A1259
Title: Vector-to-density cointegration for climate damage assessment
Authors: Kyungsik Nam - Hankuk University of Foreign Studies (Korea, South) [presenting]
Won-Ki Seo - University of Sydney (Australia)
Abstract: Global mean temperature is the standard summary of climate change, but economically relevant climate risk is distributional. A vector-to-density cointegration framework is developed that treats the temperature anomaly distribution as a density-valued functional outcome and relates its persistent evolution to a finite-dimensional vector of heterogeneous radiative forcing processes. A formal test for this long-run relation is derived and, conditional on its empirical support, estimation and simulation-based inference for forcing-specific distributional responses are developed. Empirically, stock-like long-lived greenhouse forcing is dominated by warming translation with dispersion and upper-tail expansion; chemically coupled atmospheric forcing combines moderate warming with central compression; and localized non-greenhouse forcing generates a cooling-oriented central response with outward redistribution. These findings suggest that scalar climate sensitivity may be an incomplete guide to climate-risk assessment when forcing processes differ in how they reshape temperature risk.
