Scientific transparency
Validated methods, proprietary algorithms, continuous improvement. High-resolution models, composite radar, and orographic corrections validated in the field.
Six families of meteorological calculations, combined in real time for every storm cell detected.
Convective Available Potential Energy in all its forms. Measures the energy available for storm updrafts. The higher the CAPE, the more explosive the convective potential.
Thresholds: >1000 J/kg moderate | >2500 J/kg strong | >4000 J/kg extreme
Compares the temperature of a lifted air parcel to that of the environment at 500 hPa. A negative LI indicates instability favorable to storms.
Thresholds: <-2 unstable | <-4 very unstable | <-6 extreme
Energy needed to trigger convection. Too much CIN prevents storms, too little allows widespread, disorganized triggering.
Thresholds: >50 J/kg strong inhibition | 25-50 moderate | <25 weak
Equivalent potential temperature at 850 hPa. Key indicator of warm and humid low-level air mass, the fuel for severe storms.
Thresholds: >330K storm potential | >340K strong instability
Wind difference between surface and various altitudes. Shear organizes storms, separates updrafts and downdrafts, and favors supercells.
0-6km thresholds: >20 m/s organized | >30 m/s supercellular
Measures the potential rotation of the updraft. Low-level helicity (0-1 km) is critical for tornado potential.
0-1km thresholds: >100 m²/s² rotation | >250 m²/s² tornadic
Helicity relative to storm motion. Key for evaluating mesocyclone potential and the probability of persistent rotation.
Thresholds: >150 m²/s² mesocyclone likely | >300 m²/s² intense
Vector representation of the wind profile aloft. Its curvature and length determine the expected storm type: linear, right-moving or left-moving supercell.
Combines CAPE, deep shear, and SRH to evaluate the overall supercell potential of an environment.
Thresholds: >1 supercell possible | >4 favorable environment
Composite parameter specific to significant tornadoes (EF2+). Integrates CAPE, LCL, shear, and low-level helicity.
Thresholds: >1 significant tornado possible | >4 elevated risk
Crosses CAPE and helicity to quantify both energy and rotation simultaneously. Robust indicator of severe weather potential.
Thresholds: >1 rotation possible | >2 mesocyclone likely
Ratio between buoyancy energy and shear. Determines the type of convection: multicellular, supercellular, or intermediate.
Thresholds: 10-45 supercellular | <10 over-sheared | >45 multicellular
Altitude at which an air parcel reaches saturation. A low LCL favors tornadoes by bringing the mesocyclone base closer to the ground.
Thresholds: <1000m low (tornadic) | 1000-1500m moderate
Altitude where the parcel becomes warmer than the environment and accelerates freely. The LFC-LCL difference indicates how easily convection triggers.
Theoretical top of the storm updraft, where the parcel reaches environmental temperature. Determines the maximum height of convective towers.
Identifies inversion layers that block convection. Their erosion (diurnal heating, forcing) determines the timing of triggering.
Areas where surface winds converge, forcing air to rise. Main mechanical trigger for storms in homogeneous air masses.
Intrusion of cold stratospheric air at altitude. Powerful dynamic forcing that destabilizes the column and favors deep convection.
Absolute vorticity at altitude and jet stream position create areas of upper-level divergence favoring large-scale ascent.
Forced lifting by terrain (Pyrenees, Massif Central, Alps, Rockies). Channeling of winds in valleys and convergence at the foot of mountains.
Physics-based method for calculating storm cell motion. Decomposes the mean wind and the deviation due to mesocyclone rotation (right or left).
Inspired by Bunkers et al. 2000, Weather & Forecasting
Probabilistic simulations exploring thousands of possible trajectories. Generates probability corridors rather than a single track, integrating inherent uncertainty.
Trajectory adjustments based on real terrain: deviation by mountain ranges, acceleration in valleys, blocking by ridges. Essential for accuracy in mountainous areas.
Automatic real-time classification: ordinary cell, multicellular, supercell (right/left), squall line, MCS. Each type has its own motion model.
The research behind our approach. All peer-reviewed and published in leading scientific journals.
50% increase in supercells in the Alpine region. Confirms that climate change intensifies severe storms in mountainous Europe.
Read the studyEuropean-scale study showing the increase in frequency and intensity of severe convective events linked to climate warming.
Read the studyLong-term trend analysis of lightning and hail activity in Europe. Essential data to calibrate our detection and intensity algorithms.
Read the studyEuropean Environment Agency report. €55 billion in annual economic losses from extreme weather events in Europe.
Read the reportFoundational paper of the Bunkers method for predicting supercell motion. Scientific inspiration for our Diamond Trajectory Engine and its orographic corrections.
Read the studyStorm Predict is committed to total transparency on its methods, its results, and its limits. Meteorology remains complex, perfection doesn't exist. We tell you that openly.
Algorithms continuously improve with every storm tracked. Each season brings new data, new validations, and new improvements.
Choose the path that matches you.