Probabilistic Precipitation Portal: Advancing ProbSnow and ProbQPF into a Centralized Tool for IDSS

Predicting rainfall and snowfall amounts remains a significant challenge for National Weather Service (NWS) forecasters and represents a considerable hazard to life and property across the country. By incorporating probabilistic information into precipitation forecasts, NWS forecasters are able to better communicate the uncertainty in these predictions, enabling communities and core partners to make more informed decisions. The current Probabilistic Winter Precipitation Forecast (PWPF) experiment, which includes “ProbSnow”, and Probabilistic Quantitative Precipitation Forecast (PQPF) experiment, also known as “ProbQPF”, have provided many NWS Weather Forecast Offices (WFOs) with tables and maps displaying low-end, expected, and high-end amounts for upcoming snowfall and QPF events, as well as probabilities of exceeding various threshold amounts.

These experimental products have proven to be successful tools to support probabilistic impact-based decision support services (ProbIDSS) over the past several years. However, not all WFOs have been part of these experiments, and a national-level solution was chosen as the way forward. In order to make this probabilistic information more consistent and widely available, the Weather Prediction Center (WPC) has created the Probabilistic Precipitation Portal (PPP). This new web page provides probabilistic precipitation maps, tables, and graphics for NWS offices across the contiguous United States.

The web page features a GIS-based, interactive map to quickly and easily explore probabilities and amounts of precipitation over the next 3 days. WPC also continues to provide the tables and images that were the centerpiece of the previous experiments as part of the PPP.

While these products are created at WPC, WFO forecasts remain the starting point in the forecast process. Precipitation amounts provided by WFOs and sent to the National Digital Forecast Database (NDFD) are the mode of a binormal distribution used to compute probabilities. A 60-member multi-model ensemble sets the distribution’s spread and shape. The probabilities update frequently using the latest NDFD data and projected on the standard 2.5km NDFD grid over the contiguous US.

A probabilistic approach to precipitation forecasting empowers decision-makers to account for forecast uncertainty and helps build a more Weather-Ready Nation. We will highlight key pieces of our development process and show how forecasters can leverage the PPP to enhance their messaging and decision support ahead of impactful precipitation events.