Arguments
How much will sea levels rise in the 21st Century?
Posted on 5 August 2013 by gpwayne
This post is a new 'basic' level rebuttal of the myth: "Sea level rise predictions are exaggerated."
What the science says
Sea levels are rising faster now than in the previous century, and could rise between 50cm to 1.5 metres by 2100
Measuring Sea Levels
Sea levels are rising due to thermal expansion and melting of land-based ice. Global warming is causing the oceans to absorb a lot of extra heat (up to 90%). This makes the volume of water expand, and sea levels rise. The Greenland and Antarctic ice caps, and many of the world’s glaciers, are all slowly melting. The runoff feeds into rivers and directly into the oceans. This too adds to sea levels.
Prior to the use of satellite systems, measurements were taken using tide-gauges, devices that measure the height of a water level relative to a fixed point on land. Global estimates of sea level rise were subject to substantial differences in measurement from different parts of the world.
Sea levels change all the time. They are affected by seasons, astronomical tides, storm surges, currents and density, among other influences. Tidal gauges reflect these short term influences, introducing a large margin of error.
The IPCC Fourth Assessment Report described studies that estimated sea level rise for the 20th century between 0.5 and 3.0 mm a year. The most likely range, according to the IPCC, was between 1.0 and 2.0 mm a year.
Satellite altimetry since 1993 provides a more accurate measure of global sea level rise. Three different satellites take measurements: TOPEX/Poseidon (launched 1992), Jason-1 (launched 2001) and Jason-2 (launched 2008).
Figure 1: Source - CSIRO
The IPCC projections are derived from climate models. Using both tide gauge and satellite data, we can see that sea levels are rising. Unfortunately, sea level rise is already tracking the worst-case projections, as this graph shows:
Figure 2: Sea level change. Tide gauge data are indicated in red and satellite data in blue. The grey band shows the projections of the IPCC Third Assessment report (Allison et al 2009).
In fact, the climate models underestimated the rate of sea level rise because the rapid melting of the ice sheets and glaciers was not incorporated in the last IPCC report. (It was left out because the data were not considered sufficiently robust).
Damaging Potential
Rising sea levels are widely considered to be the greatest threat posed by climate change. They threaten low-lying countries with inundation, forcing inhabitants to migrate. Coastal cities and ports could be flooded, as could cities sited near tidal estuaries, like London. Many nuclear installations are built by the sea so they can use sea water for cooling.
The potential for sea level rise is enormous. This is because the ice caps - Greenland and Antarctic - contain huge amounts of fresh water - around 70% of all the freshwater on Earth. Estimates suggest that if the Greenland ice sheet was to melt away to nothing, sea levels would rise around 6 metres. To put that a different way, a loss of just one per cent of the Greenland ice cap would result in a sea level rise of 6cm.
If the West Antarctic Ice Sheet (WAIS) were to melt, this would add around 6 metres to sea levels. If the East Antarctic Ice Sheet (EAIS) were to melt as well, seas would rise by around 70 metres.
In a process that is accelerating, all three ice caps are losing mass. While nobody is suggesting any of the ice caps will melt away to nothing, only a small amount of melting would cause great problems.
A 1% loss of ice from these three sources would produce a likely increase in sea levels of around 76cm. With the thermal expansion implied by such melting, and contributions from melting glaciers, the oceans would actually rise far more.
Predictions for future sea levels
Future sea level rises depend on a number of factors. The amount of CO2 emitted will determine how much global warming takes place. The amount of ice that melts will vary according to the amount of global warming. The same is true of thermal expansion.
Previous estimates of sea level rise have been based on a set of possible outcomes called emissions scenarios. These theoretical scenarios range from emissions which fall very quickly, to emissions that continue to rise even faster than they have already. Scientists then calculate possible outcomes for each scenario.
In the next IPCC report (AR5), due in 2014, a new method has been used. Emission scenarios have been replaced by Representative Concentration Pathways (RCP). Four trajectories were chosen, based not on emissions, but possible greenhouse gas concentrations in the year 2100. From the concentrations, the RCPs project a ‘forcing’ for each pathway (the amount of warming); 2.6, 4.5, 6.0, and 8.5 Watts per metre squared. Each pathway is named after it’s forcing e.g. RCP4.5. The lowest emission scenario is also referred to as RCP3PD, because it posits a peak warming of 3 w/m2 by 2070 (~490 ppm CO2 and equivalents), and a reduction to 2.6 w/m2 by 2100. (PD stands for Peak/Decline).
A draft version of the next report from the IPCC (AR5), due for publication soon, was recently leaked. Although the information is subject to change, the draft report says sea levels are likely to rise by between 29 and 82 centimeters by the end of the century, (compared to 18-59 centimeters in the 2007 report).
