Reconstruction and simulation of an extreme flood event in the Lago Maggiore catchment in 1868

Abstract

Abstract. Heavy precipitation on the south side of the central Alps produced a catastrophic flood in October 1868. We assess the damage and societal impacts, as well as the atmospheric and hydrological drivers using documentary evidence, observations and novel numerical weather and runoff simulations. The greatest damage was concentrated close to the Alpine divide and Lago Maggiore. An atmospheric reanalysis emphasizes the repeated occurrence of streamers of high potential vorticity as precursors of heavy precipitation. Dynamical downscaling indicates high freezing levels (4000 m a.s.l.), extreme precipitation rates (max. 270 mm 24 h−1) and weather dynamics that agree well with observed precipitation and damage, and with existing concepts of forced low-level convergence, mid-level uplift and iterative northeastward propagation of convective cells. Simulated and observed peak levels of Lago Maggiore differ by 2 m, possibly because the exact cross section of the lake outflow is unknown. The extreme response of Lago Maggiore cannot be attributed to low forest cover. Nevertheless, such a paradigm was adopted by policy makers following the 1868 flood, and used to implement nationwide afforestation policies and hydraulic structures. These findings illustrate the potential of high-resolution, hydrometeorological models – strongly supported by historical methods – to shed new light on weather events and their socio-economic implications in the 19th century.

Historical weather data for climate risk assessment

“…The second example used dynamical downscaling, which is also a widely used method. It has been applied for a number of historical weather extremes including blizzards, flash floods, or storms . Fortunato et al .…”

Causes of increased flood frequency in central Europe in the 19th century

“…Was flood frequency higher in the mid-19th century? In fact, each series exhibits prominent peaks in the 19th century, e.g., the Rhine in Basel in 1817, 1852, 1876, 1881, and 1882 (see Stucki et al, 2012), Lake Constance in 1817 (see Rössler and Brönnimann, 2018), and Lago Maggiore in 1868 (Stucki et al, 2018). However, we also note a period of low flood frequency in Basel from the 1920s to 1970s, in agreement with a low frequency of peak-over-threshold events in Basel and Ponte Tresa.…”

“…While changes in seasonality have been found for European floods (Blöschl et al, 2017), no general increase in flood frequency has so far been detected (Madsen et al, 2014). However, past records suggest that there is considerable decadal variation in flood frequency (e.g., Sturm et al, 2001;Wanner et al, 2004;Glaser et al, 2010). It is reasonable to assume that such variations will continue into the future.…”

“…However, the causes of this increased flood frequency remain unclear. Besides human interventions such as deforestation or undesigned effects from water flow regulations (Pfister and Brändli, 1999;Summermatter, 2005), this includes the role of cold or warm periods and changes in atmospheric circulation. Proxy-based studies, though focussing on longer timescales, find that in the Alps, cold periods were mostly more flood prone than warm periods (Stewart et al, 2011;Glur et al, 2013); the last of these cold and flood-prone periods in the latter study is the 19th century.…”

Simulations of the 2005, 1910 and 1876 Vb cyclones over the Alps – Sensitivity to model physics and cyclonic moisture flux

“…In fact, 20CR has proven to be a valuable input dataset for downscaling heavy-precipitation and windstorm events over the Central Alps back to the 19th century. Stucki et al (2018) showed that downscaling the ensemble 10 mean is not only computationally cheaper, but can be seen as a minimum-error and thus natural approach in well represented areas and distinctive synoptic flow conditions. For an extreme flood in 1868, they found a small smoothing effect of the associated cyclone, which induced southerly moisture flux, i.e.…”