When writing the IGOS Cryosphere Theme Report, we were confronted with the painful decision on how to spell “sea ice”; i.e., “sea-ice” or “sea ice”. Strictly speaking, it should be spelled without hyphenation when it’s a noun, and hyphenated when it is an adjective. The problem is that it is usually an adjective, but it’s use as an adjective is so common (e.g., sea ice extent), that hyphens just get in the way. The purpose of hyphenation is to remove ambiguity, yet we always think of “sea ice” as a single entity and we are never confused as to which of the two words is modifying the noun that follows. For example, consider an “old-furniture salesman” versus an “old furniture salesman”. The former sells old furniture, while the later is (probably) and old salesman that sells furniture. But in “sea ice measurements” we are not confused because ice measurements that are in the sea are the same as measurements of sea ice.
Furthermore, when we refer to characteristics of sea ice (or sea-ice characteristics) such as sea ice motion or sea ice thickness, we’re always talking about a single entity, so that the three words really are a compound noun without ambiguity. One approach would be to not hyphenate these common compound nouns:
sea ice motion, sea ice concentration, sea ice thickness, etc.,
but to hyphenate the less common compounds:
sea-ice mapping, sea-ice scientists, sea-ice models, sea-ice analysis, sea-ice services, sea-ice studies, sea-ice observations
Another approach is to never hyphenate “sea ice”. A third approach is to always hyphenate it when used as an adjective. Note that whatever we do with sea ice, we should also do with lake ice and river ice. In the end, we went with the first approach, not hyphenating when used in a common compound noun and hyphenating it otherwise. It was not a unanimous decision!
It would be a lot easier if “sea ice” was a single word (“seaice”), but we didn’t want to go that far.