Modeling high resolution space-time variations in energy demand/CO2 emissions of human inhabited landscapes in the United States under a changing climate

With urban and exurban areas now accounting for more than 50% of the world's population, projected to increase 20% by 2050 (UN World Urbanization Prospects, 2009), urban-climate interactions are of renewed interest to the climate change scientific community (Karl et. al, 1988; Kalnay and Cai, 2003; Seto and Shepherd, 2009). Until recently, climate modeling efforts treated urban-human systems as independent of the earth system. With studies pointing to the disproportionately large influence of urban areas on their surrounding environment (Small et. al, 2010), modeling efforts have begun to explicitly account for urban processes in land models, like the CLM 4.0 urban layer, for example (Oleson.et. al, 2008, 2010). A significant portion of the urban energy demand comes from the space heating and cooling requirement of the residential and commercial sectors - as much as 51% (DOE, RECS 2005) and 11% (Belzer, D. 2006) respectively, in the United States. Thus, these sectors are both responsible for a significant fraction of fossil fuel CO2 emissions and will be influenced by a changing climate through changes in energy use and energy supply planning. This points to the possibility of interactive processes and feedbacks with the climate system. Space conditioning energy demand is strongly driven by external air temperature (Ruth, M. et.al, 2006) in addition to other socio-economic variables such as building characteristics (age of structure, activity cycle, weekend/weekday usage profile), occupant characteristics (age of householder, household income) and energy prices (Huang, 2006; Santin et. al, 2009; Isaac and van Vuuren, 2009). All of these variables vary both in space and time. Projections of climate change have begun to simulate changes in temperature at much higher resolution than in the past (Diffenbaugh et. al, 2005). Hence, in order to understand how climate change and variability will potentially impact energy use/emissions and energy planning, these two components of the human-climate system must be coupled in climate modeling efforts to better understand the impacts and feedbacks. To implement modeling strategies for coupling the human and climate systems, their interactions must first be examined in greater detail at high spatial and temporal resolutions. This work attempts to quantify the impact of high resolution variations in projected climate change on energy use/emissions in the United States. We develop a predictive model for the space heating component of residential and commercial energy demand by leveraging results from the high resolution fossil fuel CO2 inventory of the Vulcan Project (Gurney et al., 2009). This predictive model is driven by high resolution temperature data from the RegCM3 model obtained by implementing a downscaling algorithm (Chow and Levermore, 2007). We will present the energy use/emissions in both the space and time domain from two different predictive models highlighting strengths and weaknesses in both. Furthermore, we will explore high frequency variations in the projected temperature field and how these might place potentially large burdens on energy supply and delivery.