PostDoc

Abstract

Citizens in continents from Africa to Asia to Europe have one thing in common, they spend between 1 and 1.5 hours per day, 5% of their lives, traveling. But how far and by what means do they choose to travel? Transportation is the third largest sector in terms of energy use. The last 30 years travel demand, the number of person km traveled per year has been growing at staggering 3.7% per year. With petroleum products, such as gasoline, diesel or kerosene, being virtually the only source of energy. With current technology travel demand therefore determines the CO2 emissions caused by transport. Biofuels were hailed as a solution only a decade ago, but is not a silver bullet for the whole system, with land use issues and questionable CO2 neutrality. Hydrogen cars and electric vehicles might be a new solution that could fill this gap. What is the impact of these innovations and economic development on travel demand? Transportation modeling attempts to answer aforementioned questions. To this end many different types of models, with different approaches, levels of detail and predictions have been proposed. They differ in scope, the majority of models are designed for local policy forecasts and some attempt to forecast the future of the entire planet. No matter the model, if they are to be interpreted correctly it is crucial to understand how they function, how accurate they are, and how they relate to other models. Recently, several of these comparisons have been published, but by comparing the results produced by multiple authors the level of detail at which the models can be compared. This investigation recreates the travel modules of three established models, TIMER, GCAM and POLES. Moreover it creates a model based on commonly used econometric models in the field, the SIMPLE model. Firstly the four models were distilled from publications, technical descriptions and cooperation with authors. The models were created in one framework with unified formulations, calibration methods, and analysis methods. Data is taken from datasets by Schafer, the TIMER model, open access database and literature research. The models were calibrated and validated on historic data, author published results, and empirical knowledge. A methodology was devised to determine ranges for Monte Carlo simulations in an unbiased manner. The Monte Carlo simulations ran on the Brutus supercomputer, #10 in EU, at ETH. The outcomes were used to produce probabilistic projections of travel demand, and implications for CO2 emissions. Secondly a comparison of the different modeling approaches is made. Results show that if the constraints of travel money budget and travel time budget are violated, then so are historic fits. Additionally there is evidence that competition based approaches perform better than per mode growth approaches such as elasticities per mode. It is also concluded that elasticities are more complicated to model than what is presented in the models here investigated; on this long a timescale they are time dependent. Time trends and saturation levels, applied to these elasticities don’t remedy this. It is next found that a feedback mechanism for income on price is essential for a properly fitting model. Lastly the TIMER-travel and GCAM models are modified to converge consumer preferences of the developing region to the industrialized region values. Thirdly projections are created, to this end the models were tested on historic validity. It was found that most model and region combinations fitted well historically and can be used to forecast the future. Some models, however, produced large errors compared to historic data for some regions. China was discarded all models and all developing regions for half the models. Next the projections were created, these were used to compare the forecasts of the different models and include the range in their inputs. Some of the results were that Industrialized region are likely to double TD in 2100 compared to 2005 and developing region are to increase this by at least 8-fold. The TD of latter will be some six times greater by 2100 than industrialized nations, compared to nearly equal TD for both today. Aircraft demand is to increase to least 12-fold. Finally the investigation makes recommendations for future improvements on models, model comparisons and datasets. One of the main conclusions is that combining lessons learned could yield better models, perhaps by making a hybrid model. Also calibrations & model development should be done worldwide to include developing region.