Other recent studies have projected comparable sea level increases. Jevrejeva 2011 for example modelled sea level rise using RCP scenarios. ) This table shows best and worst cases (RCP3PD and RCP8.5), with two in between. The figures for each projection are listed in this table:
Table 1: Projected sea level rise (m) by 2100 for the RCP scenarios. Results presented as median, upper (95% confidence interval) and lower (5% confidence interval) limits, calculated from 2,000,000 model runs. Sea level rise is given relative the period 1980–2000. (Jevrejeva 2011)
Another study (Rahmstorf 2011) obtained much the same results:
Figure 4: Sea level hindcasts and projections for different models calibrated with different temperature and sea level data. The error bars on the right indicate 90% confidence intervals (5–95 percentile, using the GISS temperature dataset); for the proxy-based projection the uncertainty is as presented in Kemp et al., 2011. (Rahmstorf 2011)
What's in the pipeline?
The 'pipeline' is a term used to describe the slow reaction of the oceans to heating (inertia). Even if we were to stop emitting greenhouse gases tomorrow, the oceans would continue to rise, driven by the heat already stored. (90% of all the sun's energy falling on the surface of Earth is absorbed by the oceans as heat). This sea level rise is said to be 'in the pipeline'.
A paper published in PNAS - Levermann 2013 - has found that greenhouse gases emitted today will cause sea levels to rise for several centuries. For every degree of warming, sea levels will rise by more than 2 meters in the next few centuries. The Earth's temperature has already risen 0.8 degrees C over pre-industrial temperatures.
Jevrejeva 2011 also found increased rates of sea level rise, even if emissions were to stabilise at 490 ppm by 2070 following the scenario in RCP3PD (RCP2.6):
Table 2: Projected sea level rise (m) by 2500 for the RCP scenarios. Results presented as median, upper (95% confidence interval) and lower (5% confidence interval) limits, calculated from 2,000,000 runs of the model. Values of sea level rise are given relative the period 1980–2000.
Conclusion
Based on the new mid-range IPCC RCP4.5 scenario - around 650 ppm CO2 and equivalents producing a forcing of approximately 4.5 watts/metre2 - the most likely sea level rise by 2100 is betweem 80cm and 1 metre. Longer term, sea levels will continue to rise even after emissions have been reduced or eliminated.
jja @50:
1) RCP 8.5 is based on the IPCC AR4 A2 scenario (see discussion at the top of page 34).
2) Treating every newly published result that suggests worse outcomes than the consensus as gospel is not science, it is advocacy. Publication in only the first step, and until a result has had a chance to be reviewed by the scientific community at large, and responses published, or the results accepted and cited, they should be regarded as provisional at best. That you are not only seizing on such results but ignoring (equally provisional) results that point in the other direction shows very clearly that it is not the science driving your view point.
3) It is true that the ice and snow albedo feedbacks are larger than expected in the models. The models, however, include many feedbacks with significant uncertainties - and just be chance we would expect some to be under-estimated and some over-estimated. Pointing to one that is under-estimated and concluding from that that the climate sensitivity is under-estimated without out a full survey of the evidence with regard to all feedbacks is, again, not science but advocacy. Afterall, from the nature of uncertainty we expect there to be some under-estimates, and finding one tells us nothing about the overall balance.
4) Measurements of climate sensitivity from paleodata by their nature include all feedbacks. Therefore, the evidence from that source is unaffected by findings regarding the strength of particular feedbacks. If we find one feedback to be stronger, then of necessity the balance of remaining feedbacks must be weaker because the total effect of all feedbacks are already included in the estimate. As an ECS of 3 +/- 1.5 C per doubling of CO2 is a robust result from paleoclimate analysis of ECS, finding an enhanced ice and snow albedo feedback does not make an ECS greater than 4.5 C per doubling of CO2 appreciably more likely.
Given points (3) and (4), I think there can be no doubt that your scenario is a "worst case" scenario, and is not supported by the current balance of scientific evidence (no matter how well supported by a cherry picked sample of that evidence).
5) Contrary to your supposition, the 2110 ZJ (RCP 8.5) result "assumes" a TOA energy imbalance of 0.686 W/m^2 in 2014, and is consonant with Trenberth's result (although notice again your selection of the largest available value as being bedrock). The 3425 ZJ (RCP 10) result "assumes" a TOA energy imbalance of 1.115 W/m^2 for 2014. Between them, therefore, they bracket the Trenberth result, and by your logic the RCP 10 value is an overestimate. Therefore, take that value and you still need to propose a mechanism to raft a volume of ice equal to 44% of the volume of Greenland's Ice Sheet. You refuse to confront this absolute necessity in your scenario, and until you do you don't have a theory worth considering.
6) I note in passing that your maths regarding energy balances is all over the place, and way off. Thus, in your comment @47 you claim my linear estimate based on RCP 8.5 and ECS2xCO2 as assuming a current TOA energy imbalance of 0.44 W/m^2 (it actually assumes 0.6 W/m^2). You further claim that using an current imbalance of 0.75 W/m^2 would result in a mean annual imbalance of 3.65 W/m^2. The correct value is 1.525 W/m^2.
Finally,
7) Your explanation of why ice sheets melted slower coming out of the last glacial amounts to asserting that ice sheets melted slower because ice sheets melted slower (ie, the albedo reduced slower). The reasoning is circular and provides no support for your hypothesis, and gives us no reason to ignore the lessons of the past.
in the link you posted the authors cite the A2r scenario, updated with current developing country projections: "
if you look at the final values of CH4 and CO2 atmospheric concentrations in 2100 you will see that the A2 scenario is approximately 15% below the RCP 8.5 high error margin estimation. You will also find that the A1Fl concentrations match that value much more closely.
