0 citations0 references

Communicating forecasts, or the art of the weather report

Weather2012Vol. 67(3), pp. 68–69

Citations Over TimeTop 20% of 2012 papers

Abstract

‘From Observations to Forecasts’ was an interesting and informative series and it was fitting that Part 14 (Eden, 2011) should address the issue of communicating forecasts. Good communication is essential to rational decision-making: without it forecasts have little or no value. Following retirement after 38 years at the sharp end of public service forecasting in the Channel Islands, I would like here to add a few comments in response to Eden's call for debate. An aviation analogy explains how I view the state of communicating forecasts. In the 1960s, most new commercial aircraft were, in essence, flown manually, but modern aircraft can fly from aerodrome to aerodrome with minimal input from the crew. However, as you take a seat you can often see your bags being manhandled onto a truck: that process has changed little over many years, your bag still takes a beating in getting from A to B. A computer-age concession is the barcode to aid recovery of bags. In my view weather forecasting is in a similar position. As a scientist I know that, by any measure I choose, forecast accuracy has improved greatly in the last 30 years, yet blogs, newspapers and laymen complain from time to time that forecasts have barely improved. Why this apparent contradiction? I believe it is due in part to indifferent communication: as with the aviation industry, one part of the forecast process (accuracy) has improved greatly, while another (communication) remains on a par with the transporting of bags on aircraft. The forecast process is a chain of events and, like all chains, is no stronger than the weakest link – communication. There is a gradual recognition of problems in this field, as the following examples indicate: the WMO Bulletin published a paper by Gill (2008) on communicating forecast uncertainty; in 2009 the BBC Radio 4 national forecast at 5.57pm experimented with various techniques such as shipping-forecast style, with background noises, using regional accents etc (at least it was a start); in March 2010 the Royal Society discussion group brought together working scientists from various disciplines to discuss uncertainty (Royal Meteorological Society website news 13 December 2011) and in June 2010 Meteorological Applications produced a special issue Communicating weather information and impact. I add to Eden's comments from experience at the sharp end, referring mostly to BBC national radio and television broadcasts and the UK Met Office (UKMO) website as examples, for the UKMO regards itself as a world leader. On 21 December 2011, a press release stated that it is regularly ranked in the top two national weather services in the world, though no source was quoted for this bold statement. Warnings are the priority in any forecast office. Where there is a danger to life, of structural damage or disruption to everyday life, a warning message should be prominent and unequivocal. As I write, I open up the UKMO website and find warnings on a cluttered page amongst adverts, twitterfeed, climate change, events and news. The top line announces ‘cold weather alert level 2’ but on the warning map for each day there is no mention of cold weather – most of England and Wales are in for strong winds. Is that clear and unequivocal? No, you have to search for it and I suggest it is not communicating the warning effectively. In contrast, every weather office home-web page in the USA opens with a prominent colour-coded warning map. I cannot exclude my former employers from criticism for, despite requests for one for some years, the Jersey Met Department website has no warnings section. Having made live BBC local radio broadcasts several times a week for 30 years, like Eden my style evolved. I followed a maxim I was told comes from the Army: Tell 'em what you're going to tell 'em, tell 'em, then tell 'em what you've told 'em. Why? It's simple. The brain can only assimilate a limited amount of information at once. By repeating it, there is a greater chance of the forecast being retained in the user's memory. I highly recommend a paper by Ayton (1988) to radio broadcasters. He discussed research on listeners' powers of recall and advised the need to speak more slowly when imparting information. I often think of his piece as I listen to forecasts on the radio, spoken at such a pace that I wonder if there is a prize for the most words used per minute. With pace goes that precious but near extinct part of speech, the pause, a short silence giving the brain time to assimilate the information just heard. I argue that in not pausing and slowing speech slightly, communication is less effective than it should be, as well as discourteous to the listener. Not having presented television forecasts, I make comments as a viewer and user. There can be little doubt that television forecasts have been ‘dumbed down’, but why and for whom? A simple example is satellite images labelled ‘Friday’, as if the viewer is incapable of telling the time. Animated sequences are played once, so by the time a viewer has located the UK and found an interesting development the animation has ended and the effect is lost. For some years pressure charts were banished from national BBC forecasts, though they never disappeared entirely from BBC Southwest and are making a comeback. How their demise aided communication is not clear. Television graphics may be clever, but do they communicate the forecast effectively? It is not clever to find in BBC national forecasts that the presenter is talking about northwest England but the camera has reached southeast England; I get so irritated I give up. Those inclined to seasickness may have looked away earlier! Further, the ‘green and pleasant land’ is a dirty, desert-camouflage colour. Clouds, white or shades of grey on satellite images, are cowpat brown and land the colour of sand. On leaving the coast, clouds (I think they are clouds) turn dark blue, a colour which depicts deep water in an atlas. As a forecaster I cannot visualise these depictions as clouds; what hope for the non-specialist viewer? Recently ‘the Weather Show’ (25 December 2011) showed winds rotating clockwise around high pressure in the southern hemisphere, suggesting graphics developers are in control, not meteorologists. The priority on television forecasts (not only the BBC) seems to be the display of clever graphics with little thought for forecast users. The media, especially television, by turning forecasters into ‘personalities’ or ‘entertainers’ undermine their authority as scientists which has also led to increasing subjectivity, something I always tried to avoid. The hot July day may suit BBQ fans but for hay fever sufferers and