A new paper by Nauels et al. concludes that even if the U.S. stays in the Paris agreement and all nations make good on their pledges, our greenhouse gas emissions by the year 2030 will be enough to guarantee at least three feet of sea level rise by the year 2300.
Unfortunately some journalists have difficulty with numbers, reporting that we’ll have three feet of sea level rise by the year 2030. Dum dum dum dum dum…
I’ve posted about sea level a lot lately. This is all exploratory of course, but this is what I think I’ve found so far. First, my alignment method which estimates both an offset and a slope adjustment for each station, in order to remove the impact of vertical land movement, seems to be working splendidly. It does leave an ambiguity, namely that the slope of the final result will be the median slope of the input data series, but that’s certainly a tolerable guess. Second, my method for PCA with missing values seems to be working. There are some strange quirks, which I haven’t touched on but they’re there. Still, it identifies groups and relationships which are certainly real.
What I’ve done most recently is use my alignment method to form an estimate for nine large regions, namely: Europe, the North American East, North American West, South America, Africa, South Asia, East Asia (which is mostly Japan), Siberia, and Australia. I’ve made no attempt to area-weight within these regions, something that ought to be done. It’s what I’m considering next.
Then I took the nine regional records and aligned them to form a global estimate. All estimates cover the time span 1900 to the present, and are based on stations with at least 360 monthly values. I’ve also removed individual stations which do not belong in a global ocean estimate, such as those along the St. Lawrence river and stations which show unmistakeable signs of severe earthquake disruption.
Here’s my overall result, aligning the regional records for a “global” estimate:
Here are yearly avearges of same:
What strikes me is the appearance of continuous acceleration since about 1960, in agreement with the result of Dangendorf et al. Here’s how the rate of sea level rise has changed, which is certainly consistent with increasing rate (i.e. acceleration) since 1960:
I also applied the PCA method to the residual series from the various regions, and the first PC reveals one of its weaknesses. It’s the Siberia series, because that is the one with many missing values so it forms its own “group.” Of course that’s true, but it’s useless for identifying physically relevant groups.
The second PC, however, shows the el Niño pattern with remarkable fidelity (PC time series in dark gray, multivariate el Niño index in red):
When we examine the “loadings” we find that the el Niño pattern is stongest on the west of N. America, also noticeable in S. America and the east of N. America, while its opposite is prevalent in Australia, Europe, and S. Asia.
There’s more to do, and there are serious shortcomings to what I’ve done so far. I think my next step will be a “gridded” approach, and once I look at that I’ll decide where to go from there.
I will also remember that, when I’ve got something more substantial, some of you might want details of the data and programs. Please be patient, I’ll try to accomodate.
Thanks to John T. Dayton Jr. for a very kind donation to the blog. If you’d like to help, please visit the donation link below.
This blog is made possible by readers like you; join others by donating at My Wee Dragon.