what i"m not so convinced by is the unpredictable nature of ice. what''s to stop an ice island from calving? or forming large fissures causing buildings built atop the ice to be torn apart? as JP mentioned pykrete could be applied in this situation, providing an ice island a homogeneous additive could give it the necessary strength to remain as a whole.

something i feel that can be capitalized on, which seems mundane at first, is the tether that is holding a given ice island to the see floor. in many caribbean islands (i realize the climate is different there versus that of an ice island, but…) buoys tethered to the ocean floor with a rope are placed in 1000"+ deep water. after a few months particulate matter in the ocean begins to collect on the rope and algae begins to grow. the algae feeds smaller fish, large fish feed on the small fish and so on. after 6 months to a year a fully fledged ecosystem forms that can sustain recreational fishing for 3-4 years. eventually the rope falls apart and the ecosystem disappears. i wonder if something similar would be possible in an arctic environment. transient ice islands that function as commercial fisheries moving every 3 to 4 years once their tether as dissolved.