Accountability in Public Meteorology: A Probabilistic Approach

Between the two main forecast methods in Numerical Weather Prediction, probabilistic versus deterministic, there exists certain quandaries that these methods present in the sphere of Public Meteorology. Although derived from highly rigorous mathematical concepts, many scientists and forecasters fail to realize that public meteorology is often dictated by preference and cost-valuation. This issue is further compounded by the general public’s often lacking understanding of meteorological concepts. Typically referred to as "public goods", weather forecasts lack observable price information as well as provisions from private markets. Traditional approaches involve economic modelling of consumer-information markets with marginal "costs and quantities" through supply-and-demand curves. A common accepted thought in meteorology is that as extreme weather events near, the public is more willing to follow warnings from meteorologists or that there will be a "marginal benefit" as information becomes more accurate and prevalent. However, a person's willingness to evacuate their residence is most often tied to the perceived costs of their property value. This is a type of "sunk-cost fallacy" which escalates commitment to protecting one's property and belongings. In fact, due to this, the demand curve may be completely horizontal, unchanging through quantities of information and set at the price of one's property. As a result, market equilibriums may only exist at higher points along supply curves effectively increasing price for "suppliers" who in this case refer to government agencies and meteorologists. Using the University of Maryland's TerpWRF model, we seek to chart how a probabilistic approach may "lower" such market equilibriums, providing less costly information for both meteorologists and members of the public. Through economic arguments, we seek to address how the public’s preference in forecast method affects their decision-making process. Furthermore, we will address the social consequences that probabilistic and deterministic forecasts present and establish the burden of responsibility that these forecasts present.