Next-Generation Air Force Weather Metrics via Bayes Cost Analysis

This research proposes a new methodology for U.S. Air Force Operational Verification (OPVER) weather forecast metrics. Military weather forecasters are essentially statistical classifiers. They categorize future conditions into an operationally relevant category based on current data, much like an Artificial Neural Net or Logistic Regression model. There is extensive literature on statistically-based metrics for these types of classifiers. Additionally, in the US Air Force, forecast errors (errors in classification) have quantifiable operational costs and benefits associated with incorrect or correct classification decisions. There is a methodology in the literature, Bayes Cost, that provides a structure for creating statistically rigorous metrics for classification decisions that have such costs and benefits. Applying these types of metrics to OPVER yields more informative metrics that account for random chance while remaining simple to calculate.

Using notional costs and benefits from Air Force operations subject matter experts, a case study was conducted by performing Bayes Cost-based OPVER on Terminal Aerodrome Forecasts and Watches/Warnings/Advisories compared to surface observations from a selection of military installations in the continental United States during the period 09 April 2019 to 30 July 2019. The case study illustrates the added utility of the new metric paradigm.