Why communicating weather forecast uncertainty is crucial for decision making



By Epa Ndahimana, AFMC/NECJOGHA

Kampala, Uganda – Ugandan farmers hold the view that gradually unpredictable weather is leading to poor yields, a reduction in crop varieties and pastures, poor animal health, rangeland related conflicts, greater expense and labour, food insecurity and reduced incomes leading to poverty.

According to the Africa Climate Change Resilience Alliance report of 2017,  many farmers in Uganda rely on indigenous weather forecasting methods and thus, find it difficult to carryout farming activities due to obstacles to correctly predict weather and seasonal changes due to climate change. This ultimately affects Uganda’s capacity to achieve the broader development objectives.

Communication is conveying knowledge so that it is reasonably understood and, typically, used to guide action. However, under climate science, this demands clear and accurate communication of uncertainty nature of weather forecasts.

Uncertainty is one component of climate science that disrupts the central reason for climate information communication. Thus, uncertainty communication is paramount for effective decision making by end users of climate services. The decision made with uncertainty in consideration avoids collision between forecasters and end users, promotes confidence and eludes unrealistic conclusion over weather forecasts by the public.

Challenges of forecasting unpredictability

However, the challenge lies within the packaging and interpretation of the forecast unpredictability. Though the language of uncertainty and probabilities is considered common sense for meteorologists, for the general public it appears as rocket science and therefore risks being misunderstood. The usefulness of communicating uncertainty means that expectations are managed, user confidence is achieved and reflects the state of the science.

One absolute risk in communicating weather forecast is mistrust and lack of faith in the science of prediction by end users. This is usually the case where absolute trust is vested in the forecasts without expectation of any change.  For instance, where a farmer repeatedly follows the advisories from the National Hydrological and Meteorological Services (NHMSs) and crop losses due to miscommunicated uncertainty, the faith in forecasts is finally eluded. To restore this confidence, they must be made to understand that weather predictions are not absolute and are never 100% correct.

It is of absolute importance to comprehend the sources of uncertainty in weather forecasts. The chaotic nature of our atmosphere perhaps makes it impossible to perfectly predict its behavior in near future.

James Bataze, a meteorologist with the Uganda National Meteorological Authority (UNMA) says that the shorter the time the easier to predict weather conditions. For instance he explains its easier to predict what will happen next week or month that what will happen in six months or year.

The expression of probabilities is one of the ways recommended to convey uncertainty to end users. The interpretation of probability is usually different and that’s why detailed explanation is paramount in communicating probabilities. For example, a 30% probability of rain in Kampala means that there is a 3 in 10 chance of rain in some place in Kampala. However, without clear knowledge of probabilities, one might wrongly think it means 40% of the daytime it will be raining in Kampala.

It is paramount to understand that communicating uncertainty is possible but with difficulties. At times users of climate services want definitive predictions like ‘it will rain or not’. The language of chances and probabilities of a particular occurrence comes with difficulties and questioning that forecasters and communicators cannot easily answer.

Communicating probabilities and uncertainties casts doubt on the confidence of predictors and often is taken as an attempt to dodge responsibility. Thus, in a country like Uganda where communicating forecasts is at infancy, effective user education is a prerequisite for users to understand why weather forecasting is not precise science.

Hence, for enhanced and effective utilization of forecasts, clear, precise and exhaustive delivery of uncertain nature of forecast forms the foundation for timely and accurate climate information that can enable development.  



Please enter your comment!
Please enter your name here