Sunday, April 17, 2016

Conflicting Baselines: a climate nerd winge

I find it incredibly frustrating and baffling that the IPCC and other major climate science bodies like NASA use a variety of unreconciled baselines for global temperature changes in the reports. Sometimes it is 1951-80, sometimes it is 1981-2010, sometimes 20thC average and so on.* I have not found a convenient set of translations between these baselines in the reports that would enable you to, say, add 0.4ºC to get from a 1880-1909 baseline to a 1951-80 baseline.
*Since climate averages are defined scientifically (at least by the recommendation of the WMO) as requiring a minimum of 30 years of continuous data, most baselines are 30 year periods, rather than single points in time.

Given that the UNFCCC negotiations are based on a preindustrial baseline ("hold the increase in global average temperature below 2°C above preindustrial levels", Cancun Agreement, 2010), yet have never (to my knowledge) defined precisely what "preindustrial" actually means nor attempted to quantify its relation to other baselines, it is unconscionable that the IPCC have not made this a far more prominent frame for all their work. In other disciplines, preindustrial is generally taken to mean prior to 1750 or so. One challenge of using preindustrial as a baseline in negotiations for a legally-binding international agreement is that high quality temperature data from direct measurements (rather than proxies) with good global coverage only extends back into the 2nd half of the 19thC (depending on how good you want the coverage to be). Even if the IPCC noted that estimates of global temperatures prior to the late 19thC have error bars too large for meaningful negotiations and suggested that the UNFCCC make, say, 1880-1909 the universal baseline for negotiations, that would be defensible.

One reason they haven't done so is that this would probably require a reconfiguration of the global goals away from neat round numbers, since paleoclimate specialists (who reconstruct temperature data prior to the instrumental record from proxy data) say that between ~1750 and 1880-1909 there was likely about 0.2ºC of warming. Shifting to a goal of under +2ºC from 1880-1909 would probably (and should) be resisted by those nations most vulnerable to warming.

Why does this matter? (Beyond obsessive concerns for clarity from scientists and those of us who appreciate precision)

It matters because the vast majority of journalists fail to mention (and may well not even be aware of) the issue of different baselines. The media thus regularly refers to goals like keeping temperatures below +2ºC without specifying their assumptions. Since most new scientific publications use a baseline considerably later than preindustrial, this means that many articles reporting on scientific findings give a very misleadingly rosy picture of the scale of ambition required to achieve the agreed UNFCCC target. It is much easier to meet a +2ºC from 1981-2010 target than a +2ºC from preindustrial (=~1750, or even 1880-1909) target.

And this matters because I expect that the number of parliamentarians who grasp these distinctions is also quite limited.