One of the strongest El Niño
events in the instrumental record occurred during late 1997 through 1998, causing a spike in global temperature for 1998. Through the mid-late 2000s this abnormally warm year could be chosen as the starting point for comparisons with later years in order to produce a cooling trend; choosing any other year in the 20th century produced a warming trend. This no longer holds since the mean global temperatures in 2005, 2010, 2014, 2015 and 2016 have all been warmer than 1998.
More importantly, scientists do not define a "trend" by looking at the difference between two given years. Instead they use methods such as linear regression that take into account all the values in a series of data. The World Meteorological Organisation specifies 30 years as the standard averaging period for climate statistics so that year-to-year fluctuations are averaged out; thus, 10 years isn't long enough to detect a climate trend.
In a BBC interview on 13 February 2010, Phil Jones
agreed that from 1995 to 2009, the global warming "trend (0.12C per decade) is positive, but not significant at the 95% significance level", though close.
This has been misleadingly reported by some news sources.
On 10 June 2011 Jones told the BBC that the trend over the period 1995 to 2010 had reached the 95% significance level traditionally used as a threshold by statisticians.