If there is something that we have all done with our smartphone – besides taking it out to look at the time, save it and forget what time it was – it is to look at the weather prediction to know, more or less, what time it is going to do, and what clothes It is convenient for us to carry.
And the truth is that we have many ways to consult the weather forecast, either through Google Assistant, the Internet browser, or through specialized applications. Surprisingly, on some occasions, the results differ according to where you consult the weather, and today we will explain why this happens.
Why is climate prediction not the same in all applications?
Right now, knowing what time you are going to do next weekend is an easy task for us, but the truth is that behind these predictions there is a very heavy work of gathering data about the state of the atmosphere, the temperature or humidity, and its corresponding application in a predictive model.
And, for you or me, time may be another variable, but for a person who is engaged in fishing, agriculture, or even construction, the climate is determined to establish the day and its duration, for What the weather predictions should be as accurate as possible.
However, measuring the accuracy of predictions is very complex, in the words of the founder of ForecastWatch, since according to the forecast model we use, there is more chance of hitting, for example, wind speed, but neglecting the accuracy of temperature, or, simply, have more chances of success in the long term than in the short.
A report from the same company compared the accuracy of the six main forecast providers, among which are AccuWeather or Weather Underground, in a total of more than a thousand locations.
This study revealed, for example, that AccuWeather predicts better the temperatures, the possibilities of precipitations and the speed of the wind, but, Weather Underground, on the other hand, uses a more precise model as for the minimum temperatures.
As you can imagine, the use of different models to determine the climate data, as well as the database available to each provider and the supervision to which these predictions may be subject are key when determining the climate. And therefore, depending on the provider used by the application where you look, you can find somewhat different data.
Logically, everyone has a margin of error, but this is not too big. You will never see that the temperature, for example, varies by 15 degrees depending on the supplier, so you use the one you use, it is unlikely that you are very misguided in your predictions.
But this margin of error is also influenced by rainfall, since, apparently, many suppliers tend to opt for predicting rain when there are doubts about it, and this has to do with the comfort of the user. And the fact is, the vast majority of people are more bothered by being told that it is not going to rain, and that it will finally rain, than the opposite. Although some companies are moving away from this so-called “wet bias” credit for the name of Peter Neilley, vice president of Weather Company – which, on occasion, both relieves us.
The data of your phone, the next step that providers want to give
On short time scales satellites are less accurate than in the long term, but barometers are incredibly useful for predicting impending changes based on the change in atmospheric pressure. And for almost seven years, many mobile phones have one.
And this, precisely, is the next step that many providers want to make, use the data from your smartphone to help improve the accuracy of their forecasts, and give you data with a lower margin of error. In addition, the transfer of these data would directly affect you, since, logically, they would be used in relation to the data of the place where you are. In fact, Weather Company assures that they would like to start using this data in 2019.