*Probabilistic sea-level projections have not yet integrated insights from physical ice-sheet models representing mechanisms, such as ice-shelf hydrofracturing and ice-cliff collapse, that can rapidly increase ice-sheet discharge. Here, we link a probabilistic framework for sea- level projections to a small ensemble of Antarctic ice-sheet (AIS) simulations incorporating these physical processes to explore their influence on projections of global-mean sea-level (GMSL) and relative sea-level (RSL) change. Under high greenhouse gas emissions (Repre- sentative Concentration Pathway [RCP] 8.5), these physical processes increase median pro- jected 21st century GMSL rise from ∼80 cm to ∼150 cm. Revised median RSL projections would, without protective measures, by 2100 submerge land currently home to > 79 million people, an increase of ∼25 million people. …*

best fit exponential

y= 2.3976 – 1.52329*(1 – e^(0.08704*x))

R^2 =0.9705

year Sea Level

2010 4.51

2020 9.56

2030 21.62

2040 50.40

2050 1.191m

2060 2.833

previous update

2060 5.48m

quadratic

y = 1.8427 + 0.17066*x + 0.0095649*x^2

R^2 =0.9646

year Sea Level

2010 4.50

2020 9.08

2030 15.57

2040 23.97

2050 34.28

2060 46.51

2070 60.66

2080 76.71

2090 94.68

2100 1.146m

previously

2100 1.33m

I’ve now changed the enso “filter” onto the exponential fit to the Jason2 plot. A bit of a transcribing Jason2 error at year 2010.1 , will be corrected for the next time. Red is piecewise linear sampling of the black Jason2 line, green is the ENSO linear sampling, lilac is the exponential .

Adjusting lag and scaling to minimise the quadrilaterals, processing mathematically rather than graphically.

So minimum sum of quads is now with a lag of .303 years (so green extending into the legend area of the Jason2 plot) and a scaling of the ENSO “strength ” units by 0.18 aplied uniformally.

The stumbling block is around 2012/2013. The next filter will be for global sea-ice, and optimise as well , like the ENSO.

AFAIK seaice is near enough fresh water, ok icebergs are 10% above water as the density change of solid and liquid phase princially. But on melting , is about 2.5% lighter than the salt water ,density 1.025. So for each cubic km that melts about 25m extra height, vol for vol. For the total 360 x10^6 sq km of global ocean surface area, I make that about 0.07mm SLR for each 1000 km^3 unit of global GIOMAS sea-ice determination loss against time.

https://sites.google.com/site/arctischepinguin/home/giomas

Offsetting for the Jason2 2012/3 peak, over and above enso correction, is likely

to suppress the current SLR, unless the global sea-ice values return to the longterm values, ie a new-normal of onging global sea-ice deficit.

Overlaid on the Jason2 plot ( blue and black) is the (red and green insead of blue) ENSO plot for the same period (used previously on these postings) , delayed .33 year or 4 months.

The straight blue line is the 24 degree .443 gradient line of the Jason plot.

I’ve differentially reduced the red side of the ENSO image by 78% relative to the blue, so the upper diagonal black lines are closer spaced than the lower , here changed to green, ones.

The Enso plot is angled at 28 degree and skewed 28 degree, horizontally scaled to match the time period and vertical scale changed along with hovering over, until best visual maximal conjuction of both plots .

If there should be any validity to this graphical playing around, I make the current Jason2 level 0.6m higher than it should be , with ENSO “correction”, considering not much left of the 6-month filter to drop it much farther. Fast global warming , end of last year, as shown by the loss of global sea-ice ?

The double grey vertical lines are the 2010 and 2015 lines on the enso plot, after edge detection to remove the red and blue solid colours. Any suggestions as to how to improve the correlation?

“Black” 6month filter data, x = year -2000

08 Feb 2017, x = 17.107,y= 7.54cm as per Aviso

29 Jan , x= 17.079 , y= 7.55cm

29 dec 2016 now y= 7.45cm

best fit exponential

y = 2.45 – (1.3832)*(1 – e^(+0.09121*x))

