In numerical weather forecasting there are two types of forecasts. Deterministic and Probabilistic forecasts. Both of these have their strengths and when you look at most forecasts which tell you the expected weather tomorrow whether it's on a chart or a data table then chances are it's a deterministic forecast which shows you how the model thinks the weather will develop over the next few hours and days.
The problem with deterministic forecasts is that they are either right or wrong. It's a little like today where some of the models show a cold front coming in from the West and some of them releasing the CAPE energy across Central and Eastern areas and developing some sharp showers and some of the models not going for that solution and keeping the atmosphere nice and stable.
In scenarios like this, meteorologists can either total up the models which go for the showers, perhaps they favour a certain model which handles these scenarios well and then release a forecast based on that, or they can use probability forecasts which is where a model is run numerous times and the average is then taken to show us the chance of something happening or not.
The Americans have used probability forecasting for years, but it's still something which the British still seem to lag behind when it comes to understanding the risk of rain, cloud, snow etc over the next few days.
Now though, you have a brand new tool from the DWD which is called the ICON EPS probability forecasts. These show you not just the ensemble mean of things like cloud, rain, temperature etc, but they also show you the probability of certain thresholds being reached. For example, what is the chance of rain tomorrow afternoon? Or what is the percentage risk of winds above 40mph next Monday?
Have a look at the new ICON EPS forecasts and take a while to get comfortable with probability forecasts as you are likely to see far more of them over the next few years as ensemble forecasting becomes more and more powerful.
METEOROLOGIST : MARSH
|