I read an interesting study by Berman, Jonides and Kaplan (2008) from the University of Michigan. They tested the effect nature walks (and nature photos) on cognitive function.
In their first study participants were given a 35 minute cognitively taxing task- repeat loads of random numbers back to the experimenter in reverse order. After this cognitive exercise the participants were sent out for a walk. The first group went to an arboretum and the second down a busy city street. Both the groups took the test again.
The result indicated that the test improved by almost 20% after wandering amongst the trees. By comparison those subjected to a busy street did not reliably improve on the test.
In the second study participants did not leave the lab but instead some stared at pictures of natural scenes while others looked at urban environments. The group staring at the nature photos beat the chaps staring at the roads and lampposts.