The severe thunderstorms risk estimates the probability of severe thunderstorms (supercells, downbursts) based on atmospheric conditions, not on observed storms. It is measured with an "ingredients-based" approach: upper-level wind shear is combined with convective energy and inhibition, counting each year the days on which the conditions favourable to severe storms coexist. The more severe days, the higher the risk.
Indicator. Number of "severe" days per year, from an ingredients-based approach (wind shear, CAPE, CIN).
Data. Daily CMIP6 variables from the Copernicus Climate Data Store (CDS): air temperature and relative humidity at multiple pressure levels, east/north wind components at 1000 hPa and 500 hPa. Scenarios: SSP1-2.6 / SSP2-4.5 / SSP3-7.0 / SSP5-8.5.
Calculation.
(1) The wind shear between 1000 and 500 hPa, that is from the ground to about 5.5 km of altitude, is computed as the vector difference between the wind at the two levels.
(2) Only for days with wind shear > 18 m/s are CAPE (convective available potential energy) and CIN (convective inhibition) computed with the MetPy library from the temperature and humidity profiles.
(3) A day is "severe" if simultaneously wind shear > 18 m/s, CAPE > 1000 J/kg and CIN < −50 J/kg.
(4) The severe days per year (N_severe) are counted, by scenario.
Thresholds → level A–F.
• Based on N_severe days/year: F: ≥15
• E: ≥10
• D: ≥6
• C: ≥3
• B: ≥1
• A: 0.
The thresholds (wind shear 18 m/s, CAPE 1000 J/kg, CIN -50 J/kg and the counts of 1, 3, 6, 10 and 15 days) are a xNatura choice without a published source.
Limits: the CMIP6 profiles are daily means, which smooth out the instability of the hottest hours of the day: few severe days are counted and the risk can be underestimated.
From physical to impact. The physical level (A–F) is combined with the production process's sensitivity through the risk matrix (6 physical levels by 5 sensitivity bands) → impact level; the damage curve converts the level into a damage factor, from which the Version 2025 value at risk is derived. The damage factor per class is an xNatura estimate without a published source: the value is indicative.
Sources.
• [Flash Flood Forecasting: An Ingredients-Based Methodology](Flash Flood Forecasting: An Ingredients-Based Methodology)
• [A Baseline Climatology of Sounding-Derived Supercell andTornado Forecast Parameters](A Baseline Climatology of Sounding-Derived Supercell andTornado Forecast Parameters)
• [The spatial distribution of severe thunderstorm and tornado environments from global reanalysis data](The spatial distribution of severe thunderstorm and tornado environments from global reanalysis data)
• [Storm Prediction Center (SPC)](Storm Prediction Center (SPC))
• metpy.
Indicator. Number of "severe" days per year, from an ingredients-based approach (wind shear, CAPE, CIN).
Data. Daily CMIP6 variables from the Copernicus Climate Data Store (CDS): air temperature and relative humidity at multiple pressure levels, east/north wind components at 1000 hPa and 500 hPa. Scenarios: SSP1-2.6 / SSP2-4.5 / SSP3-7.0 / SSP5-8.5.
Calculation.
(1) The wind shear between 1000 and 500 hPa, that is from the ground to about 5.5 km of altitude, is computed as the vector difference between the wind at the two levels.
(2) Only for days with wind shear > 18 m/s are CAPE (convective available potential energy) and CIN (convective inhibition) computed with the MetPy library from the temperature and humidity profiles.
(3) A day is "severe" if simultaneously wind shear > 18 m/s, CAPE > 1000 J/kg and CIN < −50 J/kg.
(4) The severe days per year (N_severe) are counted, by scenario.
Thresholds → level A–F.
• Based on N_severe days/year: F: ≥15
• E: ≥10
• D: ≥6
• C: ≥3
• B: ≥1
• A: 0.
The thresholds (wind shear 18 m/s, CAPE 1000 J/kg, CIN -50 J/kg and the counts of 1, 3, 6, 10 and 15 days) are a xNatura choice without a published source.
Limits: the CMIP6 profiles are daily means, which smooth out the instability of the hottest hours of the day: few severe days are counted and the risk can be underestimated.
From physical to impact. The physical level (A–F) is combined with the production process's sensitivity through the risk matrix (6 physical levels by 5 sensitivity bands) → impact level; the damage curve converts the level into a damage factor, from which the Version 2025 value at risk is derived. The damage factor per class is an xNatura estimate without a published source: the value is indicative.
Sources.
• [Flash Flood Forecasting: An Ingredients-Based Methodology](Flash Flood Forecasting: An Ingredients-Based Methodology)
• [A Baseline Climatology of Sounding-Derived Supercell andTornado Forecast Parameters](A Baseline Climatology of Sounding-Derived Supercell andTornado Forecast Parameters)
• [The spatial distribution of severe thunderstorm and tornado environments from global reanalysis data](The spatial distribution of severe thunderstorm and tornado environments from global reanalysis data)
• [Storm Prediction Center (SPC)](Storm Prediction Center (SPC))
• metpy.