Where sea-level rise isn’t what it seems

A very nice lesson from Mark Lynas on how tricky it is to interpret sea level data correctly; an excellent example of the challenge of finding reliable signal amongst the noise. This is similar to the media urge to trumpet extreme weather events as evidence of climate change.

Whilst working for the Maldives government I was always aware of the need to resist the temptation of making sweeping statements about the impacts of climate change and sea-level rise in the service of wider political ends. I saw part of my role as advisor to push back against the simplistic view that given that we know that the planet is warming, and the seas are rising, surely the impacts – in terms of erosion, flooding events and disasters – should increasingly be visible now, right?

A new paper published in the AGU’s house journal Eos Transactions shows why caution is often justified. Here (via a screengrab, as the entire thing is behind a password) is the 1993-2011 sea level trend data from Tarawa atoll, part of Kiribati in the central Pacific:

Whoa! No sea-level rise there, then. And yet of course climate campaigners – and even the Kiribati government – understandably anxious to highlight the future existential threat to the islands, have used storm surges, flooding events and suchlike as evidence of current sea-level rise impacts. Which they are almost certainly not, at least not in Tarawa atoll anyway.

To me the graph is interesting for two reasons. The first is the absence of any trend over the last 20 years towards increased sea levels in that part of the Pacific. This should be expected, because sea level rise as a computed average means that the oceans are rising in more places than they are falling, but they are falling in some places nonetheless. (Just as a few areas of the globe have got colder over recent years.) The second is the sheer up-and-down massive variability in actual sea levels, which is linked to the El Niño cycle. The author (Simon Donner, a geographer from the University of British Columbia, Canada) points out in the Eos paper that the monthly mean sea level dropped by nearly half a metre (45cm) between March 1997 and February 1998 because of switch from El Niño to La Niña conditions, and peaks of 15cm were seen in each of the recent El Niño events – which as the author points out is “equivalent to 50 years of global sea level rise at the rate observed since 2000 of 3 mm per year”.

So the problem with attributing sea-level rise impacts is the same as with attributing heat-waves, droughts, floods or other extreme events to climate change – you have to try to figure out what would have happened absent the global warming trend (in order to distinguish genuine impacts from noise), and also distinguish background changes from more direct anthropogenic interference which might confuse the picture. In a heatwave, for instance, were the extreme temperatures caused by the urban heat island effect in a more built-up area?

Read the whole thing »

Good work Mark, and thanks for the quotations from the fire walled and overpriced academic journals.