The importance of the ‘pause’ in global temperatures has never been about whether some particular trend is just below or just above zero, or about whether a specific period shows a statistically significant trend (whatever that actually means), or even about whether the trend has changed. The ‘pause’ has always been about understanding whether Earth’s climate is evolving in line with our expectations.
Several papers have now emerged suggesting that there has never been a ‘pause’. This claim is based on a rather narrow definition – if you compare linear trends over certain periods then the linear trend has not actually changed by much. For example, the recent papers by Karl et al , Lewandowsky et al (in two similar papers), and Rajaratnam et al all compared trends over specific periods and suggest that there has been no ‘pause’ or ‘slowdown’, defined by changes in linear trends. And, Cahill et al find no recent ‘change point’ in global temperature trends.
Estimating linear trends is a helpful structure (and I have used it too), but I think these papers fundamentally miss the point – we do not expect the climate to change linearly.
There was a recent period in which the observed trend was on the edge of the uncertainty derived from the raw projections from CMIP5, whichever observational estimate you use – the IPCC AR5 made this point in the Figure below (only using HadCRUT4). And, it is not just global temperature – for example, the winds in the Pacific are behaving outside expectations.
Science proceeds by taking an observation that does not match expectations and seeking to explain it. There are many possible explanations for this particular observation which have been analysed extensively, and the climate community now has a much better understanding of: the nature of climate fluctuations, the role of the Pacific and Atlantic Oceans in sequestering heat, the importance of observational coverage, how to best compare observations and models, and the role of uncertain radiative forcings.
This research has been valuable and worthwhile, even if global temperatures are about to break new records, in line with our expectations that the surface would indeed warm again.
ADDITION (19/09/15): Consider a hypothetical example where the observed warming had accelerated to (say) 0.25K/decade, but the GCMs had projected 0.4K/decade. Much of the same research would have considered the discrepency. The magnitude of the trend is not the point.