Three years after the tragedy that shook Dolomite mountaineering, a new scientific study sheds light on the causes of the Marmolada glacier collapse on July 3, 2022. The study, published in Natural Hazards and Earth System Sciences, analyzes with remarkable precision the dynamics that led to the rapid detachment of almost 70,000 cubic meters of ice at an altitude of nearly 3,200 meters, along the regular route to Punta Penía. The catastrophe claimed the lives of 11 people and left at least 7 seriously injured. The study was conducted by an international and multidisciplinary team that included Italian and European universities and research institutes such as the Universities of Parma, Padua, Trieste, Zurich, the National Institute of Oceanography and Experimental Geophysics (OGS), ARPAV, and the University of Stellenbosch. The collapse occurred in a small glacial cirque below Punta Rocca, noted for its instability. According to researchers, the mass of ice detached along a slope with inclinations of up to 40 degrees, covering more than 2.3 kilometers at an estimated speed of 80 to 90 km/h. There were no seismic signals recorded in the vicinity of the occurrence, ruling out the possibility of an earthquake. Instead, the collapse would have been caused by a combination of destabilizing factors: abnormal melting of snow and ice as a result of the previous months' record temperatures, the presence of large amounts of water trapped in deep blocked crevasses, the degradation of permafrost in the underlying rock, and the glacier bed's unfavorable geometry, which included steep slopes and unstable debris. "The glacier was in critical conditions: internal temperatures were high, the base was unstable, and the pressure exerted by the trapped water contributed to disturbing the equilibrium", says Aldino Bondesan, geographer at the University of Padua and the study's primary author. The researchers used modern technology to obtain their conclusions, including ground penetrating radar (GPR), geoelectric measurements, terrestrial laser scanners, LIDAR drones, and high-resolution satellite images. Using indices such as NDWI, multispectral analysis enabled them to detect the presence of surface water and water that had been incorporated into the glacier mass. In addition, core drilling was conducted, and sensors were implanted to monitor the temperature at depth, detecting values between -2.4°C and -3.1°C, which are indicative of a cold glacier that is near the melting point. Finally, the glacier's stability was computationally simulated using the Limit Equilibrium Method (LEM), a geotechnical engineering technique. Only the simultaneous aggregation of numerous risk variables reduced the "safety factor" below the critical threshold of 1, indicating the final breaking point.