Unlike industrial systems, agricultural systems are subject to the influence of weather patterns, soil type, geography, and management practices. Even the same agricultural product may have drastically different input structures, hence environmental impacts, in different regions. Therefore, average data with generic descriptions of material and energy fluxes are hardly adequate to capture the high degree of system variability of agricultural products. With the rising interests in bio-fuels as a means to combat climate change across the world, we strongly recommend future studies in this area to take into consideration the spatial variability of biomass growth. Just as technological and environmental variability exists across states, there is probably certain variability within a state, too, that may not be precisely captured by state average data. This does not mean, however, that state-level data should be dismissed for the research question at hand because they are still likely more reflective of local or farm-level practices than national averages. In addition, state average data are especially valuable and representative, more so than farm-level data, in situations in which massive land shift between crops takes place within a state. Nevertheless, we encourage finer-scale, more detailed studies into land shift between cotton and corn and associated environmental impacts, which could not have been conducted in our analysis due to the data limitation and resources constraints.Additional research is needed to paint a more complete picture on the impact of cropland conversion to corn: In 2005, 41 states grew corn and 17 states grew cotton, among which only 19 of the corn-growing states and 7 of the cotton-growing states had data on major inputs that can be used to generate LCIs . Among these states, only three overlap, namely, North Carolina, Georgia, and Texas. Therefore, this study does not quantify the environmental impact and their trade-offs in other cotton-growing states where conversion to corn might have happened. Nevertheless,plastic garden container environmental implications of cotton-to-corn land shift in these other states are probably worse than that indicated by Fig. 2.1 and closer to that indicated in Fig. 2.2 because cropland in southern states are generally less suitable for corn growth than the Corn Belt.
Future studies pursuing this line of research may make the effort to quantify the magnitude of land shift in each cotton-growing state when relevant data on agricultural inputs, environmental outputs, and acreage of conversion become available. Furthermore, it is worth noting that spatially detailed data are often unavailable or incomplete, although such data can improve the environmental relevance of an LCA study. In this case, one may rely on assumptions or spatially generic data to fill the gaps, and this may increase the uncertainty of the LCA results . In our study, data on agricultural inputs such as fertilizers and pesticides were available at the state level, but we often relied on spatially generic emission factors to estimate their emissions . Further, the LCA results for corn and cotton were found to be moderately sensitive to the emission factors which are likely to vary across regions . Future spatially explicit LCAs on agricultural systems may take this into account and direct efforts to estimate spatially differentiated emission factors.For the potential to mitigate climate change, reduce dependence on oil imports, and invigorate rural economic development, bio-fuel development in the USA has been supported by an array of policy measures . Among them is the federal Renewable Fuel Standard , a mandate that requires 140 billion liters bio-fuels to be produced annually from different sources by 2022. Corn ethanol is currently the primary bio-fuel and is likely to continue dominating US bio-fuels market as cellulosic and other advanced bio-fuels are far from mass production . Driven by the favorable policies and high oil prices, corn ethanol production has increased eight-fold since 2000, to the current level of about 50 billion liter per year. Early Life Cycle Assessment research on corn ethanol was largely in support of the policies aiming partly at reducing greenhouse gas emissions. As is typically done in LCA, these studies quantified GHG emissions generated at each stage of corn ethanol’s life cycle, summed them up, and then compared the results against that of gasoline. Corn ethanol was found to have 10–20 % lower life cycle GHG emissions than gasoline and, therefore, concluded to provide a modest carbon benefit in replacing gasoline . However, the conclusion was later called into question, when the land use change effects of corn ethanol expansion emerged in the literature . Converting natural vegetation or forestland to corn field for ethanol production releases a substantial amount of carbon from soil and plant biomass, creating a “carbon debt” that could not be repaid in dozens of years or even a century . Similarly, diversion of existing cropland for ethanol could generate indirect LUC effect through market-mediated mechanisms . In this scenario, corn ethanol expansion reduces food supply, which could lead to conversion of natural vegetation or forestland elsewhere in the world to compensate for the diverted grains.
