Main question: Is there a need? Who are the target audiences? Already such sites? Of course, interrelated.
Sites: many ocean-related sites, some elaborate with lots of content, none really a Web 2.0, allowing user provision, aggregated across source types, providing toolkits, images, for different stakeholder. See Julit Robinson's, Courtney Wong's analyses.
"Lack of presence doesn't mean there IS a need."
Audiences (not good term -- user/provider): need market research; review prior environmental/ocean surveys (see Juliet Carlisle's dissertation, and Cia Bates's literature review, and Juliet Carlisle's DO summary of educational institutions).
Basic need: ocean in trouble with major negative consequences for humans, animals, plants, organisms, air, water, life; thus need awareness/engagement/education; through media; potential short-term and long-term impact.
Audience/needs analysis process approach:
Three dimensions:
1. User/Providers (main categories - public, filterers/creat0rs, scientists; detailed instances: policy-makers, educators, media, journalists, parents [kids-future], scientists, tourists, data creators, media creators, entertainment producers, kids, researchers, oecan businesses, environmental advocates, ocean enthusiasts, students [different types], coastal communities, industry leaders/opinion leaders...)
2. Motivations, Connections to ocean (main categories - economic, use, enjoyment, threat, incentives to use DO itself; detailed instances: economic, recreation, analysis, spiritual, nutritional, ecological, astronomical, content providing, artistic, threats, extraction, incentives)
3. Phase of DO development, OR level of DO components (i.e., from infrastructure through users/providers)

Then: collect information on main and detailed type of user/provider, their level of main and detailed connection/motivation, and phase/level. This then is a detailed matrix. You can present all this in its most simpler form as a 3 user/provider type by 4 connection/motivation matrix, without regard to develoment or component level. This could also be one visual site map approach, where someone could rate how much they are in a user/provider category, and their level of connection/motivation, and that relevant area of DO would become available.

As to audience analysis and need identification, this matrix, aggregated, would also identify where the most action and strength of relations are. Those might be good initial areas for emphasis, publicity, site development. These could then be broken out by development phase/component level. A market/audience/business plan analysis would collect data on each of these intersections.