at clear images from Excel. A closer look at the 1970-2008 period reveals a steep upward trend that is statistically significant.
A look at 1998-2008 reveals little trend
However, 1999-2008 looks much different, with a slope almost a factor of two greater just based on removing an anomalous data point (the 1998 El Nino)
The lack of trends over a small timeframe in the last several decades is not unusual. For instance, the 1977-1986 interval contains essentially no trend
The bottom line is that intervals are imbedded in the modern anthropogenic time period which contain no statistically significant trends, and such period of one, even up to two decades can exist without significant rises in temperature. Furthermore, one can remove or add just a single year when doing these small time-frame analyses and get a much different picture of what is happening.
The authors construct probability distributions with observations and models on the likelihood of getting a decade of warming/cooling conditions over pre-industrial, modern times, and the 21st century. The likelihood of a cooling decade diminishes under a warming trend and especially in the extreme A2 scenario (which postulates a “business as usual” future), but is still not impossible . Like playing with loaded dice
And the verdict from the paper:
“Therefore, it is reasonable to expect that the natural variability of the real climate system can and likely will produce multi-year periods of sustained “cooling” or at least periods with no real trend even in the presence of longterm anthropogenic forced warming. Claims that global warming is not occurring that are derived from a cooling observed over such short time periods ignore this natural variability and are misleading.”