Farmers' accuracy interpreting seasonal climate forecast probability

Abstract

We explored errors interpreting the seasonal climate forecast (SCF) statement "In the next 3 months, the probability of exceeding median rainfall is 30%" used by the Australian media. Study one comprised 73 farmers and 12 climate staff and (1) developed a five-item multiple-choice questionnaire assessing understanding for the SCF, and (2) assessed errors interpreting the statement. The farmers who incorrectly interpreted the statement (29%) as forecasting a wetter season than normal had a lower understanding of the statement, in contrast to farmers making the correct interpretation (56%) and the climate staff. The statement appeared to be misinterpreted as a specific value of rain that was above the median by 30%, because of (1) difficulty conceptualizing 'above median rainfall' and 'below median rainfall' as distinct events, and (2) the percentage being interpreted as a specific value and not as a notion of chance. Study two, with 72 laypersons, showed more errors and slower responses with statements of probabilities less than 50% (i.e. 30%) than statements of probabilities greater than 50% (i.e. 70%), explained by the positive connotation biases of the event 'above median rainfall' and/or the use of the term 'probability'.