The devil in the detail of storms

Wintertime windstorms associated with low-pressure systems from the North Atlantic Ocean are the costliest natural hazard for Europe (Ulbrich et al 2013). These storms are associated with large pressure gradients and high background winds, but the most destructive gusts are often confined to relatively small areas within the low-pressure systems. Examples include showers and thunderstorms in the area of the cold front (e.g. Ludwig et al 2015) and the so-called ‘sting jet’ as analysed in the recent paper by Martı́nez-Alvarado et al (2018). Sting jets are relatively small and short-lived high-wind zones in the area of the bent-back warm front during the mature stage of a particular type of low-pressure system (see figure 1 for a recent example). The name goes back to a seminal paper by Browning (2004) calling the high winds ‘the sting at the end of the tail’. Sting jets typically occur in explosively intensifying storms and are mostly confined to the North Atlantic Ocean and the adjacent British Isles (Hart et al 2017). There is an ongoing debate about the details of the physical mechanisms involved in creating sting jets, including aspects such as conditional symmetric instability, frontolysis and cooling through evaporation, melting and sublimation. Up to now assessments of future risks for winterstorm damages in Europe mostly rely on coarseresolution global climate models. These show general changes in the number, intensity and track of lowpressure systems, but are not capable of resolving the actual high-wind areas within storms. So the devil in a true account of future wind risk may well lie in the details of how storms change. Generally speaking, three approaches have been applied to address this kind of problem: (1) Nesting a high-resolution regional domain into the global climate model. This is computational expensive and very demanding in terms of data storage, particularly if horizontal and vertical resolutions are high enough to realistically represent convective momentum transport from middle levels (2–5 km) to the surface, which has been shown to be crucial in some storms at least (Ludwig et al 2015). (2) Pseudo-global warming experiments, in which representative present-day example cases are simulated in high resolution, artificially modifying temperature, moisture and wind fields to test the dynamical effect of these changes. Such an approach has been used for tornadic storms (e.g. Trapp and Hoogewind 2016), but does not allow a rigorous statistical assessment. Moreover, it is challenging for mid-latitude low-pressure systems, as these travel long distances in short time. (3) Statistical approaches based on dynamical precursors computed from coarse-resolution global climate models that allow diagnosing the potential of hazardous weather on smaller scales without explicitly having to run high-resolution simulations. The ground-breaking new paper by Martı́nezAlvarado et al (2018) is the first to apply the dynamical precursor approach for sting jets to a climate change experiment. This has become possible due to the availability of model output suitable to compute all necessary fields. Interestingly, the authors show that while the total number of storms in the model slightly decreases in the future, those with sting jet precursors dramatically increase. In addition, the results by Martı́nez-Alvarado et al (2018) show an expansion of the area affected by explosively deepening sting-jet storms into northern Europe, well beyond the British Isles. A possible physical reason for this behaviour is that in a warmer world the increase in atmospheric moisture content could enhance the moist instabilities involved in creating sting jets. The coarse-resolution model data indicate a general increase in strong wind occurrence, but given the inability of the model to realistically represent sting jets, thismay still be a substantial underestimation of the real impact. This important and alarming result calls for enhanced research efforts focusing on the smallerscale details of future storm developments. Recent years have already seen the unexpected occurrence of sting jet storms in central Europe (e.g. storm ‘Egon’ over southern Germany, figure 1). Moreover, in the current climate strong gusts associated with cold-frontal convection are an even more frequent hazard for Europe as a whole. Little has been done to assess


E-mail: peter.knippertz@kit.edu
Wintertime windstorms associated with low-pressure systems from the North Atlantic Ocean are the costliest natural hazard for Europe (Ulbrich et al 2013). These storms are associated with large pressure gradients and high background winds, but the most destructive gusts are often confined to relatively small areas within the low-pressure systems. Examples include showers and thunderstorms in the area of the cold front (e.g. Ludwig et al 2015) and the so-called 'sting jet' as analysed in the recent paper by Martínez-Alvarado et al (2018). Sting jets are relatively small and short-lived high-wind zones in the area of the bent-back warm front during the mature stage of a particular type of low-pressure system (see figure 1 for a recent example). The name goes back to a seminal paper by Browning (2004) calling the high winds 'the sting at the end of the tail'. Sting jets typically occur in explosively intensifying storms and are mostly confined to the North Atlantic Ocean and the adjacent British Isles (Hart et al 2017). There is an ongoing debate about the details of the physical mechanisms involved in creating sting jets, including aspects such as conditional symmetric instability, frontolysis and cooling through evaporation, melting and sublimation.
Up to now assessments of future risks for winterstorm damages in Europe mostly rely on coarseresolution global climate models. These show general changes in the number, intensity and track of lowpressure systems, but are not capable of resolving the actual high-wind areas within storms. So the devil in a true account of future wind risk may well lie in the details of how storms change. Generally speaking, three approaches have been applied to address this kind of problem: (1) Nesting a high-resolution regional domain into the global climate model. This is computational expensive and very demanding in terms of data storage, particularly if horizontal and vertical resolutions are high enough to realistically represent convective momentum transport from middle levels (2-5 km) to the surface, which has been shown to be crucial in some storms at least (Ludwig et al 2015).
(2) Pseudo-global warming experiments, in which representative present-day example cases are simulated in high resolution, artificially modifying temperature, moisture and wind fields to test the dynamical effect of these changes. Such an approach has been used for tornadic storms (e.g. Trapp and Hoogewind 2016), but does not allow a rigorous statistical assessment. Moreover, it is challenging for mid-latitude low-pressure systems, as these travel long distances in short time. (3) Statistical approaches based on dynamical precursors computed from coarse-resolution global climate models that allow diagnosing the potential of hazardous weather on smaller scales without explicitly having to run high-resolution simulations.
The ground-breaking new paper by Martínez-Alvarado et al (2018) is the first to apply the dynamical precursor approach for sting jets to a climate change experiment. This has become possible due to the availability of model output suitable to compute all necessary fields. Interestingly, the authors show that while the total number of storms in the model slightly decreases in the future, those with sting jet precursors dramatically increase. In addition, the results by Martínez-Alvarado et al (2018) show an expansion of the area affected by explosively deepening sting-jet storms into northern Europe, well beyond the British Isles. A possible physical reason for this behaviour is that in a warmer world the increase in atmospheric moisture content could enhance the moist instabilities involved in creating sting jets. The coarse-resolution model data indicate a general increase in strong wind occurrence, but given the inability of the model to realistically represent sting jets, this may still be a substantial underestimation of the real impact.
This important and alarming result calls for enhanced research efforts focusing on the smallerscale details of future storm developments. Recent years have already seen the unexpected occurrence of sting jet storms in central Europe (e.g. storm 'Egon' over southern Germany, figure 1). Moreover, in the current climate strong gusts associated with cold-frontal convection are an even more frequent hazard for Europe as a whole. Little has been done to assess potential changes in their frequency and intensity in the future. Convection-resolving, long-term climate change experiments as those by Leutwyler et al (2017) offer new possibilities to analyse storms in much more detail, without the uncertainty coming from any precursor diagnostics. However, such models still struggle to simulate the full lifecycle of low-pressure systems over a sufficiently long time-period. Therefore a careful combination of the approaches listed above remains the most promising way to generate a much more complete picture of the future behaviour of European windstorms and the damaging high-wind areas embedded in them.