It’s safe to assume that we’ve all experienced being caught in a shower. It’s one of life’s many trials and tribulations. You see a lovely Spring day and want to experience the reenergised sunshine after a long Winter huddled in a blanket. You bravely venture outside for a walk and, 10 minutes later, the skies open up and you return soaked to the skin. You may curse the weather app or the weather presenter on the TV before your ill-fated decision and you’d rightly ask, “why can’t they get anything right?”.
If I had a penny every time I’ve been asked that very question. Of course, it is a valid question. There have been so many advancements in technology, weather models, satellites, and observational networks over the past few decades that it’s reasonable to presume that we should be able to predict when and where a shower is going to develop.
The short and frustrating answer is that forecasting the development of convective weather events, such as showers or thunderstorms, is a lot more difficult than you would expect. To start, we have to understand how convective showers and thunderstorms develop in the first place. This requires three ingredients; instability, lift, and moisture.
Instability is describing how well mixed the atmosphere is with respect to temperature – warm air wants to rise and cold air wants to sink. Generally speaking, warmer air tends to be at the surface with colder air at higher altitudes. A poorly mixed atmosphere with high temperatures at the surface produces buoyant air parcels. However, we require our second ingredient of lift to force this warmer air parcel upwards. This tends to occur from solar heating from the sun, but can also be caused by the air moving over higher ground pushing it upwards or through instability in a weather front. Once the air parcel has been forced upwards, it continues to ascend in the atmosphere due to its buoyancy. This is where our final ingredient is necessary. As the air parcel rises, the moisture it holds condenses to form clouds, eventually forming raindrops, and precipitation. If there’s no moisture, no clouds or rainfall can form.
If we know how these convective systems develop, you may think it would be a simple jump to predicting when and where they will form. Unfortunately, there are a number of factors that contribute to the uncertainty of the predictions. Weather modelling requires high quality, accurate observations and although improvements have been made through satellite and commercial aviation, the spatial extent is still poor, especially over the oceans. Secondly, although model resolution has increased significantly, it can still be difficult for these to identify and resolve small scale processes. The final significant factor is our mathematical understanding of these events. We may understand how they develop from a generalised theoretical perspective, but being able to successfully mathematically model every atmospheric, oceanic, and land process is extremely difficult.
The best analogy for these convective events is thinking of a boiling pot of water. The water is the atmosphere and the heat from below is replicating the sun. As the water is heated, the temperature at the surface of the pot is higher than that above (instability). As we heat the water further, bubbles escape and rise upwards as they are more buoyant than the surrounding water (lift). The more heat applied, the more energetically the bubbles rise – think of this as the difference between showers and thunderstorms. Attempting to model and predict the exact location and timing of where all these bubbles in the pot will develop and rise is an almost impossibly difficult task requiring us to know the micro currents in the water, when and where the most intense heat is occurring, how one bubble may impact on the development of another, and a myriad of other factors. Scale that up to the entirety of the UK and you can begin to understand why predicting these events, even a few hours ahead, is very difficult.
So what can weather models and forecasters do? Improvements are continuously being made to our theoretical and mathematical understanding which feeds into the numerical weather models. Although these models struggle to resolve the exact locations of these showers, they have improved significantly in predicting the spatial area of impact, the frequency of shower development, and their intensity. Probability forecasts try to convey the potential risk of experiencing rainfall in your given location, and the use of multiple forecast models/apps can also provide forecasters (and you) a way of understanding the uncertainty in the area of interest.
Unfortunately, this may not stop you getting an unexpected soaking but here are a couple of things you can do to reduce your chances: