Estimating the displacement in precipitation forecasts using the Fractions Skill Score

The Fractions Skill Score ( FSS ) is a popular spatial verification metric commonly used for precipitation verification. In this study we focus on analysing the ability of FSS to provide meaningful information about the displacement between precipitation in one field compared to another. A simple overlap‐adjusted use of the FSS is introduced and a number of relevant idealized cases are analysed that show that the FSS can indeed be used to determine displacement in a meaningful way. It was found that the displacement provided by the FSS is directly related to the true displacements of precipitation but with larger contiguous precipitation objects having a much larger influence. Overall, the displacement provided via the FSS compares well with the average distance to the closest neighbouring precipitation object (assuming the objects are of similar size). It is recommended that the user should use a frequency (percentile) threshold when focussing on spatial differences unless biases are known to be small and adopt the overlap‐adjusted variant of the FSS displacement. If the frequency bias is very large the FSS ‐derived displacements become less reliable. The same is true of any spatial comparison. A recipe for the use of the FSS for determining displacements is provided.