It seems like this is the sort of thing that gets the finance industry a bad name:
A financial product could be constructed with payments linked to a sea-level index, and featuring some characteristics similar to a catastrophe bond or weather derivative...
Protection would come in the form of a higher payment to the policyholder if the sea level rises more quickly than expected and a lower or zero payment if the sea level rises less quickly. Some creativity would be needed to make such a product acceptable to both the policyholder and insurer, but it is quite feasible.
But quite apart from the questionable PR which would result from creating such a product, it probably wouldn't work, as alphaville point out:
In terms of getting the technicals of the product down pat, well why not. But thinking about it – you (as the insurer) would be selling insurance on a potentially massive, truly systemic risk here. Something that could – over time – remove island nations from the map altogether. Not something you can hide from using the law of large numbers, quite possibly.
The chart accompanying the original study shows that 37.2 million people in India alone are at risk from sea-level rises by 2050, and well over 100 million in Asia alone -- as well as another 8 million in the USA. It is likely global catastrophe of that level isn't something the world financial system could escape unscathed, so while the creator of these bonds would make a nice income in the years leading up to disaster -- and an even nicer one if climate change was indeed averted -- anyone expecting that the insurance they had purchased would actually protect them against anything would be in for a nasty surprise. Almost as nasty as the people living in coastal areas.