First, Android isn't really growing at all in the USA, at least at the big two operators. ('Other smart' is almost all Android now). All the growth is coming from iPhone.
Second, there's near-zero seasonality in Android phone sales. People decide they want a phone and go out and buy whatever's in the shop at the time that looks good. Launches of 'hero' Android phones appear to have no impact at all - they may take share from other Androids, but not from iPhone and they don't increase overall sales.
Third, there seems to be a complete disconnect between Android and iPhone purchasing. One can understand iPhone sales per se going up in a launch quarter, but why don't Android sales go down in those quarters? It looks like a new iPhone launch doesn't tempt in Android buyers at all.
The implication is that there is an ongoing base of sales that goes to Android, and to some extent iPhone as well, that totally ignores product launches, and just buys a phone. Then, there's a base of people who wait to buy the new iPhone (and of course come off their 24m contact in another launch quarter, eager to buy). And this latter base is getting bigger every year, and indeed driving all of the growth.
What isn't shown in this chart, of course, is churn within the Android base: Android users moving to iPhone while new non-smart buyers shift to Android, keeping the sales steady. That's a bit more work.
AT&T and Verizon WIreless had about 65% of US smartphone sales in Q4 2012. They're the only operators (anywhere) that reported this data. Sprint only began reporting smartphone sales in Q4, and T-Mobile USA hadn't started selling the iPhone: it had had about 9% of smartphone sales, almost all Android and growing slowly
This particular split is very US-specific: the US market's pricing structure tends to conceal the price premium of the iPhone. A chart for China, assuming one could get the data, would show very strong Android growth (though also strong iPhone growth)