A0199
Title: An unconditional-quantile vector error correction model to analyze climate heterogeneity
Authors: Andrey Ramos - Carlos III University of Madrid (Spain) [presenting]
Abstract: A novel time-series quantitative methodology is introduced to analyze heterogeneity in the temperature distribution and its association with climate forcings. The approach employs a vector autoregressive model (VAR) for a range of unconditional distributional characteristics of temperature mean and quantiles alongside the radiative forcing of greenhouse gases (GHGs), including carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O). The empirical analysis is conducted across three geographical scales: the Globe, the North Hemisphere, and Europe, utilizing station-level data from the Climatic Research Unit (CRU) during the period 1880-2021. At these scales, the temperature unconditional quantiles represent temperature at different latitudes, and the results are comparable to the predictions of one-dimensional (1D) energy balance models (EBMs). However, the methodology can be adapted to more general situations with limited spatial variation, assuming there is sufficient cross-sectional or higher frequency data to derive the unconditional distributional characteristics of temperature. The proposed methodology serves various purposes, including i) Estimation of physical parameters like climate sensitivity, ii) Forecasting, iii) Identification of structural shocks and impulse-response analysis, and iv) Predictions of temperature conditional on hypothetical emissions/concentrations scenarios.