many of the very young and elderly it is a day of misery, as well as a day of anxiety for the farmer needing rain. Why not tell it as it is and avoid subjectivity? The view that pilots want to self-brief is debatable; the reality is self-brief or no brief. In Jersey, some local commercial pilots use the forecaster-briefing option. They find it very useful when fog, thunderstorms or gales are occurring or are forecast and those phenomena cannot be communicated fully in an aerodrome or flight forecast. Viewing developments as a series of satellite or radar images, with a forecaster's explanation, is a good aid to decision-making, or so pilots told me – whether to load extra fuel, change route, delay take-off etc. A good verbal briefing reduces the risk of encountering the unexpected, but in many parts of the world this is not an option. The occasional pilot who stumbles upon the Jersey office is usually astonished that such a useful service exists. In 38 years I never had formal training in marine forecasting, though there are many thousands of small boaters around our coasts. The UKMO Training College ‘doesn't do shipping’, as a worried trainee forecaster reminded me last year. The scope is limited in communicating forecasts of this type, but there are subtle ways of indicating the uncertainty which lies behind every forecast. Instruction in this field would make for a more rounded forecaster and better communicator. The internet has opened many possibilities in the world of weather, be it observations or forecasts. Professional forecasters have some excellent sites from which to interpret the likely sequence of weather, but I argue that there are serious shortcomings in the presentation of general forecasts, with lack of thought for the non-specialist user. Basic forecasts of temperature and wind speed display single numbers. Mini-mum, maximum or mean temperature – the user must guess. Is ‘today’ the period 6am to 6pm or midnight to midnight? Is wind speed a mean for the day? How are changes in direction shown? Recently a colleague took a complaint from a local boater that the local forecast of force 4 (11–16 knots) was much less than a website which indicated ‘30’ and should there not be a wind warning in force? Investigation established the units were kilometres per hour but it was far from obvious to the non-specialist user, another example of poor communication. Few sites have a legend of the weather symbols. On sites where manual input is required, what single symbol should one choose if the day starts sunny, becomes cloudy by midday with rain forecast by 5pm? There must be more creative ways of using graphics to communicate effectively. Having indicated some shortcomings, I offer a few ideas on the way forward. All forecasters should be required to undertake modules in communicating forecasts, just as they take modules in basic principles of meteorology. The object of communication should be to transfer the thoughts in the mind of the forecaster as exactly as possible into the mind of the forecast user, a skill needing training. Not all forecasters undertake media work, but all need good communication skills. Research on communicating forecasts exists, such as Ayton's (1988) paper, but recent rapid technological advances certainly leave the field ripe for further research on effective communication. I suggest specialist university departments may be well-suited to research in this field. Behind every forecast lies a degree of uncertainty but using percentages to convey the degree of confidence has not found acceptance in general forecasts. Underlying problems aside, percentages are widely used in North America, where there is extensive research in this aspect of forecasting. Perhaps the problem lies in using the word ‘uncertainty’. Would ‘risk’ be more appropriate for the general user, or odds as in the betting world and more generally understood by the public? Customers, the public, forecasters and graphics designers should all be involved in new developments. The UKMO was recently asking for comments on a beta version of their website. Would it not have been more effective to involve all concerned from the start, rather than react to ad hoc comments later, for that is the implication of the request? Knowledgeable weather enthusiasts would surely give of their time at seminars or workshops with ideas on improving forecast communication. Designers of ‘graphics packages’ should not be afraid to develop ideas used successfully by others. The adoption of a traffic-light warning system by UKMO, similar to that used in France and elsewhere, is a good example. Forecasters concentrate on science, which has allowed website designers and specialists in computer graphics to lead without much reference to the forecaster or much thought for the user. Forecaster representatives should be more forceful in pushing the view that a forecast is primarily to inform a decision-maker, rather than to entertain. Strong commercial pressures in the media should not be an excuse for lack of balance. A final thought: seasonal forecasting is no place for a weather forecaster as it is in the experimental stage and the climate change debate has more than its fair share of sceptics with entrenched beliefs. The Editor writes: Both Frank le Blanq and Philip Eden have raised important points about the communication of our subject, and further constructive responses from readers are welcome. In the next two issues (as currently planned), we have articles relevant to this theme on the preparation of forecasts and on the Met Office's Public Weather Service Customer Group. The topic of communicating weather information to the non-expert is vast and fraught with pitfalls. There is likely to be a huge range of views within the RMetS as to how this is best done, and those professionals ‘at the sharp end’ will doubtless sometimes feel they just cannot win. During the very cold weather of early February, a short letter in The Daily Telegraph criticised forecasters who said that this weather had come ‘straight from Siberia’. At best, that criticism was nit-picking: an intense cold pool had moved out from the Siberian Arctic into central Europe, and the British Isles caught the edge of the resulting extremely cold surface air. A better target for the writer (though they would not have printed it) would have been the very poor quality weather data (both forecasts and reports) that appear on the back page of this paper every day – not to mention the standard of media reporting of severe weather. We have a sensation-seeking media, as well as a plethora of weather forecasting concerns (some of very dubious credentials) seeking to be heard. How do we ensure that only the best-available advice at any time gets to the public?

Citations Over TimeTop 20% of 2012 papers

Abstract

Related Papers