Research Letter #19


This Letter starts with a simple quiz - two multiple choice questions from an article in the US-based National Weather Digest (Sink 1995). The questions, in the original study, were posed to 475 people. They were phrased to determine how the public interprets weather reports. These questions are not meant to trick us but to highlight the difficulties in sending and receiving a weather report. You will probably be inspired to think carefully before you choose an answer; do you think carefully about what the forecast means when no quiz is at hand? I am sure that I do not need to emphasize the importance of weather forecasts to fire people!

1. The weather forecast is for "a 60% chance of rain today". Do you understand this to mean:

A. Precipitation will occur 60% of the day
B. At a specific point in the forecast area (e.g. your house) there is a 60% chance of precipitation occurring
C. There is a 60% chance that precipitation will occur somewhere in the forecast area during the day
D. 60% of the forecast area will receive precipitation and 40% will not

If you picked "C" you were with the majority response (82%) but the majority was wrong. The correct answer is "B" but I suspect that the "point" in question should be the rain guage at the official weather station because this is where the rainfall result would be officially evaluated rather than at "your house". Of course, your main concern may be what happens at "your house" (or other location) so the forecast might be related to, but perhaps other than, "60%".

Here is a second, related, question from the article but I have reduced by two the number of choices given in the original.

2. Which of the following is a correct interpretation of a '25% chance of precipitation today'?

A. 1/4 of the area will get rain today
B. Out of 100 days like today, on [approximately] 25 of them rain will occur
C. There is a 75% chance that some form of water will fall from the sky today
D. 3/4 of the sky will be sunny today

If you picked "B" you were with the majority response, and you were right. The author concluded that: "The uncertainty associated with the precipitation forecasts is magnified when people have misconceptions about the meaning of the numbers and words used in the forecast". There is a whole raft of issues associated with the interpretation of the forecast and what actions are warranted as a result of it (Nicholls 1999). Let me quote from just one aspect, of many, that Nicholls (1999) - of the weather research centre in Melbourne - considers. The issue is the interpretation of El Nino forecasts. In Australia the prevalence of media stories relating El Nino to serious drought (e.g. the 1982/83 situation) now produces a public expectation that any predicted El Nino will lead to severe drought everywhere in Eastern Australia. The public are likely to be surprised if an El Nino does not lead to a severe drought, or if a drought occurs without an El Nino. Yet both of these have happened in the past. (Nicholls 1999).

One wonders how the public perceives 'fire warning' and 'extreme fire danger' messages. We probably need more insight into the understanding and acceptance of fire-related messages. Some of this insight is given on the Bureau of Meteorology web site, look under "bushfire weather" and "definitions".

Literature cited:
  1. Nicholls, N. (1999). Cognitive illusions, heuristics and climate prediction. Bulletin of the American Meteorological Society 80, 1385-97. Sink, S.A. (1995). Determining the public's understanding of precipitation forecasts: results of a survey. National Weather Digest 19(3), 9-15.

Malcolm Gill
Division of Plant Industry, CSIRO
6 April 2000