R^2 =.9712

year Sea Level rise

2060 3.30m

off the planet for 2100

previous update gave

2060 5.48m

best quadratic fit

y = 2.1459 + 0.1146*x + 0.011897*x^2

R^2 = .9646

year Sea Level rise

2100 1.33m

previously

2100 1.67m

1900 – 1929, 30 years, 0,6mm/yr

1930 – 1992, 63 years, 1,4mm/yr

1993 – 2016. 24 years, 3,4mm/yr

That could give some more meaning than calculating on very short timespans. Even if the numbers are very questionable, and the interpretation related to real change would be difficult. ]]>

y=ln((0.0045936* e^(0.5429 x)) + 1) + 3.75396

R^2 = 0. 9716

gives

year Sea Level cm

2010 4.47

2020 9.23

2030 14.66

2040 20.09

2050 25.51

2060 30.94

2070 36.37

2080 41.80

2090 47.23

2100 52.66

so near enough linear as close to log of anti-log of x, one for the skeptics, just 8cm more than the 4.44mm/yr linear over the century. But valid in maximising R^2 terms, better than the best quadratic fit, but not the exponential

and

y=-1.006213* x^3/2 + 12.17169 x – 45.1048 x^1/2 + 57.1738

R^2= 0.97184

year Sea Level

2010 4.44cm

2020 8.89

2030 9.94

2040 4.22

2050 -8.93

2060 -29.55

2070 -57.48

2080 -92.51

2090 -134.39

2100 -182.92

so scrub that one

then this one

0.5918733*x – 2.5891 + 17.682267*( e^(x*-0.274789 ))

R^2=0.9714, also better than the quadratic best and compound of linear and exponential

year Sea Level

2010 4.46cm

2020 9.32

2030 15.17

2040 21.09

2050 27.00

2060 32.92

2070 38.84

2080 44.76

2090 50.68

2100 56.60

tried exponential+quadratic compound function optimising on R^2 , giving

y= 2.9851 +e^(-0.23945*x) +0.01877*x^2 -0.0461305*x

and R^2 =0.9702

so better R^2 than the best fit simple quadratic so far, still below the .973 R^2 of best fit exponential, well above best linear R^2 of .9621 , exponential part vanishingly small later in the century.

Moved on to an Enso filter, a tidied up version of the previous graphical floating overlay. This time tabulated the differences in y for Jason-2 peaks and troughs from the best fit exponential curve for same x.

Graphed them out and formed a transparent copy to float the overlay with scaling, over the bit of the Enso-Index plot of 2008 to 2017.

Obviously subjective here, moved around for best correlation to the peaks and troughs of the ENSO plot.

Jason2 out of sync by .33year or 4months later than Enso and the y=0

of the enso plot shifted up 1 unit, so red values are less by 1 unit and blue are

larger by 1 unit. The conversion became Jason2 cm compensation = 0.28 * enso units.

Then made the opposite sign of those peak and trough correlations to the Jason plot, giving the green squares on

the 2017.13 point is .37cm over the guessed at Jason-2 level not yet in the public realm.

No particularly good fit of curve to that

eg quadratic, R^2 =0.9573

y = 6.030854 – 0.4062662*x + 0.02983813*x^2

year Sea Level

2010 4.95cm

2020 9.84

2030 20.7

2040 37.5

2050 60.3

2060 89.1

2070 1.24m

2080 1.64

2090 2.11

2100 2.64

best linear fit

y = .358034* x + 1.36104

R^2 = 0.9176

lower slope of 3.5mm/yr for the skeptics, but no sign of the Saral downturn of end of 2016/start 2017. Will try adding the (new +1) zero crossings of the Enso plot as well, to that Jason2+enso filter plot, for as-is points on the plot

Next is to try unravelling this

http://sealevel.colorado.edu/content/2016rel3-gmsl-and-multivariate-enso-index

Enso correlation plot for the earlier Jason , to give cm rather than SDs and use that on Jason2. Or repeat using Aviso-Jason1 and Enso for another stab at a correlation regime.

Here’s a reference to the age of the Hans Tausen plateau ice cap that isn’t grey literature: http://www.nbi.ku.dk/english/Calendar/Activities_07/Hans_Tausen_Iskappe/

Here’s the northern Greenland beach ridges reference, also not grey:

http://science.sciencemag.org/content/333/6043/747?sid=6dd66f17-4fe9-4f36-bda4-0b589ea22723

I don’t have access to the full article, so I don’t know if the statement on summer sea ice extent is correct.

The site does appear to be rather uncritical, to put it kindly, but that doesn’t mean you can safely tar every reference quoted with the same brush.

I didn’t see a reference on the sea level being higher. so I can’t verify that.

]]>They are entirely devoid of meaning. Fitted curves are almost never suitable for extrapolation, unless they are based on a verified physical model.

]]>