Trying out small planetary colonization concepts at a terrestrial desert is a "good idea". I believe the Mars Society is (was or will be) doing that in one of the Australian deserts. However I believe there is a conservation principle (as in conservation of enery or mass) that prohibits making the Sahara Desert into a rain forest. This is an area that I am sure Nick Hoffman could comment about expertly. However I'll take my feeble stab at it:

I have heard of attempts by people to come up with good long term climate prediction computer models. These models almost always predict that the Sahara Desert is a "rain forest" to the extreme embarrassment of the people writing the programs. Also when these codes are showing the Sahara Desert as a rain forest, they typically show Europe as being covered with a glacier or other parts of Africa that are normally fertile being covered with deserts. The southern shores of the Mediterranean were actually quite fertile and liveable two thousand years ago, i.e. Carthage, Cyrene, etc.. For that matter, most of Australia was covered with rain forests about 15,000 years ago. It seems to require only a very slight perturbation to change a rain forest into a desert and visa-versa. For years the Egyptians have been quietly advocating flooding large parts of the Sahara with water from the Mediterranean or Nile in the hope of improving their local environment. However the Europeans have been bucking the Egyptians on this because it would change the salinity of the Mediterranean and create environmental havoc. This is the main reason why terraforming Mars is a good idea: You can only make Mars better with no bad environmental side effects.

Gary