That figure's crude and somewhat misleading. Most of these "species" are listed as "new" merely because they don't fit into existing taxonomic keys, and the regional experts are stumped. In a survey report, such animals might be referred to as something like...
Station 3633A: Owenia sp. 38
True species status requires an extraordinarily vigorous literature search and organism workup. Often this process takes years. During this time many of these "species" turn out to be damaged, juvenile, different sexed, or otherwise altered versions of pre-existing species. After all this filtering, your 13,000 figure drops to something far less, I believe <1,000 is what's being reported. That's still a bloody lot of critters, however.
I've got an unidentified deepwater anemone in my lab that's probably a new species. It can therefore be tentatively called "new". If I ever get an anemone taxonomist to spend months to years analyzing it, it may prove in time to be a new species. This is how the system works.