Each spring and summer, the U.S. Department of Agriculture recruits hundreds of experts to examine wheat fields and report estimated yields.

Jude Kastens says he can do the same from the comfort of his office, with the help of a satellite.

Kastens, a researcher with the Kansas University-based Kansas Applied Remote Sensing (KARS) program, is among a team that is perfecting the use of satellites to predict crop yields. He says the satellites are as accurate as traditional methods and operate at a fraction of the cost.

“One person is doing the yield forecast for the whole country,” he said.

The KARS predictions work this way: A satellite 520 miles above Earth reads the amount of chlorophyll  the chemical that makes plants green  for a 16-mile-wide stretch every seven days. Each “pixel” of the picture produced is 250 acres.

A computer then compares the chlorophyll data to data from past years when crop yields are known. Compiling data from more years should continue to make the predictions more accurate, the researcher said.

Though the satellite reads the chlorophyll level of all plants  not just the crops  the general health of all plants is a good indicator for yields, Kastens said.

Early predictions in the mid-1990s were often inaccurate. But this year’s satellite predictions for winter wheat  the first announced publicly by KARS  were so accurate compared to other predictions they even surprised KARS researchers.

For instance, the KARS data forecast on April 22 called for 34.5 bushels per acre in Kansas. A team of more than 50 wheat industry experts surveyed 442 fields in Kansas April 30 through May 1 and predicted the same number.

“It’s shocking it was that close,” said Kevin Price, KARS associate director and geography professor.

Final USDA yield figures won’t be released until the end of September. The most recent projections said Kansas produced 33 bushels per acre, compared to a KARS projection of 33.6.

For the entire United States, the USDA predicted 38.9 bushes per acre, compared to a KARS projection of 37.2.

In addition to the saved money, the satellite method allows researchers to project yields earlier in the season and for a more detailed area.

Winter wheat isn’t the only crop the researchers are predicting. They also predict soybeans, corn, oats, barley, sorghum and spring wheat. And they use plant health to project pheasant populations.

Though Kastens and Price would like the USDA eventually to abandon the traditional field predictions for satellite methods, they realize that’s not likely soon, even though NASA and the USDA provide funding for KARS.

“There’s been a lot of reluctance to adopt this new technology,” Kastens said.

Bill Tierney, an agriculture economist with Kansas State University, said the biggest advantage for satellite predictions was the detail in area they were able to provide. USDA estimates generally include around 15 counties.

“If you’re in the grain business, grain merchandising business or feedlots or flour mills, you want to know what the grain production will be along a rail line or in a 30-mile radius of a feed lot,” he said.

But Tierney said farmers wouldn’t be likely to simply jump to a new system.

“I think what (USDA) would have to do is run them in parallel, so the public that’s being served by these production estimates would get a while to develop confidence in them or not,” he said. “There’s always a concern if you’re relying on a satellite and it goes fritz, you don’t have a crop estimate.”