Towards mega-torrential rainfall?

The catastrophic events caused by the DANA (cold drop) on October 29, 2024, in Valencia (Spain) lead us to reflect on whether we are facing a trend toward extreme rainfall events never before seen. The mega-rainfalls of 10/29/24, as well as the resulting flooding, caused 228 deaths, as well as extraordinary economic damage. How could such a catastrophic outcome have occurred in a country like Spain with a long tradition of weather forecasting? Are we facing a mega-event, never seen before?

The province of Valencia has suffered torrential rainfall events for as long as recorded memory exists. The extreme rainfall that occurred in Alcira in 1864 is a memory that still remains today. Valencia is a land where "rain doesn't know how to rain" (Raimon, 1984). Extreme events have occurred in 1957 (1,414 hm³), 1982 (2,225 hm³), 1987 (2,656 hm³) and 2000 (2,196 hm³). From this perspective, the 1,526 hm³ accumulated in the province of Valencia in 2024 do not represent an absolute record. However, the torrential rains of October 29, 2024, had catastrophic effects. Why?

It's true that political and administrative errors occurred, but the event's uniqueness stems largely from the fact that it occurred in an area where it had never rained to this extent before. DANAs have a very localized territorial effect. They can occur in areas where they have never occurred before.

The research seeks to evaluate the return period of rainfall produced at the head of the Chiva torrent (Poyo torrent), where accumulated rainfall in 24 hours exceeded 600 mm. The traditional methodology (based on the Gumbel distribution, using the data series available at the Chiva station) predicted a much lower rainfall than recorded: 248 mm. Meanwhile, the gridded data generated by Copernicus (E-Obs, 0.1°) show an even lower limit for 500 years: 138 mm. From this perspective we would certainly be facing a mega-event of extreme rainfall.

The results suggest the need to develop more detailed risk maps. Maps that consider not only the rainfall recorded in a specific location, but also its immediate geographic surroundings. They also highlight the enormous difficulty of developing gridded databases like E-OBS when there is not a large number of meteorological stations in each area.