# Sea level rise projections bias

Sea levels, recently updated with 10 new data-points, reinforce the hiatus described as a ‘pothole’ by Josh Willis of NASA’s Jet Propulsion Laboratory, Pasadena, Calif., who says you can blame the pothole on the cycle of El Niño and La Niña in the Pacific:

This temporary transfer of large volumes of water from the oceans to the land surfaces also helps explain the large drop in global mean sea level. But they also expect the global mean sea level to begin climbing again.

Attributing the ‘pothole’ to a La Nina and the transfer of water from the ocean to land in Australia and the Amazon seems dubious, given many land areas experienced reduced rainfall at the same times, as shown above.

A quadratic model of sea-level indicates deceleration is now well-established and highly significant, and if present conditions continue, sea level will peak between 2020 and 2050 between 10mm and 40mm above present levels, and may have stopped rising already.

Reference to a ‘pothole’ in a long-term trend caused by short-term La Nina, while ignoring statistically significant overall deceleration, is another example of bias in climate science.

Call:
lm(formula = y ~ x + I(x^2))

Residuals:
Min 1Q Median 3Q Max
-8.53309 -2.39304 0.03078 2.45396 9.17058

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -2.264e+05 3.517e+04 -6.438 7.40e-10 ***
x 2.230e+02 3.513e+01 6.348 1.21e-09 ***
I(x^2) -5.490e-02 8.772e-03 -6.258 1.98e-09 ***

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 3.448 on 222 degrees of freedom
Multiple R-squared: 0.9617, Adjusted R-squared: 0.9613
F-statistic: 2786 on 2 and 222 DF, p-value: < 2.2e-16

And the code:

figure5<-function() {
x <- time(SL); y <- SL
l=lm(y ~ x+I(x^2))
new <- data.frame(x = 1993:2050)
predict(l, new, se.fit = TRUE)
pred.w.clim <- predict(l, new, interval="confidence")
matplot(new\$x,pred.w.clim,lty=c(1,2,2), type="l", ylab="Sea Level",main="Quadratic Projection of Sea Level Rise",ylim=c(-10,100),lwd=3,col=c(2,2,2),xlab="Year")
lines(SL)
}

## 0 thoughts on “Sea level rise projections bias”

1. Sherro1 says:

Of course, this is one of the critical parameters to watch for the next few years because it’s so easy for John Citizen to comprehend.
I’m trying to get some dislogue going with JPL who have recently issued a press statement that the radius of the earth is now confirmed constant with a one sogma error of 0.2 mm y^1. However, even if this WAS the error, taking the 700,000 years of Vostok Ice core, we have a radius change of 140 m. This would have an effect on day length, daily deltaT (day-night) and all sorts of orbital effects if the change was one of density rather than weight. Early days yet.

• Anonymous says:

David, thanks for the suggestions, I might do that. I am coming to the view that there is now enough evidence to ‘call the top’.

2. Unless there’s some physical connection, modeling a curve with a polynomial generally has no predictive value.  Especially higher order polynomials where the big coefficients take the projection into really wild flights of fancy.

• Anonymous says:

Yeah I thought someone would say that.

1. Both linear and quadratic coefficients are highly significant. You are effectively saying that the trajectory is going to change wildly and diverge outside the confidence limits for some unexplainable reason. This model is assuming things are going to proceed much as they have in the past. I like that assumption. Assuming chaos is not useful and unphysical.

2. This is one quadratic coefficient to capture the reduction in rate of increase. Its not multiple higher order terms that do lead to overfitting. 3. Such models complement physical models. The question is, why is sea level decellerating? Because forcing is declining? Because solar insolation is falling and is the main driver of recent changes?

3. Rkorbz says:

This entire prediction is of course now garbage once you add the 2012 data

• Davids99us says:

The 2012 data agrees well actually.

4. Pingback: wypozyczalnia lawet

5. Pingback: zobacz oferte

6. Pingback: anti aging

7. Pingback: see this website

8. Pingback: zobacz tutaj

9. Pingback: strona

10. Pingback: jak otworzyc sklep internetowy