China's CO2 emissions present obvious temporal and spatial distribution characteristics. Therefore, the study of spatiotemporal dynamics of CO2 emissions could provide useful information for the government and policy-makers on viable CO2 emissions mitigation in China. Using Chongqing as a case study, we investigated the spatiotemporal dynamics of CO2 emissions at the county level (38 counties) from 1997 to 2012.The mathematical statistical method, spatial autocorrelation, and rank size rule were employed to evaluate the CO2 emissions change in detail. The results showed that all of the counties in Chongqing have experienced a rapid growth of CO2 emissions, but the two dimensional structure of CO2 emissions has not changed. The Global Moran's I clearly decreases with a small fluctuation, and these values gradually decrease from 0.56 in 1997 to 0.40 in 2012.In addition, the HH clusters are concentrated in some counties in the downtown areas. Based on the rank size rule analysis, the slope values q decrease from -1.35 in 1997 to -0.88 in 2012, indicating a clear scattered pattern of CO2 emissions in Chongqing at the county level. It has also been proven that the proportion of second industries and the urbanization rate are more important impact factors for CO2 emissions than the population.
Keywords: CO2 emissions; Chongqing; county level; spatial autocorrelation; spatiotemporal dynamics.