Every continent has a dry strip between about 15º and 35º on its west coast, but they vary drastically in size:

  • The Baja/Mojave and Atacama are fairly narrow, stretching about 300 miles inland at the widest (using the "25 cm annual precip" definition of desert.
  • The Namib, about 600 miles.
  • The Great Australian Desert covers about 2,000 miles, most of the continent.
  • The Sahara, Arabian Desert, Dasht-e-Lut, and Baluchistan together stretch over 5,500!

These are all caused by the winds of the horse latitudes, in turn caused by the continental high-pressure zones, AKA subtropical ridges. But why does this produce just a slim strip of desert on some continents but a massive span southwest of Eurasia?

The size of the continent is the obvious answer, but if so why is Australia almost all desert, despite being the smallest continent, surrounded by warm ocean? Looking at the Andes, the Rocky Mountains, the mountains of Iran, Angola, and Namibia, and the Great Dividing Range, it appears that each subtropical desert tends to end at a mountain range, but not always: the southern Namib climbs half a mile before stretching across half of South Africa, and the Sahara has several highlands.

  • $\begingroup$ Consider that a lot of North Africa, the Middle East, and even some of Australia wasn't actually desert until after humans had worked on it for a good many years. $\endgroup$
    – jamesqf
    Jun 20 '15 at 17:42

I think the size of the continent is a big contributor. If there's very little cloud cover, and little rain, the air just keeps drying out the ground. Try finding a good map of surface or lower-level winds, and just compare how long it takes an air parcel to go from over ocean to the Sahara. Sometimes the winds will come straight from the oceans, but often they come from over land: the Near East and Asia.

But yes, all your examples end at mountain ranges. Or rather, start there. At the horse latitudes and on the equator-ward edges of the subtropical ridges, winds mostly blow east to west, not west to east. A number of these deserts are thus in rain shadows of a sort (https://en.wikipedia.org/wiki/Rain_shadow). This kinda ties in with the above: Since the wind blows east to west, the air that blows over the Sahara hasn't just been over Africa but also much of Asia, drying out along the way.

My explanation for Australia would be this. It doesn't have major mountain chains, so it's not in a rain shadow. But it doesn't have mountains, so there's nothing to initiate orographic precipitation. It's convective precipitation or nothing, once you get inland. And since the ground is dry, convective precipitation doesn't have much of a water source, so it doesn't rain much and the ground stays dry. Someone with more background on Australia should chime in on that, though.

  • $\begingroup$ +1, thank you! I'm used to thinking of wind as primarily west-to-east, but of course you are correct that's reversed at certain latitudes. Note that Australia does have a modest mountain range along the east coast, perhaps that's just enough to draw out most of the moisture coming west from the Pacific Ocean. So the relatively broad rainy parts of the Americas at these latitudes exist because there are no substantial mountain ranges to the east of them? $\endgroup$
    – user331
    Jun 30 '15 at 3:19
  • $\begingroup$ That's what I think is happening. There's also the South American monsoon, though; wind blowing towards the southwest, in from the Atlantic, in southern hemisphere summer. My guess is that if there were something that forced that wind to blow straight west, the southern Amazon would be a desert. Unfortunately, I don't know of any papers on the subject... $\endgroup$ Jun 30 '15 at 14:53