In a recent post the question of sea level rise and the rate of increase came up. Here are some sobering calculations, based on the IPCC AR4 scenarios and figures, and taking into account what has been left out of them.

In the SPM released in February, sea level rise was offered ‘excluding rapid dynamical changes in ice flow’. At the same time, the underlying assumption of the table showing the increases was that the rate of sea level rise would be roughly linear relative to the decade 1993-2003.


Taking the SRES B2, A1B and A2 scenarios as the ‘middle ground’ (and arguably the most likely of the scenarios), the IPCC plumps for a rise of around 4.2mm per annum, or 42 cm by 2090-99 (give or take).


On RealClimate, a comment was posted on why a rise of 1m by 2100 was plausible. At the same time, the recent presentations by Jim Hansen, which have also been much discussed on the various fora, present similar reasoning for a similar or greater rise, taking the ice sheets into account and considering the likelihood of increasing contributions to sea level in a BAU scenario, and a feedback effect in the Polar regions.


Here are some calculations based on the numbers which have been provided by the IPCC, and taking into account the probability that dynamical changes in ice flow are restrained but measurable. This still does not take into account the ‘collapse’ scenario.


The 3 scenarios give a range of sea level rise of 20-51cm by 2095. The IPCC also allows that the GIS and AIS contributions may add 10-20cm to that level, increasing the upper range to 71cm.


Total observed rise 1993-2003 is given as 3.1mm/yr +- 0.7. Unaccounted for discrepancy between estimates and measurements is 0.3 +- 1.0mm/yr. Of this, the GIS contribution is calculated as 0.21 +- 0.07, the AIS 0.21 +- 0.35 mm/yr. Note the large error allowances. Comparing the unaccounted rises (median positive error 0.8mm/yr) and the ice sheet contribution errors (0.42 +- 0.42mm/yr , ie: up to 0.84mm/yr) might suggest that there could be a connection between the two. This could reasonably be interpreted as suggesting that an ice sheet contribution of 80mm this century is within the existing error calculations. This would take the IPCC mid-range estimate of 42cm up over the half-metre mark.


Then we can take into account the unlikelihood that ice sheet contributions are going to remain static. Even allowing for a small dynamical response of +5%/decade, based on the ’93-’03 estimate of 0.42mm/yr, by 2053, the ice sheet sea level contribution will have gone up to around 24.4mm. This is only a fraction of the ‘allowance’ of 10-20cm by 2095 offered in the SPM. So we have to conclude that, for ice sheets to contribute five to ten times as much as they currently do, we would need to see dynamical increases in contributions in the order of 15-30% per decade, compounded. Is this possible, or likely? This depends on how much damage has already been done, and how sensitive the ice sheets are to further increases in local temperatures. Given the rate of Arctic sea ice loss to date, and recent evidence from both ends of the cryosphere, it is not beyond the bounds of possibility.


Taking these possibilities into account, and being conservative in anticipating adversity, a new range of estimated sea level rise by 2050 comes out between 21 and 71cm, the lower figure being based on the estimate used for the WG2 document, the higher being a reasonable increase (within the error allowances) based on rapid growth in ice sheet contribution over the next 4 decades. One conclusion from the above calculations is that it is reasonable to imagine an increase in global sea level by 2053 of around 50cm.


Why does this matter? The recent publications of the impacts and adaptation summary of the IPCC bases its imagined problems on a sea level rise of 33.6cm by 2080, or around 21 cm by 2050. The consequences of such a rise are already seen as very damaging. What consequences are we likely to see if the estimates are, in fact, too conservative, as has been implied? Firstly, we can expect the impact time scale brought forward by thirty to forty years. Secondly, the impacts on coastal systems by 2080 would be two or three times worse than the estimates currently allow for.


Given that it is generally accepted that we cannot avoid the next fifty years of warming whatever we do, the conclusions might be:


that policy makers and the public should be alerted to the possibility that investing in land or property in coastal areas is a poor risk;

that several developing countries need urgent help to develop adaptive strategies now, not soon, and that this must include large cash inflow into vulnerable areas;

that large-scale human migration from developing into developed countries is very likely to increase dramatically as survival and the already poor quality of life become increasingly difficult to sustain, and that the global economy and political stability is very likely to be increasingly compromised as the next four decades unfold.


When is the time for the developed nations to act?