Back

Hiatus delays unprecedented warming rates

Blog post by Richard Betts (with Figure 3 by Ed Hawkins)

Current global temperatures are often discussed in terms of their unprecedented nature when compared to the last few thousand years. An interesting paper in Nature Climate Change by Steven J Smith and colleagues examines the rate of warming projected by the CMIP5 ensemble and suggests that the rate of warming is unprecedented also. However, we note here that their projections are not constrained by the current observations which do not show such strong warming rates at present, and are unlikely to do so in the next few years.

Smith et al rightly point out that the rate of change is a critical factor in climate change impacts, especially if the change is so fast that it cannot be adapted to. We discussed this in the IPCC Working Group 2 report:

Figure 1. (a) 20-year rates of global warming from observations and CMIP5 simulations. (b) Corresponding ‘climate velocities’, quantifying the shift of local temperature states across the landscape as the global climate warms. (c) Displacement rates of groups of species – comparison with climate velocity gives an indication of whether species can adapt to warming by tracking the movement of preferred temperature states across the landscape. For further details see IPCC AR5 WG2 Chapter 4, ‘Terrestrial and Inland Water Systems’ – this is Figure 4-5 from that chapter.

Smith et al note that the projected rates of warming are faster than those seen in reconstructions over recent centuries, and discuss the implications of uncertainties in transient climate response and aerosol forcing – it’s an interesting and useful analysis. However, much of the media coverage of this paper focussed on a somewhat surprising statement in the abstract:

The rate of global-mean temperature increase in the CMIP5 archive over 40-year periods increases to 0.25±0.05 °C/decade by 2020

This is surprising because global mean warming has proceeded more slowly over the last decade and a half or so – the so-called warming slowdown or ‘hiatus’. After a warming rate of around 0.26 °C/decade in 1984-1998, the rate slowed to about 0.04 °C/decade in 1998-2012 (IPCC WG1 Chapter 9 Box 9.2).

Will warming really accelerate again enough to give a 40-year warming rate of 0.25 °C/decade by 2020?

The authors do acknowledge the observed recent slowdown, and they present the observed warming rate in one of their figures, which uses a simpler climate model (see figure 4a in Smith et al.). However they do not factor this into the projected 40-year warming rate by 2020. (To be fair, we did not do this in our IPCC figure either!)

We suggest that one way to account for recent observations is to use the constrained near-term projections presented in IPCC WG1 Chapter 11 (Figure 2). This figure took the range of warming rates from CMIP5 and applied these to a more recent baseline. This acknowledges that while the projected warming rates may be realistic, the warming in coming decades is now starting from a lower point than in the projections.

Observational constraint of near-term warming – this was figure 11.25 in IPCC WG1 Chapter 11, ‘Near-term climate change: projections and predictability’. See Section 11.3.6.3 for further details.
Figure 2: Observational constraint of near-term warming – this is part of Fig. 11.25 in IPCC WG1 Chapter 11, ‘Near-term climate change: projections and predictability’. See Section 11.3.6.3 for further details.

Using this in combination with the observed global mean temperature allows us to estimate the 40-year warming out to 2035 taking the hiatus into account:

40-year warming rates for the HadCRUT4 observations (black), CMIP5 projections for RCP8.5 (red) RCP4.5 (light blue) and RCP2.6 (dark blue), and for the observations plus the IPCC’s ‘assessed likely range’ (grey plume).
Figure 3: 40-year warming rates for the HadCRUT4 observations (black), CMIP5 projections for RCP8.5 (red) RCP4.5 (light blue) and RCP2.6 (dark blue), and for the observations plus the IPCC’s ‘assessed likely range’ (grey plume).

Doing this, and assuming of course that the IPCC’s ‘assessed likely range’ is realistic, we find that the 40-year warming rate is actually expected to reduce a little over the next 5 years or so. The 40-year rate peaked at about 0.18 °C/decade in 2011 and then drops off a little by 2015. This is due to a few cold years in the 1970s dropping out of the 40-year period. The 40-year rate is then projected to be between 0.14 and 0.18 °C/decade by 2020. By this method, Smith’s central estimate of 0.25 °C/decade looks unlikely to be reached by 2020.

The wider point made by Smith et al still stands – there is a clear risk of rates of climate change exceeding those in recent centuries, especially if greenhouse gas emissions continue to increase. Recent 40-year warming has already been unusual in the context of the last millennium, but possibly not completely unprecedented. However, our brief analysis here suggests that the recent hiatus may have put off the time at which 40-year warming exceeds that seen in the last millennium. Nevertheless, this may still happen within the next two decades. As Smith et al discuss, a key factor will be the transient climate response.

 

Re-blogged from Climate Lab Book