New Weather Model Will Improve Storm Forecasts

10/15/2014 (Permalink)

Twelve hours before the Washington, D.C., area experienced severe winds from a derecho, the High-Resolution Rapid Refresh (HRRR) model correctly predi

By Doyle Rice, USA Today

The National Weather Service put a new weather model into operation, a hyper-local program that promises to improve forecasts and warnings for severe weather. Weather models are complex computer programs that help forecasters analyze and predict the weather.

The weather service says the new one will be able to pinpoint neighborhoods under threat of tornadoes and hail, heavy precipitation that could lead to flash flooding or heavy snowfall -- and warn residents hours before a storm hits.

The High-Resoluton Rapid Refresh weather model is the first to include hourly updates as well as a visual resolution sharp enough to simulate individual thunderstorms, with grid points about 2 miles apart, the National Center for Atmospheric Research says.

The closer together the grid points, the better the chance of accurately predicting where it will rain. The HRRR has four times better resolution than previous weather models.

"HRRR will be updated hourly ... which is what's needed for predicting fast-changing storms and something not done before by the NWS," said Stan Benjamin, a meteorologist with the National Oceanic and Atmospheric Administration's Earth System Research Laboratory.

The model has been used experimentally for four years, including during the 2012 Washington, D.C., derecho -- a severe straight-line windstorm -- which it accurately predicted.

Is the new model all it's cracked up to be? Two scientists not affiliated with the weather service say yes:

"HRRR is a major advance for the National Weather Service and an important one," said Cliff Mass, a meteorologist at the University of Washington. "Large improvements are possible in short-term forecasting ... which we call nowcasting ... and HRRR is a substantial technological advance."

"HRRR is a unique and absolutely great tool for the energy industry, sports venues, transportation companies, etc., with countless applications for people and property affected by weather in the very near term," said meteorologist Ryan Maue of the private forecasting firm WeatherBell.

A caveat from Maue: "This is only one model representation using one set of physics and assumptions. There are countless configurations of the weather model that could generate drastically different forecasts ... As with any weather model, there will be monumental successes and busts."

"The human forecast is not going to be replaced, nor can we rely solely on computer generated forecasts for severe weather warnings or automated warnings," he added.