Spatial consistency and bias in avalanche forecasts – a case study in the European Alps

摘要: Abstract. In the European Alps, public is provided with regional avalanche forecasts, issued by about 30 forecast centers throughout winter, covering a spatially contiguous area. A key element in these forecasts the communication of avalanche danger according to five-level, ordinal European Avalanche Danger Scale (EADS). Consistency application the avalanche levels individual essential to avoid misunderstandings or misinterpretations users, particularly those utilizing bulletins different centers. As quality of avalanche difficult verify, due categorical nature of the EADS, we investigated goodness focusing on spatial consistency and bias, exploring real from four winter seasons (477 days). We describe operational constraints associated production communication bulletins, and propose methodology quantitatively explore spatial consistency and bias. note that level agreed significantly less often when compared across national center boundaries (about 60 %) than within (about 90 %). Furthermore, several showed significant systematic differences terms more frequently using lower (or higher) levels than their neighbors. Discrepancies seemed be greatest analyzing the proportion 4 – high 5 very high. The size warning regions, smallest geographically clearly specified areas underlying products, differed considerably between forecast Region size also had impact all summary statistics parameter influencing level, but it also limits variations level. Operational constraints avalanche forecasts variation ways EADS interpreted locally may contribute inconsistencies may potential sources for misinterpretation users. All issues highlight need to further harmonize process way avalanche hazard communicated increase consistency hence facilitate cross-border interpretation traveling