HR: 1340h
AN: B33A-0226 [Abstracts]
TI: Modeling Seasonality in German Carbon Dioxide Emissions From Fossil Fuel Consumption
AU: * Gregg, J S
EM: jgregg@space.edu
AF: University of North Dakota, Department of Space Studies
Box 9008, Grand Forks, ND 58203
United States
AU: Andres, R J
EM: andres@space.edu
AF: University of North Dakota, Department of Space Studies
Box 9008, Grand Forks, ND 58203
United States
AB:
A method is developed to determine seasonal fossil fuel consumption patterns by using monthly sales data to estimate the
relative monthly proportions of the total annual carbon dioxide emissions for Germany. From these data, the goal is to
develop mathematical models that describe the seasonal flux in consumption for each type of fuel, as well as the total
emissions for this nation. The time series models have two components. First, the general long-term trend is determined with
regression models. After removing the general trend, two alternatives are considered for modeling the seasonality. The first
alternative uses the mean of the monthly proportional consumption to predict the seasonal distribution. A second alternative
is to use an ordinary least squares autoregressive model. This model is chosen for its ability to accurately describe
dependent data and for its predictive capacity. It also has a meaningful interpretation, as each coefficient in the model
quantifies the dependency for each corresponding time lag. Most importantly, it is dynamic, and able to adapt to anomalies
and changing patterns. To model the monthly fuel consumption, the annual trend is combined with the seasonal model. The
models for each fuel type are then summed together to predict the total carbon dioxide emissions. The prediction error is
estimated with the root mean square error from the actual estimated emission values. The result is a quantitative description
of carbon dioxide emissions from fossil fuel consumption.
DE: 9335 Europe
DE: 1610 Atmosphere (0315, 0325)
DE: 0400 Biogeosciences
SC: Biogeosciences [B]
MN: 2004 AGU Fall Meeting