Figure 2 (top) shows one estimate of GMST (HadCRUT4) and the scaled forcings for the full 1765-2015 period, using the regression parameters derived over 1850-2015.
An alternative approach to estimate GMST further back in time is to use direct observations from long instrumental records and calibrate them against each of the five global mean temperature datasets.
CET covers just 0.005% of Earth's surface but is highly correlated with GMST on multidecadal time scales (Sutton et al.
using the overlapping periods (1850-2015), and adopt the same parameters to scale CET back to 1659 as an estimate of GMST (Fig.
These two periods produce consistent estimates for the warming to 1986-2005: 0.75[degrees] [+ or -] 0.10[degrees]C (for 1765-1800) and 0.64[degrees] [+ or -] 0.08[degrees]C (for 1720-1800) when using HadCRUT4 for GMST. The other global temperature datasets give larger values for the warming to 1986-2005, by up to 0.09[degrees]C (Fig.
Moreover, the Intergovernmental Panel on Climate Change (IPCC), which reflects the scientific consensus on climate change, adopted the term hiatus in its Fifth Assessment Report, and even gave it a definition "as the reduction in GMST trend during 1998-2012 as compared to the trend during 1951-2012" (IPCC 2013, Box TS.3).
Two analyses of the GMST time series have failed to find any statistical evidence for a slowdown (Foster and Abraham 2015), or a distinct changepoint in the rate of warming (Cahill et al.
1 shows all possible 15-yr trends in GMST for the period 1970-2014 (i.e., 1970-84, 1971-85, and so on; N = 31) for four different datasets.
The linear trend in GMST (established by ordinary least squares on annual global means) is statistically significant for the last 15-yr period (ending in 2014) for three of the four available datasets: GISS (trend, b = 0.08 K [decade.sup.-1]; test statistic, t = 2.20; level of significance p < 0.05), the dataset of Cowtan and Way (b = 0.10 K [decade.sup.-1], t = 2.41,p < 0.05), and the most recent NOAA dataset by Karl et al.