Good article. A big factor that it seems you did not cover is the early spring rains at altitude in regions that depend on snowmelt for irrigation.

One of the most catastrophic events in these regions is a warm spring with rain in the mountains that melts the snow. This leads to devastating floods followed by drought later in the summer. It seems to me looking at your graphs that these events may become the norm not the exception over time.

Thanks. You are correct about warm spring rains on snow producing floods. With irrigation directly from streams, mid summer flow is dependent on contributions from above around 5000 ft, depending on latitude, if I recall correctly. Irrigation from reservoirs can fill and capture some of the flow, but it is much better to "store" that water on the ground, rather than having to pass it through the system.

I've also heard that both less snowpack accumulation overall and earlier melt from higher than normal temperatures (before irrigation season) are also other aspects to this problem;
http://www.news.cornell.edu/releases/Jan97/water.hrs.html
http://www.usatoday.com/weather/climate/globalwarming/2008-02-01-water-c...
http://ccc.atmos.colostate.edu/newsapr1.php

The essay models above predict precipitation, they don't specify form. I don't think we'll see near the pack in the future either.

We've been very fortunate this year, after years of precipitation deficits, to get the snow we have. For lower elevations, the form difference has been often just a few degrees. The storms come in from the Pacific, moist and relatively warm.