Correcting aggregate energy consumption data to account for variability in local weather

There is a growing need in the atmospheric modeling community for city-scale energy consumption data to estimate the magnitude of waste heat emissions in urban areas. While energy consumption data are widely available at aggregate space and time scales they are often difficult to obtain at the finer scales needed in such applications. Simply assuming that local consumption patterns mirror those at coarser scales can lead to significant errors. We, therefore, present a method for correcting coarse-resolution energy data for use at the urban scale. The method is developed and validated using state and city-scale electricity data from cities in the US. Our approach develops regression models relating state-level sector-specific energy consumption to statewide temperature variables. These relations are then applied to temperature data for the city of interest to estimate city-scale consumption. This approach has been validated using residential electricity consumption data for three US cities e Houston, Los Angeles and Seattle. The fine scale weather correction scheme was found to be superior to the alternative of using the aggregate (state-level) data, reducing root mean square errors in estimated consumption by 8e40%. Much of the remaining error is believed to be a result of the assumption that the state-level building infrastructure (including heating and cooling equipment) is similar to that in each of the cities.