Irreversible Adaptation and Knightian Climate Uncertainty

We model a setting where a firm must choose when to adapt to climate-induced resource scarcity based on objective expectations of climate change (climate projections), and is influenced by both risk (annual variability in climate projections) and ambiguity (inability to assign probabilities to various climate projections). We show that the effect of risk on the timing of adaptation depends on the presence of diminishing returns in the firm's production technology, while the effect of ambiguity depends on whether there is ambiguity in resource supply trends, instantaneous variance, or both. We apply our model to the decision-making process for adopting water-saving irrigation technology in California's Central Valley. By analyzing sixty-four downscaled climate projections for water availability in the region, we quantify the climate variability and ambiguity faced by agricultural firms and determine the optimal timing for adaptation.