https://orcid.org/0000-0002-4189-9740
-
Views
-
Cite
Cite
Miriam D Cuba, Daniela Castro-Camilo, Marian E Scott, Miriam Cuba, Daniela Castro-Camilo, and Marian Scott’s contribution to the Discussion of ‘Inference for extreme spatial temperature events in a changing climate with application to Ireland’ by Healy et al., Journal of the Royal Statistical Society Series C: Applied Statistics, Volume 74, Issue 2, March 2025, Pages 312–313, https://doi.org/10.1093/jrsssc/qlaf006
Close - Share Icon Share
Extract
We thank the authors for highlighting a critical topic for climate resilience and adaptation: incorporating climate change trends at marginal and joint levels in the study of the evolution of extreme temperatures in space and time.
The authors model the temporal changes of temperature return levels in Ireland, overcoming common challenges of real-world data: missing observations, biased spatial coverage, temporal nonstationarity, and extremal spatial dependence. The first two are initially addressed by incorporating the threshold exceedances of climate model data in the marginal model for the threshold exceedances of the station data by assuming that both random variables follow the generalized Pareto distribution (GPD) and linking their respective scales. Specifically,
...
where and refer to space and time, and represents covariates at time t. Linking datasets to improve estimates is not new and constitutes the cornerstone of statistical downscaling (Giorgi et al., 2001), which aims to combine high-resolution data and locally sampled observations. For instance, Gelfand et al. (2003) and Berrocal et al. (2010) assume a Gaussian framework and propose
