So I was looking at the latest sea level data from AVISO in France. A recent publication found that sea level rise is accelerating, at 0.084 ± 0.025 mm/yr2 (In just the satellite era.) I get 0.061 ± 0.007 mm/yr2 when I fit the data to a second-order polynomial (the same method used by Nerem et al, the paper mentioned in the previous sentence.), but I'm sure my error bar (2σ) is too small because I didn't include autocorrelation. Anyway, I get about the same number they do.
The acceleration has been about constant for about two years, but maybe there's a little uptick at the end:
Two comments about this graph:
1) the missing error bars (white gaps) are, I'm pretty sure, due to a bug in Excel, and
2) again, the error bars here are without considering autocorrelation.
(As I've written before, I don't know how to do the calculation of error bars for a 2nd-order (or higher) polynomial in the presence of autocorrelation. If anyone reading this knows, I'd appreciate a comment with more information.)
The acceleration changes relatively slowly, but once it starts changing it takes some time to stop. It has a lot of inertia, you might say. So its slight upturn now will probably continue into larger values,especially since we're entering an El Nino -- see the 2015-2016 period in the graph, with its monster El Nino.
I'm working on doing to 3rd-order polynomial fit. It's not obvious it will be better....
No comments:
Post a Comment