I must admit it really bugs me when people talk about weather presenters "reading the forecast", the implication being that they are simply reading out a script and that they know nothing whatsoever about meteorology.
The truth is the vast majority of broadcast meteorologists, to give us our proper title, have studied at institutions like the Met Office and many have spent time forecasting on RAF bases.
To be honest, I don't really mind being called a brolly dolly or a weather girl, quite frankly it's nice to be thought of as a girl given that I'm a married mum of two closer in age to 40 than 30. But there is more to the job of weather presenting than meets the eye, so here's my insider's guide.
If you want to be a good weather presenter then some knowledge of weather, climate and forecasting is essential. My job is to translate all the meteorological jargon spewed out by the forecasting models into everyday language, so that you know whether it's going to be hot, cold, wet or dry.
MORE FROM THE WEATHER CHANNEL UK: The rainiest places on Earth
No matter what TV channel I'm working for, or what time the shift starts (4.30am for breakfast telly!) the first thing I need to do is get on top of the weather story. That means I need to know what's just happened, what's happening now and what's going to happen.
The best way to do this is by looking at synoptic charts and reading the guidance - both written and pictorial forecast output from various computer models. At the same time I keep an eye on current weather observations and check the radar and satellite imagery to see if what's actually happening matches the forecast.
Most weather forecasts are quite detailed for the next 24 to 48 hours and then become more generalised for the days that follow. As for how next week or next month is looking, it's not possible to give specifics. The limitations of forecast models, supercomputers and the intricacies of our green and pleasant land (we are, after all, a small mid-latitude island sitting at a crossroads for various weather patterns) means it just isn't possible to give anything other than an idea of nationwide trends - such as warming up, cooling down, turning wetter and so on.
The good news is that accuracy has improved over recent decades with a four-day forecast today as accurate as a one day forecast was 30 years ago. Who knows what the next 30 years will bring?
In the meantime, now that I'm up to date with the forecast it's time to sort out the graphics to help me tell the weather story. Long gone are the days of hand-drawn pressure charts and magnetic weather symbols, indeed some broadcasters don't use symbols at all and prefer computer animations of cloud, rain and snow. I like to use a mix of both where possible.
The problem with symbols is that it's pretty hard to sum up the entire day's weather at a specific location in just one symbol. It might be sunny all morning, cloud over later with one or two showers at lunchtime then dry up for the afternoon. If you pick the shower symbol then people might wrongly interpret that as likely to rain all day. This deterministic versus probabilistic style of forecasting often leads to cries of "you got the weather wrong".
In theory, showing the movement of a weather front across the UK allows viewers to see what time the rain is likely and gives them a fuller picture of the day's weather. But again there are problems with this as the timing of the front might change, for example.
And that's where presenters come in. Once we've got the weather graphics ready it's up to us to explain the charts and point out the areas of interest, mention uncertainties in the forecast and add some extra value. The temperature disc might say 17C but will that feel warmer or colder than the day before? What effect will the wind have? And so on.
You might think the next stage would be to write a script to accompany the weather graphics that would then be loaded onto autocue so we could read out the forecast in the same way that news presenters read their scripts.
There are a couple of reasons why this doesn't happen. The first is because most weather presenters stand in front of a blank green screen and a process called "chroma keying" replaces anything green with the weather graphics so that the final image you see on your TV shows the presenter in front of the graphics.
Look at me
In reality, there is nothing behind me so I rely on monitors in the TV studio to see what charts I'm on and where to point - and when I look into the camera autocue system, instead of showing a script it shows me in front of my graphics exactly as viewers see on their TV screens. It's worth noting here that the image I see in the camera does not move like a reflection in a mirror, instead it goes the opposite way, which can take some getting used to. There's often a slight time lag as well which is also rather disconcerting.
Green-screen technology means no green clothing and even certain shades of blue can look funny. Not all weather studios use green screens, some use grey or blue and some have large TV screens or projectors behind the presenter so they can turn round and actually see their graphics (instead of faking it). So why don't we use autocue scripts when there's a real screen there?
The other reason is that weather broadcasts are often used as buffers in programmes. We might get told we've got a couple of minutes but if the item before us runs over and the next programme must start at a specific time then that two minutes can be chopped at very little notice. This is the case on BBC Radio 4 for example where one must absolutely not crash the top of the hour and the famous pips.
On the other hand, it might be that the programme has run light and so you need to fill, or perhaps a guest for the item after you hasn't arrived yet so please just keep talking. Hang on - who is asking me to keep talking?
Voices in my head
Whether I'm doing a live or a pre-recorded weather bulletin I wear an earpiece so I can hear everybody in the gallery. The main voices in my ear are those of the director who cues me and the PA (production assistant) who gives the timings. Some people like closed talkback so they only hear messages for them, but I like to be on open talkback so I can hear everything going on in the gallery and if chaos is about to happen I get a headstart on what I can do to help.
Being able to talk to time is a skill, so it helps that we ad-lib rather than read a fixed script and it does get easier with practice. The down side of ad-libbing is that we can accidentally say the wrong thing, especially after a run of early starts or night shifts. It might be a simple case of confusing east and west, the tongue-twisting fog and frost, or the more dangerous spoonerism of the Kent countryside...
All in all, being a weather presenter is harder than you think. We aren't told what to say, we do check the forecast, we prepare our own graphics which we operate ourselves using a clicker while ad-libbing to a duration that can change mid-broadcast. We point at charts that aren't really there while voices in our head shout out numbers so we know how long we've got left.
I'm not complaining though as I love my job - even if I do get the blame when it rains!