One of the biggest hurdles with wind and solar energy is that both can be highly variable, which has raised doubts about whether either source is reliable enough to provide "base load" energy, the minimum amount of energy that must be available to customers at any given hour of the day.

Jacobson said that the variability can be overcome.

"The most important thing is to combine renewable energy sources into a bundle," he said. "If you combine them as one commodity and use hydroelectric to fill in gaps, it is a lot easier to match demand."

Wind and solar are complementary, Jacobson said, as wind often peaks at night and sunlight peaks during the day. Using hydroelectric power to fill in the gaps, as it does in our current infrastructure, allows demand to be precisely met by supply in most cases. Other renewable sources such as geothermal and tidal power can also be used to supplement the power from wind and solar sources.

"One of the most promising methods of insuring that supply matches demand is using long-distance transmission to connect widely dispersed sites," said Delucchi. Even if conditions are poor for wind or solar energy generation in one area on a given day, a few hundred miles away the winds could be blowing steadily and the sun shining.

"With a system that is 100 percent wind, water and solar, you can't use normal methods for matching supply and demand.  You have to have what people call a supergrid, with long-distance transmission and really good management," he said.

Another method of meeting demand could entail building a bigger renewable-energy infrastructure to match peak hourly demand and use the off-hours excess electricity to produce hydrogen for the industrial and transportation sectors.

Using pricing to control peak demands, a tool that is used today, would also help.

Jacobson and Delucchi assessed whether their plan might run into problems with the amounts of material needed to build all the turbines, solar collectors and other devices.

They found that even materials such as platinum and the rare earth metals, the most obvious potential supply bottlenecks, are available in sufficient amounts. And recycling could effectively extend the supply. 

"For solar cells there are different materials, but there are so many choices that if one becomes short, you can switch," Jacobson said. "Major materials for wind energy are concrete and steel and there is no shortage of those."

Jacobson and Delucchi calculated the number of wind turbines needed to implement their plan, as well as the number of solar plants, rooftop photovoltaic cells, geothermal, hydroelectric, tidal and wave-energy installations.

They found that to power 100 percent of the world for all purposes from wind, water and solar resources, the footprint needed is about 0.4 percent of the world's land (mostly solar footprint) and the spacing between installations is another 0.6 percent of the world's land (mostly wind-turbine spacing), Jacobson said.

One of the criticisms of wind power is that wind farms require large amounts of land, due to the spacing required between the windmills to prevent interference of turbulence from one turbine on another.

"Most of the land between wind turbines is available for other uses, such as pasture or farming," Jacobson said.  "The actual footprint required by wind turbines to power half the world's energy is less than the area of Manhattan." If half the wind farms were located offshore, a single Manhattan would suffice.

Jacobson said that about 1 percent of the wind turbines required are already in place, and a lesser percentage for solar power.

"This really involves a large scale transformation," he said. "It would require an effort comparable to the Apollo moon project or constructing the interstate highway system."

"But it is possible, without even having to go to new technologies," Jacobson said.  "We really need to just decide collectively that this is the direction we want to head as a society."

Jacobson is the director of Stanford's Atmosphere/Energy Program and a senior fellow at Stanford's Woods Institute for the Environment and the Precourt Institute for Energy.