It is being increasingly recognized that the uncertainty in weather forecasts should be quantified and furnished to users along with the single value forecasts usually provided. Probabilistic forecasts of “events” have been made in special cases; for instance, probabilistic forecasts of the event defined as 0.01 inch or more of precipitation at a point over a specified time period (PoP) have been disseminated to the public by the Weather Bureau/National Weather Service since 1966. Many of the single value (or categorical) guidance forecasts produced by the Meteorological Development Laboratory (MDL) have probabilistic forecasts underlying them, but distribution to users is limited. Within the past decade, ensembles of operational numerical weather prediction models have been produced and used to some degree to provide probabilistic estimates of events easily dealt with, such as the occurrence of specific amounts of precipitation. In most such applications, the number of ensembles restricts this “enumeration” method, and the ensembles are characteristically underdispersive.

However, fewer attempts have been made to provide a PDF (Probability Distribution Function) or CDF (Cumulative Density Function) for a continuous variable. MDL has used the error estimation capabilities of the linear regression framework and kernel density fitting applied to individual and aggregate ensemble members of the Global Ensemble Forecast System of the National Centers for Environmental Prediction to develop PDFs and CDFs. Our goal is to provide probabilistic guidance for all surface weather variables used in routine public and aviation forecasts in gridded form in the National Digital Guidance Database (NDGD).

In this paper, we will provide the results of applying our techniques to both developmental and independent data, together covering three years of data at over 1600 stations.