In the paper of Taillardat et al. (2019) “Forest-based and semiparametric methods for the postprocessing of rainfall ensemble forecasting,” a version of Eq. (13) has been provided using the incomplete beta function, in order to use only beta and incomplete beta functions:where 0 < ξ < 1 and B(;,) and B(,) denote, respectively, the incomplete beta and the beta functions.
Unfortunately, the appearance of the incomplete beta function has been made by an analytical continuation of the hypergeometric function, which was not permitted in our case. The correct Eq. (13) thus is as follows:where 0 < ξ < 1 and 2F1(,;) and B(,) denote, respectively, the hypergeometric and the beta functions.
We state that this error does not affect any discussion or conclusions in the original article. Indeed, Eq. (13) has not been used through our study, and the CRPS has been computed numerically. We thank Romain Pic for pointing out this error.
REFERENCE
Taillardat, M., A.-L. Fougères, P. Naveau, and O. Mestre, 2019: Forest-based and semiparametric methods for the postprocessing of rainfall ensemble forecasting. Wea. Forecasting, 34, 617–634, https://doi.org/10.1175/WAF-D-18-0149.1.
