Recent Contributions to the Understanding of the Uncertainty in Upper-Air Reference Measurements

Measurement uncertainty of climatic variables in the Upper Troposphere/-Lower Stratosphere (UTLS) is a key for the understanding of climate warming. Beyond traditional measurement uncertainty, the focus of the atmospheric science community is moving to an uncertainty which is related to the spatial and temporal scales, and to the comparison of different instruments. This paper reviews some recent contributions to the analysis of data related to radiosonde and satellite observation. In particular, techniques based on functional data analysis approach, spatiotemporal models, regularisation, and cross-validation are applied to radiosonde profile data and satellite data.

L’incertezza della misura di variabili climatiche nell’alta troposfera/bassa stratosfera (UTLS) e fondamentale nella comprensione del riscaldamento globale. Tale incertezza non riguarda solo la tradizionale incertezza strumentale ma coinvolge anche aspetti legati alla dimensione spaziale e temporale ed al confronto fra strumenti diversi. I metodi statistici considerati riguardano la functional data analysis, i modelli spazio-temporali, anche 4D, e le tecniche di cross-validazione.