Scuba-diving will increase your memory performance. I refer here to a famous experiment conducted in 1975 that examined Tulving and Thompson's encoding specificity principle, which later became what's known as context-dependent memory.
These concepts essentially refer to the fact that you remember things better when the environment you're in more closely matches the environment you learned the things in the first place.
The unorthodox start
Duncan Godden and Alan Baddeley heard of a phenomena in which deep-sea divers were much worse at remembering the events of their dive much on land than when they were in the water. The natural response was to, of course, send people to go diving and test their memory with a list of words.
Godden and Baddeley found that that if you learned the list of words underwater, you recalled it better underwater. If you learned the list on land, you recalled it better on land.
The external setting matters
This particular effect has since been explored in a number of settings. For example, if you learn in a noisy environment, you'll recall better in a similarly noisy environment (but too noisy and you're back to recalling worse).
The memory task and the recall task should match
In fact, it even matters how similar the recall task is to the memory task. If you learned a list of words, then you should write out that list to recall it. If you're picking up unsorted flashcards instead, you're going to remember far fewer words. This is called transfer-appropriate learning.
The internal setting matters too (e.g your emotions)
Your internal environment is also important. What I mean by that is, if the way you feel matches from when you learned something to when you're trying to recall it, you'll do better. It's called state-dependent learning.
So. If you want to boost your memory, try to match as closely the way you learn to the way you'll be expected to remember.
It's all part of the way our brain operates. We can only (take in so much new information)[/articles/memory-basics/]. And our memories are (liable to distortion)[/articles/endowment-effect/]. By making sure the brain has as much information about the memory as possible, you're more likely to get it activating the right sets of neurons.
Proust once wrote:
Remembrance of things past is not necessarily the remembrance of things as they were
But at least remembrance of things past can be way better if you set things up the same.