Forecasting Cut-Off Lows Events with MPAS: A Study of Valencia's Historic Rainfall in October 2024

Between 29 and 30 October 2024, Spain experienced one of the most intense and destructive natural disasters in its history, predominantly affecting the Valencian Community but also parts of the Murcia region and the province of Albacete. The floods impacted approximately 75 municipalities, affecting over 400,000 inhabitants, damaging around 100,000 homes and 137,000 vehicles, and resulting in a total of 232 fatalities across Spain, 224 of which occurred in the province of Valencia alone. This extreme meteorological event not only recorded the highest rainfall accumulation in Spain’s history, with 771.8 mm in just 14 hours at the Túris station in Valencia but also highlighted inefficiencies in the authorities’ ability to convey extreme danger alerts to the population. The State Meteorological Agency (AEMET) issued the alert at 07:36 on 29 October, but it was passed on by the local authorities only 20:11, approximately 12 hours after the event started and the onset of precipitation, which significantly increased the risk to the population. In this study, simulations were conducted using the NCAR/MPAS model with a global resolution mesh of approximately 92 km, which converged to a finer mesh centred on Spain with a resolution of 25 km. The resolution increase was smoothed due to the numerical scheme used by the model, which employs Voronoi hexagons. The MPAS was initialised with initial conditions from the NCEP/NOAA dataset, obtained at 00Z for the period 23–29 October 2024. The study aimed to evaluate how far in advance it would have been possible to predict the configuration and position of the centre of the Cut-Off Low (DANA), the atmospheric phenomenon responsible for the extreme precipitation totals. The goal was to determine how early the risk associated with the DANA could have been identified, regardless of the precipitation totals forecasted by the model, focusing solely on the atmospheric phenomenon itself. The MPAS simulations revealed that as early as 24 October, the DANA configuration could be identified, based not only on the position of its vorticity centre at 500 hPa but also on the intense moisture transport at 850 hPa originating from the Mediterranean, which surface temperature was approximately 2–3°C above its average, directed towards the Valencian region. This pattern persisted in all simulations initialised between 24 and 29 October, with some precipitation cores showing accumulations of 200–300 mm between 29 and 30 October in the Valencian region. Thus, this study encourages reflection on the extent to which meteorology should rely on precipitation totals forecasted by atmospheric models when issuing alerts and warnings, or whether such alerts could instead be guided by the configuration of specific atmospheric phenomena. This approach could potentially increase lead times, as forecasting wind fields generally involves lower uncertainty compared to precipitation. Such an increase in lead time could be crucial to save lives in extreme weather events like this one.