Uncertainty in the weather
The first week of January has seen a widespread dump of snow across the UK, particularly here in Scotland. But how are our weather predictions developing with evolving technology?
A common issue pubic weather forecasts tend not to mention or account for uncertainty this word has different connotations depending on the audience. While scientists see uncertainty as a valuable aspect that informs and contextualises predictions and findings, the public often see uncertainty as a negative, implying that the scientists don’t know what they are talking about. This can result in the public not taking weather warnings seriously as the predictions were “uncertain”. On the other hand, when uncertainty is not talked about, variations in weather changes that deviate from the mean prediction are often scoffed at and the forecasters are held accountable for giving incorrect predictions.
This recent study from a group in China uses a combination of current prediction models while also explicitly representing three kinds of uncertainty, in a framework they call GenEPS (Generative Ensemble Prediction System). The results of this study showed a marked improvement in accuracy of weather event prediction and suggests that the more high-resolution data produced and observed, the more the framework can be trained, and the more accurate and acute future predictions can be.
This paper is a good example of how important uncertainty is in science!
Please read, rate and review this study here:
Congyi Nai, Xi Chen, Shangshang Yang, Ziniu Xiao, Baoxiang Pan