The first thing people often say to me after I explain what I do is: “oh, so you want to be the weather girl?”
This isn’t really true (I wan’t to be the female Brian Cox – there’s a serious deficit of female science nerds in the media) but I can sort of see why you might be fooled.
After several weeks (ok, months) of burying my head into annual reports I’ve finally had a bit of time to get back to some more science communication. I’ve just finished a course in weather and TV presenting, which was probably the most fun training course I’ve ever done.
It involved getting to grips with the way the weather forecast is presented live on air, as well as how it all fits in behind the scenes in the studio. Science-wise, we put together a presentation of the forecast for the next 24 hours, plus the outlook for the rest of the week, by delving into the model output. For me that was the most interesting part, because I use an operational weather prediction model in my own research, though of course for a slightly different purpose.
One thing that surprised me a little was how divergent the model solutions became, even just 18 hours ahead. For instance, one of the global models predicted temperatures tomorrow of around 27 degrees in London, whereas the French high resolution model figured it might hit 34. My best guess would be that it will be somewhere in the middle, but I will be interested to see how things pan out tomorrow.
I suppose that was one of the take-home messages for me. Knowing which models do best in which situations can really help meteorologists choose which models they give more weight to. It takes time to figure that out, and that expertise is priceless for accurately predicting the weather.
Another thing I thought was interesting (and useful for modellers to remember!) was that high resolution models don’t necessarily produce the most accurate results. In fact, sometimes lower res models like GFS (the American global model at 0.25° resolution) got the synoptic situation over the UK better than the higher res AROME or ARPEGE models (at 0.01° and 0.1° resolution, respectively). The higher res models got more complex features, but sometimes these were misplaced or even spurious, i.e. showing something that wasn’t there in the observations.
From my own research I know that there is a benefit to using high res models when you have complex terrain, or other fine-scale features that are important to resolve because they influence the way the atmosphere behaves. However, it also highlights the fact that models are imperfect, and we must use them as a tool, rather than taking them as gospel. Like I said, I found the way that ‘proper’ meteorologists used a variety of models to get the big weather picture really inspiring – it requires them to understand the limitations and strengths of each model, not just in sum, but also in different situations.
It was also great to hear from a TV reporter about what sorts of things TV journalists want to hear from scientists and experts. As with science itself, it’s important to be concise. Keeping things simple, but above all short, can really help communicate your meaning effectively. It’s something I definitely need to work on.
I think my TV career might require a bit of work too – it’s hard being in front of a camera and not stumbling over your words! Turns out it’s also really hard to reach Scotland on the green screen when you’re short like me, but I’m not sure I can really do anything about that.
In the meantime, I guess I’ll just work on my dazzling TV presenting skills…