2. This is the limitation of science, in a non-linear environment. PIOMAS has been verified by Cryosat-2. This has not been utilized in the AR5. The non-linear response of multiple coupled systems to the loss of arctic sea ice will produce traumatic shifts that are unforetolled by the AR5. The paper I quoted is published in 2008, this is not a "new" paper.
3. Your estimates of climate sensitivity are based primarily on paleoclimate data. However, you do not have a statistically signficant number of sample points for the current climate regime. So, you throw out the anomalous data that indicates that there is a much higher climate sensitivity during interglacials. You justify this because you do not understand why a runaway warming did not occur if sensitivities were actually that much higher during the interglacials.
In addition, the issues that I have brought up are not uncertainties, they are unknowns. The difference is that the unknowns are not contained in the current models. The body of evidence indicates that these unknowns (defined as proesses with impacts that are determined to be real, potentially catastrophic, but currently undefined and so not included in the model) will have effect, and that the potential effects of their cumulative impacts will surpass even the worst case scenarios.
The 3,425 ZJ cumulative energy deposition is your value not mine. I assume a current TOA imbalance of .75 W/m^2 and a 2100 TOA imbalance value of 3.5 W/m^2 (linear and proportional to RCP 8.5). Yeilding an average TOA imbalance of 2.125 W/m^2 between now and 2100, producing the cumulative energy deposition of 2,939 ZJ. I used RCP 8.5 for continuity of discussion, I do believe that RCP 10.0 is more likely given the albedo and carbon cycle feedbacks (as well as emission scenarios) posited in @16.
6. The .44 TOA came from your post@ 43 when you said, "That formula predicts a current TOA energy imbalance of 0.44 W/m^2, and a 2100 imbalance of 1.76"
you did not show your math so I reproduce it here:
1. TOA imbalace will be have constant proportionality to RF between now and 2100 (you say this is conservative)
2. Current TOA is .63, RF is 1.635(to pre-industrial) Ratio is 35.8/100
3. you then used RCP 6.5 for your calculation, I used RCP 8.5
4. 8.5 * 35.8/100 = 3.27 W/m^2
5. If I use a starting TOA of .75 then my values are slightly higher for TOA in 2100 = ~3.5
7. The rate of warming from the last interglacial is approximately 200X slower than current warming rates. This is an undeniable truth. The ice will necessarily melt more slowly. The Milankovich cycles are not GHG driven events. I am sure you know this.
I would ask that future discussions by AR5 reviewers list clearly and succinctly whether or not the following unknowns are contained within their models.
1. The cessation of Anthropogenic SOx emissions in 2070
2. The reduction of the OTC by 80% in 2060 and its effects on North Atlantic carbon sink.
3. The effect of drought on the Amazonian basin by year 2050 on natural carbon sequestration
4. The potential for boreal peat carbon emissions in a +4C environment
5. The step-change in annual surface temperatures produced in the arctic for an ice-free state in the summer.
6. The total emission potential from the ESAS under a regime of consistent +10C water temperatures in late summer.
7. The localized ocean warming effect on the West Antarctic Ice Shelf under a regime of 80% reduction of the OTC.
8. The projections of Western United states drought under a regime of a summer ice-free arctic in 2030.
9. Accelerated Permafrost emissions due to an ice free summer state in 2030.
10. The effects of these non-analyzed carbon cycle and albedo feedback scenarios within an environment where Carbon Capture and Sequestration is NOT utilized by developing countries and ECS for 2XCO2 is actually 4.5C.
The above would be a very useful study. It would also provide a reasonable basis for the collective mea culpa due to the human race (and the Nobel Committee) when the arctic ice DOES disappear in 2030 and catastrophic warming events begin to reveal these unknowns to be real and disasterous.
I'm noticing some people are quoting Jevrejeva et al 2014 as proof that mean sea level rise is nowhere near the often quoted 3 mm/year, and in fact is more like 1.8 mm/yr and decelerating.. Curious what is the story of this recent study?
Knaugle:
From Jevrejeva 2014:
"There is a good agreement between the rate of sea level rise (3.2 ± 0.4 mm·yr− 1) calculated from satellite altimetry and the rate of 3.1 ± 0.6 mm·yr− 1 from tide gauge based reconstruction for the overlapping time period (1993–2009). The new reconstruction suggests a linear trend of 1.9 ± 0.3 mm·yr− 1 during the 20th century, with 1.8 ± 0.5 mm·yr− 1 since 1970."
It appears the skeptics are quoting the sea level rise over the 20th century. Since the sea level rise is accelerating, using this longer time period lowers the reported rate. If you measure the most recent 30 years Jevrejeva agree with the oft reported 3.2 mm/yr from satellites. Skeptics appear to claim deceleration from the 1970 to present rate.
That seems ok if you don't mind overlooking the fact that sea level rise is accelerating. If we average over the past 1000 years the rate will be almost zero! No problem!