Articles | Volume 7, issue 2
https://doi.org/10.5194/ascmo-7-53-2021
https://doi.org/10.5194/ascmo-7-53-2021
23 Sep 2021
 | 23 Sep 2021

Forecast score distributions with imperfect observations

Julie Bessac and Philippe Naveau

Data sets

Dataset for illustration examples Julie Bessac https://github.com/jbessac/uncertainty_scoring

Model code and software

Codes for illustration examples Julie Bessac https://github.com/jbessac/uncertainty_scoring

Download
Short summary
We propose a new forecast evaluation scheme in the context of models that incorporate errors of the verification data. We rely on existing scoring rules and incorporate uncertainty and error of the verification data through a hidden variable and the conditional expectation of scores. By considering scores to be random variables, one can access the entire range of their distribution and illustrate that the commonly used mean score can be a misleading representative of the distribution.