Quantifying driving ensemble influence on operational convection-permitting ensemble spread

Convection-permitting ensembles (CPEs) are a common short-range forecasting tool designed to quantify the uncertainty in convective-scale processes, but their usefulness is limited by insufficient spread. While most efforts to improve spread have targeted the CPE itself, previous studies have shown that the “parent” driving ensemble can have a large impact. Additionally, few studies have examined the parent-child relationship for precipitation patterns, which are important for forecast guidance production but require the use of neighbourhood-based metrics for robust evaluation. By comparing spatial statistics between an operational CPE and the global ensemble used to drive it, we investigate the leadtimes and regimes under which the CPE diverges from the driving ensemble and link this to the spread-skill relationship. As a compliment to existing methods, we introduce the Parent-Child Fractions Skill Score (pcFSS) that directly compares precipitation patterns between the two ensembles. Both ensembles are similarly underspread when the lateral boundaries dominate, though the CPE demonstrates better spread during earlier periods. Additionally, under convective regimes, the CPE shows the potential for larger forecast differences compared to the global ensemble. These regimes are associated with larger CPE spread, but also lower skill. Conversely, both ensembles show similar spread and skill under mobile regimes. Ultimately, we show that using the pcFSS in conjunction with existing methods provides a broader understanding of CPE behaviour by highlighting instances of stronger and weaker driving ensemble influence.