Value of forecaster in the loop

Abstract

Over the past 20 years, there have been many advancements in the process of weather. Specific contributions to improvements in the forecast process come from improvements in numerical model forecast guidance and the timely availability of many new observational data types. As numerical forecasts guidance improves, it has become important to document the value a human forecaster adds to the forecast process. Because of collateral duties and career requirements, United States Navy forecasters often find it difficult to become proficient at forecasting. In this study, a basic forecast process was used to identify the skill in forecasts of ceiling and visibility made by human forecasters and produced by a statistical modification to numerical model fields. Terminal Aerodrome Forecasts (TAF) were collected for eight military air stations and compared to the performance of the Localized Aviation Model Output Statistics Program (LAMP). Hit rates and critical success indices were used to identify forecast skill. Using various timelines and categorical partitions, all results showed that there were little to no statistical differences between the TAF and LAMP in 2013. Finally, a case study was examined to highlight the capability of probability forecasting as an improvement to the forecast process.