Isn't the Artic a floating ice mass? If it melts, how would that raise the ocean levels? (ice water in a glass - ice cubes melt, water level actually drops, as water expands when frozen). If they were talking about Antartic ice, I'd understand (that's sitting on a land mass, not in the ocean already). What obvious thing am I missing?

Well first of all doesn't the Arctic ice sheet actually create quite a lump *above* sea-level? When what you see above sea-level, higher temperatures (i.e. melting) and gravity interact you'll end up with rising sea-levels. But in any case whatever ends up in the ocean, be it from the Arctic, Antarctic or simply snow-capped mountain ranges, ends up raising the sea-levels globally. That's just the way liquids work.

Ben's right on the first point. Floating ice displaces a volume of water equal to its total mass, so when it melts its density increases and it ends up taking up exactly the same amount of space as the water it displaced. But the article's just lousy. The hypothesized sea level rise from global warming isn't from Arctic ice, it's from the Greenland ice sheet and Antarctic ice primarily.

One big concern for people in northern Europe is how rising temperatures might mess with the Gulf Stream. Without the Gulf Stream, we'd be in for quite a chill.

I've heard it suggested that the thermal expansion of sea water may have as much of an effect on sea levels as melting ice. There just isn't that much ice, relative to the amount of ocean. I remember having to explain to someone who saw "Waterworld" just how completely unrealistic the premise was.

Of course, if the gulf stream shuts down, that could change: sea levels could drop, as they have in the past under what may be similar circumstances. Not that St. Petersburg would be any better off...