Economic Value of Weather and Climate Forecasts

Rating:
Summary: THE Reference
Review: The book represents the best reference for "distributions oriented" verification of probabilistic forecasts. Developed and described in the context of meteorological forecasts, the DO paradigm is robust, and has ubiquitous applications, when one recognizes that EVERY forecast is inherently based on a conditional probability distribution.

The focus on meteorological forecasts does not make applications in other disciplines transparent, in a tutorial way, but anyone who thinks rigorously about the "quality" of forecasts will make the connection. An essential, and universal concept demonstrated from meteorology is the realization of "quality-value reversals" in which forecasts with higher "quality" (quantified by traditional error measures such as correlation or RMS error) have lower "value" in terms of net benefit of the forecast when used in a specific decision making context. This fundamental concept is well extablished in other fields with high-stakes forecast-based decisions (, e.g. medical decision making) and should be fundamental to any development/evaluation of forecasts (meteorologiical or otherwise).