Modeling dependence dynamics of air pollution: Time series analysis using a copula based GARCH type model

Abstract

This paper investigates the dependence structure between the Air Pollution Index (API) of Shenzhen and corresponding regional, national levels based on copula based GARCH models. In particular, time varying normal copula and time varying SJC copula are compared and employed to model the dependence structure. Comparison with the results of DCC-GARCH model is made. We find that there exists significant asymmetric upper and lower tail dependence between Shenzhen and regional, national levels; tail dependence captures the change in dependence better; dependence structure change across time. Our findings have implications for environmental management. © 2013 Springer-Verlag Berlin Heidelberg.