Changing hydroclimate dynamics and the 19th to 20th century wetting trend in the English Channel region of northwest Europe

Abstract

Northwestern Europe has experienced a trend of increasingly wet winters over the past 150 years, with few explanations for what may have driven this hydroclimatic change. Here we use the Old World Drought Atlas (OWDA), a tree-ring based reconstruction of the self-calibrating Palmer Drought Severity Index (scPDSI), to examine this wetting trend and place it in a longer hydroclimatic context. We find that scPDSI variability in northwestern Europe is strongly correlated with the leading mode of the OWDA during the last millennium (1000-2012). This leading mode, here named the 'English Channel' (EC) mode, has pronounced variability on interannual to centennial timescales and has an expression in scPDSI similar to that of the East Atlantic teleconnection pattern. A shift in the EC mode from a prolonged negative phase to more neutral conditions during the 19th and 20th centuries is associated with the wetting trend over its area of influence in England, Wales, and much of northern continental Europe. The EC mode is the dominant scPDSI mode from approximately 1000-1850, after which its dominance waned in favor of the secondary 'North-South' (NS) mode, which has an expression in scPDSI similar to that of the winter North Atlantic Oscillation (NAO). We examine the dynamical nature of both of these modes and how they vary on interannual to centennial timescales. Our results provide insight into the nature of hydroclimate variability in Europe before the widespread availability of instrumental observations.

Evaluating downscaled products with expected hydroclimatic co-variances

“…This is in strong contrast to dynamical downscaling (regardless of the regional climate model chosen), which preserves non-stationary physical 315 relationships among variables and is thus able to capture such changes to meteorological phenomena. Such considerations may serve as limitations to the credibility of downscaled projections, particularly when considering the dominant (e.g., Baek et al 2019Baek et al , 2021 and/or non-stationary (e.g., Baek et al 2020;Scholz et al 2022) nature of internal atmospheric variability in driving hydrologic hazards. Importantly, our work highlights expected covariances 320 during convective and frontal precipitation as process-based evaluation diagnostics that can be applied universally to a wide range of statistically downscaled products.…”