Abstract
This paper examines available data, develops a strategy and presents a monthly, global time series of fossil-fuel carbon
dioxide emissions for the years 1950–2006. This monthly time series was constructed from detailed study of monthly
data from the 21 countries that account for approximately 80% of global total emissions. These data were then used
in a Monte Carlo approach to proxy for all remaining countries. The proportional-proxy methodology estimates by
fuel group the fraction of annual emissions emitted in each country and month. Emissions from solid, liquid and gas
fuels are explicitly modelled by the proportional-proxy method. The primary conclusion from this study is the global
monthly time series is statistically significantly different from a uniform distribution throughout the year. Uncertainty
analysis of the data presented show that the proportional-proxy method used faithfully reproduces monthly patterns in
the data and the global monthly pattern of emissions is relatively insensitive to the exact proxy assignments used. The
data and results presented here should lead to a better understanding of global and regional carbon cycles, especially
when the mass data are combined with the stable carbon isotope data in atmospheric transport models.