Minding Ag's Business

Do Weather Models = Yield?

Can Silicon Valley statisticians predict crop yields down to field level, using high tech models alone? That's been the premise of Climate Corporation's unique Total Weather Insurance policies since 2006 and a big motivation for Monsanto's impressive $930 million purchase of the agricultural insurer Nov. 1.

At a press conference in San Francisco last week, Climate Corporation executives stressed that Google-like algorithms can not only model weather, but personalize agronomic advice and yield forecasts down to millions of individual fields nationwide. It will be launching a crop advisory service charging $15/acre for corn and $7.50/acre for soybeans starting in 2014. Monsanto believes Climate Corporation's expertise will compliment its FieldScripts precision farming service.

Technology is definitely advancing at warp speed, but some land grant academics, crop insurance agents and farmers remain skeptical that today's weather models accurately equate to yield, especially on a site-specific locations. Some drought-prone Great Plains farmers cite flaws in the company's crop insurance products. Statisticians say winter wheat is one of the hardest crops to model, as it has amazing ability to rebound from extreme weather. Others say yields for corn aren't just one event, but multiple events that can vary by geography, soil types and farming practices. Mississippi farmers raise soybeans light years differently now than they did 17 years ago, so historic weather-yield relationships need to account for technology change.

"If you're measuring precipitation, does it come in equal amounts or a 5" deluge at the end of the season?" says University of Nebraska's Brad Lubben, director of the North Central Risk Management Education Center. Weather and yield models are proxies for reality, but they're not perfect, he adds.

Using weather data to forecast crop yields "is the fundamental question ag economists want to figure out," says Rod Dejesus, an economist at North Carolina State University who studies weather and yield forecasts for crop insurance purposes. While he hasn't seen Climate Corporation's models specifically, he believes the science is still a work in progress. "It's harder than people think. There's lots of debate on it in this in the literature, and no resolution," he says.

Crop insurance policies on Great Plains farms still show some imperfections in Climate Corporation's models.

McCook, Neb., dryland farmer Jason Boner had plenty of dead corn in his fields this season, but little to show for his Climate Corporation TWI policies in 2013. His part of the state remains in extreme drought, as it has been since 2012.

"We had not received any moisture through last winter and we were coming off the historic drought of 2012, so my thinking last spring was any insurance was good insurance," Boner says. "Crops are a gamble out here and with a soil moisture profile of hardly anything, I raised my federal crop insurance to 80% coverage and took out the Climate Corporation policy."

TWI insurance covers specific perils like night-time heat stress or drought, but at the time Boner bought policies for 320 dry land acres, TWI sales literature estimated that his insurance would start to pay on 84 bpa yields and max out at 20 bpa, paying indemnities up to $51,000. Instead, yields on his fields averaged 18 bpa and 20 bpa, but he'll end up owing the company $6,000 in premiums.

"If 18 bpa isn't enough to maximize a claim, what are they insuring against?" he says.

Gerard Pochop, a farmer from Atwood, Kan., also farms in the northwest part of the state listed as in an extreme two-year drought, but he has had a similar experience. An insurance adjuster appraised his spring wheat crop in one field at a half bu. to 1- 1/2 bu. per acre due to freeze and drought before he abandoned it earlier this year, but he just received a bill from Climate Corporation saying he owes the balance of his premium. "This insurance is a lot more expensive than federal crop because it's supposed to fill the deductible on multiperil insurance," he says. "If you have a total disaster, you ought to collect something."

Boner doesn't blame weather "basis" risk for his problems--the mismatch between local weather and what a weather station miles away might record or what Climate Corporation models estimated. "We had a few minor disputes over temperature and rainfall, but nothing that should cause things to be this far off," he says. He just doesn't see the correlation between weather and yield used in their sales literature.

Climate Corporation says it has improved it weather models every year of operation and it no longer relies on weather stations to benchmark weather policies. Instead it is modeling temperature and precipitation to 2.5 mile grids and contends that is more accurate than gauges. In areas where too much moisture is a problem, Climate Corporation now asks if fields are tiled, which can account for higher than expected yields under some conditions. Only a small fractionof policies register complaints, CEO David Friedberg says.

Climate Corporation Vice President Jim Ethington says he had heard complaints from parts of western Nebraska and northwest Kansas this season, but that area yields were highly variable. "That could be due to events not covered by the weather insurance policies," he said, or a mismatch with farming practices. "It is a disconnect, hopefully we want to get to the bottom of it," he says "Across all of our customer base, we think we are doing a pretty good job of it."

In central Minnesota TWI policyholder Mark Nowak agrees his weather insurance is working as advertised. Because it is so expensive, however, Nowak uses it only on high risk fields subject to flooding from a drainage ditch. However, this was the first year in 40 that excessive May moisture prevented him from planting some fields. TWI accurately forecasted his local rainfall, he says.

Meanwhile, Great Plains farmers like Boner speculate the cool but dry summer exposed flaws in Climate Corporations yield forecast. "I took this policy because the drought of 2012 was quite hard on me and my family financially, so as a younger farmer I needed some additional protection. I don't know why this policy didn't work, all I know was my 2013 crop was nearly a failure and I need this policy to come through for me," Boner says.

Kansas farmer Pochop agrees. "With multiperil insurance, you know you're going to get paid if you have a disaster. These TWI policies are like going to Las Vegas. They've got a lot of bugs to work out."

Follow me on Twitter@MarciaZTaylor



To comment, please Log In or Join our Community .