Proper scoring rules for multivariate probabilistic forecasts based on aggregation and transformation

Article Article scientifique Article Proper scoring rules for multivariate probabilistic forecasts based on aggregation and transformation eng 2025-03-13 author Pic Romain 0000-0001-9263-3742 429414 Recherche et enseignement GSEM (RESEM) author Dombry Clément author Naveau Philippe 0000-0002-7231-6210 author Taillardat Maxime 0000-0003-2039-4476 23-58 10.5194/ascmo-11-23-2025 2364-3579 2407.00650 https://ascmo.copernicus.org/articles/11/23/2025/ <p>Proper scoring rules are an essential tool to assess the predictive performance of probabilistic forecasts. However, propriety alone does not ensure an informative characterization of predictive performance, and it is recommended to compare forecasts using multiple scoring rules. With that in mind, interpretable scoring rules providing complementary information are necessary. We formalize a framework based on aggregation and transformation to build interpretable multivariate proper scoring rules. Aggregation-and-transformation-based scoring rules can target application-specific features of probabilistic forecasts, which improves the characterization of the predictive performance. This framework is illustrated through examples taken from the weather forecasting literature, and numerical experiments are used to showcase its benefits in a controlled setting. Additionally, the framework is tested on real-world data of postprocessed wind speed forecasts over central Europe. In particular, we show that it can help bridge the gap between proper scoring rules and spatial verification tools.</p> Forecast verification Weather forecasts Scoring rules Spatial fields French National Research Agency (ANR) new TRends in EXtremes, prediction and validation ANR-20-CE40-0025-01 T-REX EU Horizon2020 Energy-oriented Centre of Excellence II 824158 EoCoE-II CNRS-INSU ExtremesLearning French National Research Agency (ANR) PEPR TRACCS ANR-22-EXTR-0005 French National Research Agency (ANR) EXtremes, STatistical learning and Applications ANR-23-CE40-0009 EXSTA FR true https://doi.org/10.5281/zenodo.14982271 https://doi.org/10.5281/zenodo.7429236 https://doi.org/10.5281/zenodo.10201289 Advances in statistical climatology, meteorology and oceanography 11 1 f44e9ce3-80a4-4669-9e6d-8551646a7af2 Article (Published version) Public Article_scientifique_from_crossref.pdf 2103831 application/pdf 383e4fd37ce33db9545d2efe4e334771c4a396a3b4bed79a0f29ff6af5c1c04f Vignette Public ascmo-2024-14-f01-thumb.png 46420 image/png openAccess

Title	Published in	Access level	OA Policy	Year	Views	Downloads

Proper scoring rules for multivariate probabilistic forecasts based on aggregation and transformation	Advances in statistical climatology, meteorology and oceanography			2025	59	33

Archive ouverte UNIGE

Pic, Romain

account_balance Affiliation entities

Affiliation entities