Modeling aviation's global emissions, uncertainty analysis, and applications to policy

Abstract

(cont.) fuel burn results below 3000 ft. For emissions, the emissions indices were the most influential uncertainties for the variance in model outputs. By employing the model, this thesis examined three policy options for mitigating aviation emissions. More stringent engine certification standards, continuous descent approach procedures, and derated take-off procedures were analyzed. Uncertainties of the model were carefully accounted for in the fuel burn and emissions scenarios of the policy options. The considered policy options achieved roughly 10-30% reductions in NOx emissions. However, HC and CO emissions rather increased due to higher emissions production rate for the CDA and derated take-off. In addition, the NOx emissions reductions in some cases were not statistically significant given the uncertainty in the modeling tool.

Air travel continues to experience fast growth. Although the energy intensity of the air transport system continues to improve, aviation fuel use and emissions of many pollutants have risen. This thesis focuses on developing, assessing and applying a system model to evaluate global aircraft fuel consumption and emissions, and to examine technological and operational measures to mitigate these emissions. The model is capable of computing how much emissions are produced on a flight-by-flight, fleet and global basis and where in the atmosphere the emissions are deposited. These are important questions for aviation environmental policy-making. Model development was followed by a comprehensive uncertainty analysis. It involved comparisons of reported versus modeled results at both the modular and system levels. On average, the aggregate-level composite fuel burn results showed about -6% difference from reported fuel burn data. A statistical analysis showed that this mean shift was a combined contribution of the key uncertainties in aircraft performance and operations. A parametric study followed to rank-order the effects that the key modeling uncertainties had on estimates of fuel burn and emissions. Statistical methods were developed to analyze both the random and systematic errors of the modeling tools. The analyses showed that the uncertainties in engine and aerodynamic performance had the largest impact on system errors, accounting for around 60-70% of the total variance in full-mission fuel burn results. The uncertainties in winds aloft and take-off weight explained another 20-25%. LTO procedures, which consist of engine throttle setting, rate of climb/descent and flight speed, were the most influential uncertainties that drove the variance in