Meteorologists defend the value of predictions, although they acknowledge that they will never be one hundred percent accurate “There are forecasts that are still extremely complex,” they say
In meteorology there are no absolute certainties. Uncertainty is an inherent element in the forecasting process, due to the extreme sensitivity to the initial conditions of atmospheric modelling. Despite the fact that mathematical models have greatly improved in recent decades, there are forecasts that are still extremely complex to perform, as is the case of isolated depression at high levels (DANA), which has caused havoc in large areas as it passes through the peninsula. This is confirmed by Rafael Sánchez, a meteorologist from the delegation of the State Meteorological Agency (AEMET) in Galicia.
“Some situations are easier to forecast than others, and DANAs are probably among the most difficult.. Prediction models have improved a lot in recent decades, but there are forecasts that are still extremely complex to carry out,” says Sánchez.
The passage of the DANA through the Community of Madrid, where the heavy rains moved 50 kilometers with respect to the predictions, has caused other quagmires. Some politicians criticized the alert level –red, the maximum– declared by the Aemet, and the activation of the ES-Alert system of the General Directorate of Civil Protection and Emergencies 112 that last Sunday reached Madrid phones advising not to leave the house if it was not strictly necessary due to the extreme risk of storms, and demanded that the agency “tune more & rdquor; in the predictions.
Meteorologists regret that the reliability of the predictions and the activation of this massive alert have been called into question.
Climatologist, postdoctoral researcher and professor at the University of Santiago de Compostela (USC) Dominic Roye, Head of the Data Science area at the Climate Research Foundation (FIC); and Darío Redolat and Toni Rubio, climate change and meteorology consultants, call it “irresponsible& rdquor; open the voices that question the reliability of weather forecasts.
“It is very irresponsible to open a sterile debate on the reliability of weather forecasts”
Professor at USC, member of the Foundation for Climate Research
“In a context in which the appearance of extreme events is becoming more frequent, where there is clear evidence of a trend towards a greater concentration of precipitation in a large part of the Iberian Peninsula, it is highly irresponsible to open a sterile debate on the reliability of weather forecasts. This drift can motivate the distrust of citizens in science and in the knowledge that is behind the forecasts that move the notices. The conclusions of this public debate will depend on whether these words do not influence and sow doubt about the emergency alert system on future occasions,” the experts say in an article published by the Science Media Center (SMC). They also explain that the meteorological models will continue to improve in the coming years, “but they will never have a one hundred percent hit rate & rdquor;. “Small variations in the initial atmospheric conditions can lead to drastically different results,” they point out.
The Aemet meteorologist in Galicia explains that the basis of the predictions are numerical models: mathematical simulations of the behavior of the atmosphere. “By means of meteorological observation (surface observatories, atmospheric soundings, satellites, buoys…) the initial conditions of the atmosphere are determined and then the model tells us how it will evolve in the following days & rdquor ;, he explains. However, The atmosphere is a non-linear system, very sensitive to variations in the initial conditions., so that a small deviation in these conditions implies that the real and simulated atmospheres are rapidly separating in their evolution. “And, logically, we never get to know the starting conditions of the atmosphere with total precision, so we necessarily carry an uncertainty that normally makes weather forecasts beyond 10-12 days not useful. Uncertainty can also be estimated, which is why it is common to give predictions accompanied by a certain probability, especially when they are forecasts for several days from now,” he explains.
“There are situations that are easier to forecast than others and the DANAs are probably among the most difficult”
Aemet Galicia Meteorologist
According to Royé, Redolat and Rubio, knowing how to communicate uncertainty and not feed citizens’ mistrust regarding the reliability of predictions is important in order not to sow doubts about alert systems in the event of upcoming emergencies such as this latest DANA, and defend alert notices such as those sent via mobile in Madrid and are already used in other countries, such as the United States in case of severe storms. “It is necessary that, in the hours between the forecast and the moment when the expected event should occur, there is in-depth surveillance of it. This can be done with existing real-time monitoring tools (remote sensing, satellites, radar and weather stations). In addition, surveillance itself must be coordinated with civil protection and with the parties involved, following up on the event through an optimized protocol,” they say.
He Spanish protocol ES-Alert It is part of the European EU-Alert program, a population alert system (PWS) that works, in addition to extreme weather episodes, in other cases such as industrial accidents, fires and seismic phenomena.