Friday, November 09, 2007
By Andrea Johnson
A Chicago-based information company, Lanworth, Inc., is using twice-daily satellite images to look at crops.
With their technology and methodology, company officials hope Lanworth becomes a common name for pre-USDA forecasts.
Lanworth estimates the 2007 U.S. corn yield at 13.11 billion bushels and soybeans at 2.58 billion bushels. Their data was released five days before USDA released the official Oct. 12 report.
USDA's Oct. 12 numbers were 13.3 billion bushels for corn and 2.60 billion bushels for soybeans.
USDA's numbers were lower than some market analysts expected.
Nick Kouchoukos, Lanworth director of information services, says the company stands by their numbers. They expect USDA will reduce 2007 U.S. corn and soybean production by at least 200 million bushels each in the November or January reports.
“We didn't expect USDA to bring the (corn) number down until November, but by reducing yield by more than 1 bushel/acre, they are already showing a more conservative estimate,” he said.
Lanworth specializes in extracting information using Remote Sensing, Geographic Information Systems, Economic Modeling and Software Development.
Founded in 2000, the company was formerly known as Forest One. The name fit their focus on timber supply studies, GIS-based land management systems, due diligence on land acquisitions and higher-and-better-use analysis.
Lanworth now offers more services to help government, agriculture, real estate, utility companies and transportation.
The company currently has about 23 employees and maintains offices in Chicago and Jackson, Mississippi. The production staff includes three Ph.D. level analysts, 10 junior analysts and technicians, and five programmers.
“We built our team and our methods and our knowledge in the forestry domain,” said Kouchoukos. “Once we got our confidence and skills up, we were ready for the challenge of agriculture.”
Lanworth uses several types of information to analyze crop data.
The company uses digitized Farm Service Agency (FSA) maps to understand the distribution of farm land at a very detailed level.
In addition to maps showing farmland distribution, FSA provides digitized maps that indicate soil types and fertility. Lanworth then incorporates digital soils data provided by USDA's Natural Resource Conservation Service.
Satellite-derived maps are used to determine acres planted to corn, soybeans, wheat and other crops in the spring.
Lanworth follows weather patterns, and surveys field changes on a daily basis.
“We begin to follow how weather affects the growth of those crops,” he said. “That establishes the basis for monitoring crop progress, and it provides a solid foundation for yield forecasts.”
The satellite data comes from two NASA satellites with polar orbits. The satellites travel in almost north and south orbits and collect images of Earth.
One of the satellites is called Terra and the other is called Aqua. Each satellite collects images of Earth as it makes one revolution each day.
“We settle for coarser resolution imagery that is taken every day,” said Kouchoukos. “What we get in that trade is the ability to see how that field is changing over time.”
By looking at differences between the rates at which fields green up in the spring, or timing on when they reach their peaks, Lanworth can make accurate predictions about crop progress and size.
Kouchoukos acknowledges that USDA remains the definitive source of agricultural information.
USDA has methodology developed over decades. At the end of the year, USDA's numbers will be right on the mark.
“As USDA goes through time, they make successively smaller revisions until they arrive at their final estimate,” he said. “It takes awhile for their sampling and survey strategy to begin to appreciate all the variability that's out there.
“So when you're trying to understand the whole by taking small measurements of the parts - if the sampling strategy is good, you'll ultimately arrive at the right answer. What we try to do with satellites is take in everything with a single glance.”
Their evaluation team includes computer image experts and agriculturists.
“If there is an area where a problem is developing, we not only see that problem, but we can immediately measure how big it is in area,” he said.
Lanworth sees their value in two ways - they can offer data and information that is less than 24 hours old. They also note the variability across regions.
If they see something emerging on the imagery that looks like a problem or an area where yields are especially good, Lanworth sends out a team to take a look in the fields. They also use aerial photography to help decide what they are seeing.
Kouchoukos believes this information has value to feedlots or ethanol plants and biofuel facilities. Lanworth data could help those entities determine if there is enough local feedstock available for their enterprises.
“There's a lot more uncertainty about who's going to plant what and where,” he said. “Our data gets that information out early.
“If you're trying to operate a production facility in an area where you're not sure what your local resources will be, our data can help you understand that early and make arrangements as necessary.”
Traders find value in the accuracy and timeliness of the information.
Kouchoukos also thinks farmers can gain local knowledge by learning about the general trends and plantings in their area.
Lanworth information is sold as subscriptions that could range from entry level prices of $1,500-$2,000 to several thousand dollars.
Real-time market information is available covering corn and soybeans in the United States; wheat in the United States, Australia and Ukraine; and palm oil in Malaysia, Indonesia and Thailand. Corn and soybean data from Brazil will become a standard product in 2008.
“We do want to become a household name, absolutely,” Kouchoukos said.
Farm & Ranch Guide
Copyright 2007 Farm & Ranch Guide