Land Use Change Study Uses Incorrect Data

Cindy Zimmerman

When you do a study that claims biofuels are resulting in increased land use for agricultural production you should be sure the land you are counting is not actually a subdivision.

The Renewable Fuels Association (RFA) points to a new study published in the academic journal Biomass and Bioenergy that exposes major methodological problems in recent studies paid for by the National Wildlife Federation (NWF), and concluded that U.S. biofuels expansion has not caused a detectable increase in the U.S. food prices.

The paper includes a captivating image showing how satellite tools mistakenly characterized large tracts of urban housing in Lemoore, Calif., as “cropland.”

“The real-world data showed no evidence of food price increases or other lands converting to agriculture because of biofuel,” according to the study, which was conducted by scientists at the University of Idaho and U.S. Department of Agriculture (USDA). The research was funded by the National Institute of Food and Agriculture and USDA Office of the Chief Economist.

The new analysis found that the type of satellite data relied upon by NWF is error-prone, unreliable, and “misleading.” According to the report, “The automated [satellite] land use classification errors were biased towards classifying ambiguous land as agriculture.”

Specifically, the authors manually inspected actual land uses to see if the satellite imagery used by NWF correctly classified the land use. The researchers found that 10.9% of actual non-agricultural land was misclassified as agricultural land by the satellite data. Further, while automated classification using the satellite data showed an 8.53% increase in agricultural land from 2011-2015, the manual classification indicated no significant land use change at all.

“The use of satellite data is prone to error in classifying certain land uses, such as distinguishing between cropland used to grow hay, and pasture land for grazing…Although an automated satellite image classification provides a convenient way to quantify land use change, the results could be misleading if not carefully verified,” the authors explained.

Read the analysis.

Ethanol, Ethanol News, RFA