Why is time so hard to predict?

At the start of civilization, people had to hallucinate with the variety of meteorological phenomena that took place on our heads. Rays, thunder, lightning, rain, wind, sun, clouds, rainbow, snow, hail … Everything was new and unknown! That is why our first instinct was to relate these atmospheric circumstances to the gods, to invent varied myths and to develop rites to calm the anger of deities.

Currently, science has enabled us to verify that all these elements arise due to very different causes. The air temperature, humidity, solar radiation, wind intensity and direction, seawater temperature, atmospheric pressure, cloud coverage or precipitation They are some of their conditions. We know it because we have been trying to predict atmospheric phenomena for years, not only to know if they should take the umbrella or sunglasses when they leave the house, but also because they play a very important role in navigation, aviation or agriculture, among other things.

«The more observation data we have, the more quality the weather forecast will have. That is why these parameters are registered by Multiple of spread weather stations over the entire earth’s surfaceAs well as through aircraft, satellites, teledetection instruments, globes-es and other instruments. The development of technology has been the key in the evolution of this science, “explains Rubén del Campo, spokesperson for the Spanish Meteorology Agency (AEMET).

All this information is based on mathematical models that start on the physical comparisons that control the movements of the atmosphere, and the calculations are responsible for powerful computers that greatly accelerate the work to meteorologists. Yet it is complicated of the prediction of time complicated, especially in the long term.

The reason is that “the atmosphere is a chaotic system that limits our ability to predict its evolution. Therefore, even with a lot of starting data, From the start there are small shortcomings and/or errors in the observations That, with the passage of time, more and more patents are becoming, “he says about the field. As a result, a prognosis is more reliable for two days than five.

Another important obstacle is that Mathematical models themselves also have their limitations. “Computers simulations are not perfect and can also lead to prediction errors,” the specialist adds.

The more the future we make the meteorological prediction, the more errors will collect and the less the conditions for the realistic will be sustained, so that, as time progresses, the prediction will always be far from reality. For all this, «The period in which the predictions remain valid around two weeks in ideal circumstancesBut it is usually less, between a week or ten days, “says the spokesperson. That is why, today, December 17, 2020, we cannot be sure whether it is rough or sunny in Reyes within 20 days.

Periodic updates

So that the prediction is as truthful as possible, The AEMET updates its time award four times a day (At 0 o’clock, at 6, at 12 o’clock already 18). For their part, adverse phenomena remarks are updated twice a day, as well as those that can be activated by observed phenomena. The rest of the predictions (provincial, regional, national, fire risk, ultraviolet radiation, mountain …) are updated every 24 hours.

And is it easier to predict time in one place than on another? «It is not so much dependent on the territory and the type of time is expected. When a stationary anti -clone dominates on the peninsula, we can predict the time in most regions for several days with sufficient reliability; But When a storm is approaching or a Dana is the thing complicatedBecause small variations in the direction of the wind, or in the position of the center of depression, can mean major changes in time at a certain place, “he explains from the field.

We can say ” The regions with soft and without large mountains, the prediction of time is ‘easier’Because the prediction models can simulate that surfaces with greater precision. On the other hand, in front of the places where the orography is complicated, there are times when the resolution of a model is not enough and this cannot interpret how a certain mountain interacts with the atmosphere, “he adds. All this tries to solve Increase resolution modelsThat simulates the lighting of a country better and which are able to predict small -scale atmospheric phenomena that can be very violent, such as storms.

In the future we will also be working on obtaining weather data from mobile telephone equipment and networks, and even The ‘crowdsourcing’ is increasingly taken into accountThat is, obtaining meteorological data that, by its scale, can go unnoticed by conventional observation networks and that are collected thanks to the involvement of citizenship. To do this, the Aemet has developed the web application SinobasWho has more than a thousand registered users and also has a Twitter profile (@aemet_sinobas).