There is no part of biofuel greenhouse gas (GHG) emission analysis that is as complex and uncertain as Land Use Change (LUC) GHG emissions. It has been an active two years for biofuel LUC modelers. In late February 2022, the U.S. Environmental Protection Agency’s (EPA) Office of Air Quality and Transportation, in consultation with the Department of Agriculture (USDA) and the Department of Energy (DoE), conducted a virtual public workshop on biofuel GHG modeling.
The workshop was dominated by discussions of the complexity and uncertainty of LUC GHG emissions. Dozens of experts testified and presented data and arguments that estimated a wide range of LUC GHG penalties for biofuels. Some attempted to make the case that there have been millions of acres of sensitive lands converted to cropland and that soil carbon losses after that conversion were large. Others, such as Dr. Stephen Ogle, Colorado State University/National Resources Ecology Laboratory, who is the lead compiler of soil carbon and nitrous oxide emissions data from agricultural lands for the U.S. EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks, stated that U.S. cropland soil carbon stocks are increasing consistently over time, and they exceed the soil carbon losses from any land use change that has occurred. “In 2021, cropland soils were estimated to sequester 51.8 million metric tons of CO2, primarily due to reduced and no-tillage practice adoption in the U.S.”
After reviewing comments made during the workshop, as well as comments submitted during the subsequent public comment period, the EPA decided to conduct a biofuel GHG model comparison exercise to advance the scientific understanding of available models. On June 21, 2023, the EPA published the results of the model comparison exercise. Five models; ADAGE, GCAM, GLOBIOM, GTAP and GREET CCLUB, were used to compare two biofuels: corn starch ethanol and soybean oil biodiesel.
Tables of results for Land Use Change GHG Emissions:
Notice the extremely wide range of LUC GHG emission results estimated by these well-known models and modelers. How can very smart modelers arrive at such a wide range of conclusions? Two factors: 1. Extreme variation in the estimated area of land that was converted, and 2. Extreme variation in estimated GHG emissions from soil after this land was converted.
This wide range has prompted many experts to say that using models to predict biofuel LUC GHG emissions cannot be done with reasonable accuracy.
Just recently the Treasury Department, in consultation with the EPA, USDA and DoE provided guidance for the IRA 40B Sustainable Aviation Fuel Tax Credit program. Once again, LUC emissions of corn ethanol and soy oil biodiesel dominated the deliberations. Fortunately, cool heads prevailed as LUC emissions for both corn ethanol and soybean oil biodiesel were estimated to be modest. Perhaps their deliberations were influenced by a recently published study1 on cropland abandonment across the U.S. Using high-resolution satellite imagery not previously available to the public, and automated computer programming, these researchers were able to determine with a high degree of accuracy that more than 30 million acres of cropland had been converted back to pastures, grassland, wetlands and shrublands from 1986 to 2018.
Perhaps we can finally put the theory to bed that cropland acreage has greatly expanded due to biofuel feedstock demand?