Ensemble Forecasting

Written on: 02nd Nov 2021

Ensemble forecasting is a way for weather forecasters to assess the likelihood of a particular weather type in a particular area on a particular day. How are they created and what do they tell us?

Supercomputers are incredibly powerful, observations from all over the world are fed into the supercomputers several times a day, these observations come from planes, ships, radars and weather stations just to name a few. The computer then does some wibbly wobbly magic maths, analyses all the data that's fed into it and then creates a weather forecast. 

It is of course, impossible to know the exact conditions at every location across the planet, we can only feed data into the machine from areas where data is being recorded, so what happens in the areas between these locations? Well.. essentially, we make assumptions and educated guesses based on what is most likely to be occurring in that area. The starting data is tweaked slightly, so maybe a degree is added or subtracted to the temperature, humidity increased or decreased and so on, minor adjustments are made to the data and the supercomputer then runs the forecast again.. and again.. and again. The GFS model does this 30 times, the ECM model 56 times. 

The end result is 30 or 56 different weather forecasts showing possible outcomes to the end of the forecast period. The end result looks like this;

Every single line on the chart above is an individual forecast based on slightly tweaked starting data. On the left hand side of the chart is the present, the ensembles are tightly packed together and this gives high confidence in the forecast. As we progress into the future moving to the right, the lines become increasingly chaotic, less packed together and uncertainty in the forecast increases. 

There are many different types of weather that can add uncertainty into the forecast and ensembles are a great way to assess not only probability of a particular weather event, but also the confidence we as forecasters can have in the forecasts we are producing.