While the concept of iLUC has become widely accepted in academic and policy arenas , quantification of iLUC emissions is known to be difficult and highly uncertain . Plevin et al. , for example, estimated the range from 10 to 340 CO2e MJ−1 y−1. This wide range is due in large part to a lack of quality data and detailed understanding as to how the global agricultural market would respond to bio-fuels expansion . In contrast, the direct land use change emissions can be relatively accurately quantified . Previous studies used the concept of carbon payback time to measure the magnitude of dLUC effect of corn ethanol. While the initial carbon debt due to land conversion may be large, it can be repaid over time by the annual carbon savings corn ethanol yields in displacing gasoline because corn ethanol has lower life cycle GHG emissions. The first dLUC study estimated that 48 years would be required for corn ethanol to pay back its carbon debt if the Conservation Reserve Program land is converted and that 93 years would be required if central grassland is converted .Gelfand et al. conducted a field experiment on CRP land conversion to measure its carbon loss. They found that approximately 40 years would be required for the use of corn ethanol to pay back this carbon loss with the converted land under no-till management. In another study, Piñeiro et al. arrived at a similar estimate of approximately 40 years for the payback time for CRP land conversion to corn ethanol. However, these studies were based on several oversimplifications that may substantially affect their results. First, these studies assumed that newly converted land has the same yield as existing cornfields, neglecting the potential yield differences of newly converted land. In particular, CRP land is generally less fertile than cornfields that have been in continuous use . Thus, corn ethanol from CRP land generates lower annual carbon savings, hence a longer payback time. Land with extremely low yield may even fail to provide any carbon savings, in which case the carbon loss due to land conversion is permanently lost. Second, the dLUC studies relied primarily on life cycle assessments based on early bio-fuel conversion processes that did not reflect the productivity improvements that have occurred in the past decade due to yield and energy efficiency increases at both the corn growing and ethanol conversion stages . Recent studies have shown that corn ethanol’s carbon benefit has increased to up to 50 % , compared with earlier estimates of 10–20 % . The productivity of the gasoline production system over the same period of time has been fairly steady . The productivity improvements in the corn ethanol system result in greater amounts of annual carbon savings that, if considered, would yield a shorter payback time than previously estimated. Finally, the dLUC studies used the global warming potential 100 to assess the global warming impact of corn ethanol, gasoline, and dLUC emissions. This approach assumes equal weights to GHGs emitted at different times. More recent literature explores the application of different weights to GHG emissions emitted in different times. First,plastic pot from a scientific point of view, increasing background GHG concentrations in the atmosphere result in a diminishing marginal radiative forcing for a unit GHG emission . The rate at which the relative radiative forcing effect of a unit GHG emission diminishes depends on future atmospheric GHG concentrations.
Reisinger et al. , for example, estimated that the 100-year Absolute Global Warming Potential of CO2 from 2000 to 2100 could decrease by 2 to 36 % under various GHG concentration scenarios. Second, a series of articles have attempted to synchronize the temporal system boundary under which life cycle emissions are taken into account and the time horizon under which characterization factors are derived. For example, if GWP100 is to be used, one can set the temporal system boundary to the next 100 years and account for the radiative forcing effects that occur within that time horizon . One of the rationales is that the efforts to reduce GHG emissions today is perhaps more valuable than those in the future because climate change may bring about irreversible damages to the planet . In this class of literature, simple climate-carbon cycle model like Bern model or simple first-order decay model is used to calculate atmospheric load of GHGs over time, and corresponding radiative forcing . Background concentrations of GHGs are, however, generally assumed to be constant in the literature. Third, some argues that future climate change impacts should be discounted at certain rates using the net present value approach . These approaches use different rationales and involve varying degrees of subjectivity in, e.g., the choice of emission scenarios and discount rates. For the sake of simplicity, however, these approaches are collectively referred to as dynamic characterization method in this paper. The objective of this study is to re-examine corn ethanol’s CPT, taking into account the potential yield differences of converted land and technological advances within the corn ethanol system. We also examine how dynamic characterization of GHG emissions changes the CPT using one particular approach as an example. We focus on conversion of CRP land primarily for ease of comparison with previous studies and also because there is evidence indicating that conversion of CRP land to cornfield has occurred with the expansion of corn production in the past decade . We start with estimating the amount of annual carbon savings that can be generated by corn ethanol from an average cornfield and how the amount changes over time. For this analysis, we use the Bio-fuel Analysis Meta-Model with several modifications . Specifically, because the base year of EBAMM is 2001 , we incorporate into the model historical data on the process inputs and outputs of corn growth and ethanol conversion for 2005 and 2010 to reflect the system’s productivity improvements in the past decade . We project further productivity improvements to 2020 using projections in the Greenhouse gases, Regulated Emissions and Energy use in Transportation model . We assume that technology advancement stabilizes after 2020 . Detailed information is provided in Appendix B. We then incorporate yield differences into the model to approximate the amount of annual carbon savings that the CRP-corn ethanol system provides. The CRP program, established by the Food Security Act of 1985, is intended to retire highly erodible and environmentally sensitive cropland from production . Because highly erodible land is less productive in general, the program enrolls land with lower productivity indirectly . Additionally, due in part to the early payment scheme— the maximum acceptable rental rates—farmers tended to offer their low-quality land for CRP consideration while retaining productive land for continuous cultivation . As a result, CRP land appears less productive than other types of cropland, including land that shifts into or out of the cultivated cropland from less, other intensive uses . Direct measurements of crop yield on CRP land are scarce, but measurements of crop yield on marginal land, including CRP and shifting land, can be used as indications of the relative yield differences between CRP land and average croplands .