Detection and attribution of urbanization impact on summer extreme heat based on nonstationary models in the Yangtze River Delta, China

Owing to a significant increase in the proportion of areas with impervious surfaces, the occurrence risk of summer extreme heat events in the Yangtze River Delta (YRD), China need to be revisited from the perspective of nonstationarity. This study applied the nonstationary generalized extreme value (GEV) models for marginal fitting of heatwave days (DH) or heatwave intensity (MIH). And the dynamic Copula-based models are used to fit the dependence structure of DH and MIH. The urbanization effects on the variations of univariate and bivariate risk of the summer extreme heat can be attributed via nonstationary risk analysis. The nonstationary models incorporating the western pacific subtropical high intensity index (WHI) and percent of impervious surface area (PISA) as covariates of the parameters exhibited a better performance for marginal distribution fitting of MIH and DH. According to analyzed results, the bivariate risk (BR) of MIH-DH has increased by 15.2%–17.8% for stations in YRD during 1988–2017 while the univariate risk (UR) of DH and MIH has increased by 18.2%–19.4% and 32.9%–35.4%, respectively.