You can read the abstract here and I have uploaded a pdf of the poster here. There is a picture of the poster below, click on it to make it readable, although you will need to download the pdf to make out some of the fine print.

Some of the numbers changed a little bit between the time I submitted the abstract in August and now. I found one or two small errors and recalculated the uncertainty range using Monte Carlo analysis. “Min” and “Max” values bracket the 90% confidence interval. In the title I used “directly” to distinguish the physical effects of emissions on ocean volumes from the more “indirect” (and bigger and better-known) contributions of emissions to sea-level rise via the effect of emissions on global warming.

The gist of the study is that emissions of water and carbon dioxide from burning fossil fuels add to the mass and volume of the oceans, causing sea-level rise. Something like this had been done previously by Gornitz et al in 1997 for the period 1980-89. At that time, the contributions to sea-level rise from fossil-fuel consumption were rightly deemed to be negligible, especially given the big uncertainties in the other major contributors to sea-level rise, such as from the melting of land ice and from thermal expansion of the oceans.

However, since the 1980s, CO2 and H2O emission rates from fossil fuel consumption have increased by more than 75%. At the same time,  the uncertainties in the estimates of the other big factors that contribute to sea-level rise have been reduced and are now more closely reconciled with the improving observations of sea-level rise. The estimates I have made of the contributions of combustion emissions to ocean volumes are now of a similar magnitude to the residual between contributions to and observations of sea-level rise, although there are significant remaining uncertainties in the sea-level balance sheet.

The study uses the socio-economic models for the Representative Concentration Pathways 2.6 and 8.5 to estimate sea-level rise from fossil-fuel emissions out to 2100. I am not aware of any previous work that has attempted this. The estimates amount to about 1-2% of the quantities attributable the sum of the better-known culprits of sea-level rise.

Although small in relative terms compared to the wasting of ice sheets and the thermal expansion of sea water, the quantities involved are huge in absolute terms.  For example, the amount of water added to the the surface hydrological cycle from the burning fossil fuels to date is roughly equivalent to the volume of water in Lake Erie. The combined future contributions of water and carbon dioxide to sea level rise are comparable to the projected contribution of melting glaciers in major regions like, for example, South America, or they are about half of the amount from the melting of Greenland’s peripheral glaciers.

The chemical effects of dissolution of CO2 into the oceans—ocean acidification— have been extensively studied, but the physical consequences for sea levels have not received much attention previously. One mole of CO2 (44 grams) dissolved into sea water results in a volume increase of about 32 millilitres. Densities are increased also following CO2 dissolution, but in normal ocean conditions these concentrations and density effects are very small compared to the important density changes arising from temperature and salinity variations. However, in the rare places where CO2 concentrations are very high, such as in volcanic lakes like Lake Nyos in Cameroon, CO2 dissolution can cause density stratification of the lake water. This stratification  can be destabilized by bubbles of CO2 coming out of solution, with disastrous results. The tragic and catastrophic degassing of the lake in 1986, which killed 1700 people,  released a few hundred thousand tons of CO2  and lowered the lake level by up to a metre. This is the same kind of CO2 dissolution/volume effect that I have referred to here, but in reverse.

I should perhaps add that the sea-level rise that I have estimated from fossil-fuel emissions—at most an additional centimetre by 2100—has no policy implications. I just calculated the amounts because I was curious how big they might be.

Thanks to my Skeptical Science colleagues, especially Sarah A. Green, Peter Jacobs and Martin Stolpe, and to my old friend and colleague David Smythe for their advice on this project. Keywan Riahi, Detlef van Vuuren and Ben Marzeion kindly provided data.

If there are mistakes, they are all my doing, of course. I would be happy for anyone to point out any errors, either by commenting here or by email.