Will Satellite Mean Sea Levels Continue to Rise?
Post on 14-Apr-2018
218 Views
Preview:
Transcript
8/2/2019 Will Satellite Mean Sea Levels Continue to Rise?
http://slidepdf.com/reader/full/will-satellite-mean-sea-levels-continue-to-rise 1/7
Will Satellite Mean Sea Levels Continue to Rise?
Mean sea level changes have become another hallmark of the Anthropogenic Global Warming
(AGW) hypothesis. Recently attention has turned away from the traditional sea level gauges tosatellite altimeter measurements. These show a much greater increase in mean sea level than
terrestrial gauges.
This brief study takes a look at the new and developing global record from satellites. Is global mean
sea level change following a predictable trend and will it continue upward?
The University of Colorado at Boulder provides four datasets for Jason/Topex (J/T) satellite mean sea
level variations which differ in the degree of correction applied. All four, including pre plotted
charts, are available here: http://sealevel.colorado.edu/results.php
In this analysis, just one dataset will be used. It is the fully corrected record, with inverse barometerapplied and seasonal influences removed (http://sealevel.colorado.edu/current/sl_ib_ns_global.txt).
First a plot of the data is essential:
A local polynomial regression smoothing fit (Gaussian kernel) has been applied to reveal the trend.This smoothing is used throughout this document.
8/2/2019 Will Satellite Mean Sea Levels Continue to Rise?
http://slidepdf.com/reader/full/will-satellite-mean-sea-levels-continue-to-rise 2/7
There is a strong rising trend that is visually close to linear. Adding a linear fit to the data produces
the following plot:
There is a good agreement with a straight line represented by the equation:
Y = 6098+3.051*X, R2
= 0.9338. Thus the mean rate of rise in sea level is 3.05mm/year.
However, looking at the data it would appear that a different trend might fit the data; a sinusoidal
function. The next plot applies a sine fit to the data:
8/2/2019 Will Satellite Mean Sea Levels Continue to Rise?
http://slidepdf.com/reader/full/will-satellite-mean-sea-levels-continue-to-rise 3/7
A sine fit to the data yields a slightly better value for R2
= 0.9435 and it is described by the equation:
Y = 26.11*SIN((X*0.1393)-259.6)+4.453
The period of the sine is about 45 years.
R2
does not always reveal the fit accuracy particularly well so plots of the residual errors betweenmeasured values and the models were created, as below:
Looking at this graph, it is clear that the sine model describes the data significantly better than the
linear version most of the time.
Could it be that mean sea level rise is at least partially cyclic? If it was then that might help explain
the apparent difference between the significantly lower linear trends of long records from tide
gauges and the short satellite dataset. This divergence is clearly shown in an interestingpresentation from Laury Miller et al of NOAA, on page 2 (see below):
http://www.oco.noaa.gov/meetings/OCOSR/07/presentations/tuesday/NESDIS_4_Miller.ppt
It is also available here: http://www.docstoc.com/docs/5441583/NOAA-Lab-for-Satellite-Altimetry-
Global-Sea-Level-Rise/
* To understand how a cyclic variation can affect the apparent linear trends, please see the following
link: http://wattsupwiththat.com/2010/04/12/the-new-math-ipcc-version/#more-18382
8/2/2019 Will Satellite Mean Sea Levels Continue to Rise?
http://slidepdf.com/reader/full/will-satellite-mean-sea-levels-continue-to-rise 4/7
Are there any other indications that mean sea level trend may not be linear?
Let us take a look at page 8 of that same presentation, as shown below:
8/2/2019 Will Satellite Mean Sea Levels Continue to Rise?
http://slidepdf.com/reader/full/will-satellite-mean-sea-levels-continue-to-rise 5/7
Taking the first graph and comparing it to my sine-fit by aligning both axes, it is clear that there are
some considerable differences as can be seen below:
Large bias errors of several centimetres are not unknown between satellite records as is shown on
page 5 of the Miller presentation (reproduced below). Miller et al show a bias error of 140mm
between Topex and Jason. Thus it is conceivable that a bias of just 20mm could easily be presentbetween GeoSat and the bias corrected J/T data.
8/2/2019 Will Satellite Mean Sea Levels Continue to Rise?
http://slidepdf.com/reader/full/will-satellite-mean-sea-levels-continue-to-rise 6/7
Shifting the two records to account for 20mm bias error, as shown below, would therefore seem
reasonable. Doing such suggests that the GeoSat data may in fact be in agreement with J/T and
following a cyclic pattern similar to the sinusoidal trend fitted here. Is this just coincidence?
Miller et al, however, beleive that the GeoSat data is in error and have corrected it, turning a
negative slope into a positive one that consequently appears to argee with a linear trend fitted
through the J/T record as shown below:
Is this manipulation really valid or is it based on a preconceived notion that the mean sea level trendshould be linear?
8/2/2019 Will Satellite Mean Sea Levels Continue to Rise?
http://slidepdf.com/reader/full/will-satellite-mean-sea-levels-continue-to-rise 7/7
Just for completeness, the Miller plot is overlaid with the latest data plotted with the sinusoidal fit:
Of course, whilst a simple sine model is reasonable for this short period of time, it is unlikely to fit
adequately over longer time frames. However, mean sea level could still have a signifiant cycliccomponent but with a more complex form.
Conclusion:
This analysis demonstrates that a sine fit to the Jason/Topex satellite mean sea level record is
presently a better representation than a linear fit. As such it is suggested that the assumption of
linear mean sea level trends may not be valid on decadal and longer time scales. Global mean sea
levels could have a cyclical nature possibly with a period of about 45 years. If this is the case, then
with or without an underlying linear or quasi-linear (over the record period) trend, this cyclicity
might explain the discrepancy between long term tide gauge measurements and satellite records.
Will mean sea levels continue their upward linear trend or begin to flatten and possibly fall?
Place your bets. Only time will tell.
Jonathan Drake 16 May 2010
Questioning Climate
Revision: Typo correction 24/5/2010
top related