«It is a mistake that whoever decides to send an alert is a politician. “It must be automatic”

In fact, its HySEA ‘software’ is used by the governments of countries such as United States, Italy, Chile or Spain to predict, in a matter of minutes, the impact that a tsunami may have on the coast. His model was brought to Chile after the great controversy that occurred in the Latin American country in 2010, when the warning of a tsunami and an earthquake that caused 512 dead, 16 missing and 800,000 people victims. The tsunami was the second most important in the history of that country and the fifth in the entire world and, however, the president at the time, Michelle Bachelet, had ruled out, in a public appearance, that no tsunami was going to occur and made a call for peace of mind to the population. These statements were based on a diagnostic error by the Hydrographic and Oceanographic Service of the Chilean Navy (SHOA). “Many heads rolled there, and since then, our model is the one they use to avoid this type of catastrophe,” explains Macías.

Its prediction model is something like the Aemet (State Meteorological Agency) for tsunamis, with the difference that these gigantic waves move much faster than any atmospheric phenomenon and force quick decisions to be made to launch quick alerts to citizens. and be able to save lives.

HySEA has been praised by Americans, a population, says Macías, “who, unfortunately, and through experience, has learned to react to these phenomena.” In his opinion, the main lesson that Spain should learn is that Political profiles should not make decisions in times of emergencybut rather the protocols should be automated. Humanity, this is the case of Chile, learns from mistakes and our country, he says, must “make urgent changes to its protocols to prevent another DANA like the one in October from throwing the country into chaos.”

—How does the most reliable tsunami crystal ball on the planet that your team developed work?

—Here in Spain we work with the IGN (National Geographic Institute). The ‘software’ is used so that if the epicenter of an earthquake is in the sea, our simulation is launched and in a matter of minutes you can have an approximate idea of ​​how much its impact will be. Taking into account this, risk levels such as those of the Aemet are established. It is done by segments of the coast and the alert is sent to the population through a system like Es-Alert. This warning reaches people who are located on the part of the coast where the wave can reach.

—If there are models like this that allow us to react in a matter of minutes, why do we act so late?

—First of all, we must point out the antecedents that existed in our country. In the recent memory of Spaniards, a rather critical idea had been established regarding mobile phone alerts. In September 2023, let us remember, the first Civil Protection alert had been launched in Madrid and there were those who said that it had been excessive and inaccurate in its prediction because, in addition, it had ended up unloading in Toledo.

—So we should rethink the model that Aemet uses? Is it working correctly?

—Aemet is an achievement and its approximations are very tight. Aemet or the software that we devised for tsunami prevention are simulations based on mathematical models for geophysical flows. The difference is that the prediction of earthquakes or tsunamis is not subject to as much uncertainty as the atmospheric phenomena that end up producing, for example, a DANA. They are more chaotic physical behaviors, since precipitation can travel kilometers. There is a margin of error that will always exist and that can cause unforeseen damage to the population.

—Do you mean that we will never be able to 100% avoid a catastrophe no matter how many technical advances we implement?

—Technology can mitigate tragedy, but not eliminate it. In Spain we have sufficient resources but, however, what happened in Valencia has shown us all that the action plans have not worked and the reason is that there is a lack of automation, fixed protocols in decision-making.

—Is it better that the decision be made by a machine than by a politician?

—In a crisis context it is more difficult to make decisions about when to send an alert. These decisions must be made calmly, coldly, and not at the moment of the emergency, because they will be made incorrectly. No one, least of all a politician, should be in charge of giving the green light for an alert to be sent. It should be automatic: if Aemet foresees this scenario Y at X time for a population Z, the alert should be automatically activated. This way we would be giving objective answers and not basing action on subjective considerations.