A Comprehensive Analysis of Estimating Land Use Change Emissions Induced by Global Aviation Biofuels Production Using Economic Equilibrium Models

Xin Zhao, Purdue University


This dissertation is comprised of three independent but closely related essays, which provide new insights on estimating biofuels induced land use (ILUC) change and related emissions from distinct perspectives. The first essay in this dissertation developed the framework for estimating aviation biofuels induced land use change emissions using GTAP-BIO and its coupled emission model, AEZ-EF. We translate International Energy Agency (IEA) projections into appropriate shock sizes by region and pathway. We make modifications in GTAP-BIO to introduce the aviation biofuels pathways and update AEZ-EF based on new data. Twelve aviation biofuels pathways, produced using four technologies including Hydroprocessed Esters and Fatty Acids (HEFA), Fischer-Tropsch biojet (FTJ), Alcohol-To-Jet (ATJ), and Synthesized Iso-Paraffins (SIP), were evaluated and documented in this study. The results showed that the four cellulosic crop FTJ pathways provided negative ILUC emission results, due to the high soil carbon sequestration and biomass carbon from producing cellulosic crops. The palm oil HEFA biojet pathway had the highest ILUC emission intensity, mainly because of the high deforestation and peat oxidation in Malaysia and Indonesia. Only one pathway, palm oil (open pond) HEFA produced in Malaysia & Indonesia, had total life-cycle emissions higher than petroleum-based jet fuels. These results indicate that using aviation biofuels could help reduce aviation emissions. The same experiments conducted in GTAP-BIO in the first essay were tested using GLOBIOM as well. However, the results from the two models are in some cases quite different. Thus, in the second essay, in collaboration with the GLOBIOM team, we compared the land use change and emission results from the two models to explore the main drivers of the differences. We provide a detailed analysis of the differences in model framework, model data, and emission factors. The most important drivers include livestock rebound response in GLOBIOM, palm related issues, foregone sequestration on abandoned land, cropland intensification responses through multi-cropping and use of unused land, trade modelling framework, and land use change patterns in Brazil. Most of these drivers led to relatively larger ILUC emission intensity from GLOBIOM compared with GTAP-BIO. Based on the investigations of the key drivers, the GTAP-BIO and GLOBIOM teams worked closely on reconciling some of the data and assumptions in the models to resolve the key issues raised. As a result, the ILUC emissions gap between the two models shrank for all pathways due to the reconciliation process. The differences remain relatively large for all HEFA pathways, mainly due to the unresolved livestock rebound responses. In the third essay, we proposed a land use modelling framework in which land transformation is modeled using a physical land transformation functional form, additive CET (ACET). Land productivity changes due to land transformation are adjusted based on biophysical information from the land demand side, and land conversion cost is explicitly accounted for so that welfare traceability is maintained. The simulation results demonstrated that the new approach could, in a theoretically consistent manner, (1) directly provide traceable physical land use change results, (2) flexibly handle land productivity heterogeneity with biophysical information, and (3) provide detailed welfare decomposition in light of land productivity adjustment and land conversion cost. Even though the third essay was not directly linked to the first two essays in this dissertation, the developed land use modelling framework could be applied in future empirical studies estimating biofuels induced land use change.




Tyner, Purdue University.

Subject Area

Environmental economics|Agricultural economics

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server