Indicators and predictions of climate change

Forecasts provide indications of possible future climate changes that are needed to inform policymakers, strategic planners, and the public. This chapter compares the usefulness of climate forecasts obtained from three distinct approaches: univariate statistical methods, models based on physical processes, and the judgments of experts. It evaluates the methods on four criteria: (1) accuracy and lack of bias, (2) uncertainty assessment, (3) acceptability by users, and (4) computational efficiency. While acknowledging the limitations of the methods, it concludes that all three approaches can have a valuable and complementary role in indicating future climate changes, given the diverse needs of forecast users.