Heatwave statistics under changing baselines: threshold sensitivity across Mediterranean climate regimes

Heatwave characteristics in Mediterranean climate regions are commonly quantified using percentile-based temperature thresholds, yet the influence of climatological baseline choice on these metrics is often overlooked. Here, we examine how different baseline windows affect heatwave thresholds and how these changes propagate through heatwave statistics across the Mediterranean region. Using ERA5-Land daily maximum and daily mean temperatures, heatwaves are identified under a consistent percentile framework (P95–P99) while varying the baseline period (1981–2010, 1981–2020, 1991–2020).

Rather than focusing on absolute climatological values, we analyze spatially aggregated differences in heatwave count, duration, and magnitude induced solely by baseline selection. The results reveal systematic and climate-dependent sensitivities: warmer baseline windows lead to elevated thresholds, fewer detected events, longer maximum durations, and mixed responses in heatwave magnitude. These effects are not uniform but vary strongly with elevation, land cover, and Köppen–Geiger climate zones. Dry and transitional Mediterranean climates show the largest sensitivity, while forested and high-elevation regions are comparatively less affected. Heatwaves defined using daily maximum temperature exhibit stronger baseline dependence than those based on daily mean temperature.

The findings demonstrate that baseline selection alone can substantially alter heatwave statistics, highlighting the need to explicitly account for threshold sensitivity when comparing heatwave characteristics or assessing climate-related risks in Mediterranean-type regions.