Je vais juste recopier le commentaire de Kral Friedrich :
I have gone digging up some papers in addition to the one cited in the post above on the issue of Ramsar:

Hendry and others, Human exposure to high natural background radiation: what can it teach us about radiation risks?, Journal of Radiological Protection 2009, http://is.gd/LPiUlf

They confirm the point made by Mark Lynas above that there are no “proper epidemiological studies”. The problem with Ramsar is that the population is so small that an 0.5% risk increase would not show up with any statistical significance.

Mosavi-Jarrahi and others, Mortality and morbidity from cancer in the population exposed to high level of natural radiation area in Ramsar, Iran, International Congress Series 2005, http://is.gd/A6CS0r.

They have looked at the cancer mortality data and found actually less risk in the higher level areas, but with so small absolute numbers that there is no way to “falsify” LNT with that.

It does falsify LNT for the idea that 15 Sv over a lifetime should lead to 75% cancer rate though, as pointed out above by ColinG.