The Angry Corrie 66: Nov-Dec 2005


Chart rundown?

Ken Stewart on an unsettled spell for the BBC weather forecast

BEFORE THE BBC's "Change in the Weather" was introduced in the spring of this year, I had given little thought to what changes should be made. Perhaps I should have done, as the subject had been well aired in general terms in the magazine Weather (see http://www.royal-met-soc.org.uk/weather.html). Views ranged between what those with special and professional interests might want, and what would best suit what one of the letter-writers called "Tony and Tina": the ordinary public. There was even the purist view that no forecasts should be broadcast at all as they couldn't be done properly in the time - but then you can say that about news as well. In retrospect, clearly things were ripe for change as the previous pattern of forecasting had existed for many years and the capabilities of computer graphics have hugely increased in recent times. All the same, I'd missed much of the pre-publicity, so was only dimly aware of what was about to happen and the actual changes came as quite a shock.

On 15 May the initial impact on me (and on others I spoke to at the time) was one of horror, at both the TV forecast and the website version, http://www.bbc.co.uk/weather/ We suddenly had a map with a tilt which made Scotland appear very distant. The English Channel took up more screen-space than Scotland. The sunny areas, portrayed in a sandy brown colour, appeared desert-like. The rainfall graphics were misleading for showers: the blue puddles were essentially random in showery areas, so a static graphic could mislead the unwary into thinking they would avoid rain.

Early on, I thought BBC forecaster Jay Wynne would fall over as he ducked and weaved to keep up with the mad graphics. The omissions were serious: no synoptic charts, no wind arrows (even though the letter-writers in Weather - before the event - had emphasised that these were the key drivers of weather). The website also took a big turn for the worse. No Atlantic charts (and very amateurish UK charts), no satellite images, no 24-hour rainfall history, and charts which required a "full-screen" setting to use the controls. The practice of using a "snappy" headline such as "Dry for most" or "Rain in the north, sunny elsewhere" was - and remains - fraught with risk. I was not amused by one lunchtime "Hot and sunny" headline when I was sitting indoors heavily clad and with my heating on.

image from source document

These were all details, however. The huge overall objection was that a deliberate decision has been taken to deny the public basic information with which to form an opinion on the likely weather in their own context. Instead, the setup was now such that we had to accept what the forecaster told us without the information to put it in context. An analogy would be with ancient priests who understood eclipse timings but made sure the people were not allowed such information.

These points were broadly ignored in the responses which appeared on the "Ask Helen Young" feature on the website; predictably, it stuck to more bland and superficial issues. As I had heard that the BBC were saying people felt "disenfranchised" by seeing synoptic charts that they did not understand, I wrote in and turned this round to people being disenfranchised by their absence. The BBC's response carefully avoided the issue, but said their research showed that: (a) "the audience wanted more clarity and more detail", and (b) "people were often confused by the amount of information within a forecast". This seems to me typical of the contradictions that may be drawn from such research.

Around this time (early June), improvements began to creep in. The tilt of the UK graphic was improved - though it's still unsatisfactory. The BBC said the original aim had been "to show the horizon without omitting any land areas and had the advantage of clearly showing the transition from day to night" - never mind the weather! Website synoptic charts were also now properly labelled, and Atlantic charts came back. On TV, BBC Scotland led the way with Heather Reid referring to the "good old isobar chart"; forecasts networked from London have now followed. Wind arrows have also quietly been restored.

Problems for hillwalkers and climbers remain. Typically when away from home we have no access to the internet and have to rely on TV and radio. The Radio Scotland mountain forecast was for a long time broadcast at 6:59pm, and as such tended to clash with the 6:55pm TV forecast; also, it was on FM only for much of the year and hence not receivable in mountain areas. This has recently improved: the mountain forecast has switched to 7:13pm at the end of the news bulletin, and is now available on AM.

So where are we? The worst aspects of the change have been mitigated and there has clearly been some response to the protests; but why did anyone at the BBC ever consider the 15 May version to be acceptable? In response to 2½ pages of highly critical letters (including one of my own) in the August issue of Weather, the BBC's Andrew Lane made the astonishing claim that "we have not changed our policy on the display of pressure charts, despite
evidence that most people do not understand them."

Changes are still being made, for example the interval on the rainfall forecast has become three hours rather than one, and the forecast period has been extended proportionately. The BBC claim they have a new system which ensures consistency between forecasts (it will "eradicate national and regional discrepancies"), yet I still find cases where the text forecast is seriously at variance with the icon forecast. The "engaging" graphics give an illusion of accuracy - for example there have been website graphics suggesting that rain would reach into central Scotland well after it was obvious this would not occur. And overall, there remains the suggestion that we have to take "their" word for it, rather than being encouraged to think for ourselves.

At least what we have now is reasonably useful, but why did it take such a route to get there?


Ed. - Two additional thoughts. The change to a colour-coded chart rather than a symbol-based one has made the TV forecast almost unwatchable for anyone with a black-and-white set - and there are still a large number of these in use, for instance as second sets in domestic situations. The brown sun-shading is particularly hopeless, recalling the nightmare of trying to distinguish the teams when watching Ireland (green shirts, white shorts) play rugby union against Scotland (blue shirts, white shorts) on a black-and-white set in the 1970s.

More generally, the BBC - at least in its initial 15 May format - seems to have given away most of the advantages that made it a better forecaster than ITV/Sky/Five, and to have voluntarily descended to their level. It's as though one of the broadsheets decided to turn tabloid because they reckoned no one should be faced with words of more than two syllables. Even now, several months on, I keep half-thinking I've tuned into the latest reshowing of To Die For, where Nicole Kidman plays a forecaster in all-teeth-and-handwaving mode. This is particularly the case with Sarah Wilmshurst and Carol Kirkwood - excellent meteorologists though they no doubt are - as both have a tendency to look and sound like they've dropped in from a children's channel.

Forecasts should be distinctive because of the quality of information, not because of how telegenic the forecasters are, or how slick the latest software. The dread words "dumbed" and "down" ought not to be an issue, as weather forecasting is one place where style must not be allowed to triumph over content. Although it might never happen that a walker, climber or kayaker gets in trouble through having ventured out on the strength of a poorly presented, needlessly generalised or plain wrong forecast, this is perhaps a little more likely now than it was before May.

Incidentally, there does seem something very predictable about UK forecasts declining in quality at the same time as broadcasters become fixated with American weather.


TAC 66 Index