Atmosphere, Journal Year: 2025, Volume and Issue: 16(4), P. 480 - 480

Published: April 19, 2025

Highly resolved and accurate daily precipitation data are required for impact models to perform adequately correctly measure the impacts of high-risk events. In order produce such data, bias correction is often needed. Most those statistical methods correct probability distributions by modeling them with either empirical or parametric distributions. A recent semi-parametric model based on a penalized Berk–Jones (BJ) test, which allows automatic personalized splicing non-parametric distributions, has been developed. This method, called Stitch-BJ model, was found be able showed interesting potential in setting. present study, we will consolidate these results taking into account seasonal properties an out-of-sample context considering dry days probabilities our methodology. We evaluate performance method this setting against more classical as Gamma, Exponentiated Weibull (ExpW), Extended Generalized Pareto (EGP) Results show that separation necessary intra-annual non-stationarity. Moreover, distribution consistently well better than all other considered over validation set, including distribution, used due its robustness. Finally, while correcting day can easily applied, their relevance discussed temporal spatial correlations neglected.

Language: Английский