Recognizing that probabilistic weather forecasts are critical to meeting the air traffic efficiency goals of NextGen, RAL is working under FAA sponsorship to develop a probabilistic 1–10 hr C&V forecast product (CVF). This probabilistic approach acknowledges the inherent uncertainty in the forecast process and seeks to quantify and convey that uncertainty directly to the user. Forecast probabilities can be intelligently interpreted by human decision–makers, and can be used quantitatively in decision support systems where a variety of inputs and consequences are considered. In mature form, CVF will forecast the probabilities for ceiling and visibility conditions ranging from fully obscured to clear
RAL's CVF development combines existing operational forecast resources through use of new techniques, and typically yields forecast skill that exceeds that of its input resources. For example, blending of LAMP forecast data with time–lagged ensembles derived from the Rapid Refresh Model (RAP) yields the improved IFR ceiling forecast skill shown for the Philadelphia International airport shown at the top of the figure. Similar results for forecasts of IFR visibility are shown at the bottom. Under NOAA funding, development is seeking to determine how CVF techniques might best enhance the tools available to operational forecasters today.
CVF development will include evaluation of the forecast skill increases attainable through use of time–lagged ensembles from the High Resolution Rapid Refresh and NAM models, as well as use of RAL–developed statistical forecasting methods. Conventional ensembles comprised of members from the NAM and Rapid Refresh will also be evaluated. The first–generation experimental CVF product was evaluated by the National Weather Service in November, 2012.