Advances in Numerical Weather Prediction

Numerical Weather Prediction

The Great Lakes add heat and moisture to the air surrounding them. Adding heat and moisture results in a lowering of barometric pressure, which our computer models can now show. Upper left image shows the additional snowfall (shaded areas) in inches and lowered surface pressure (red lines in millibars) that occurs in the National Weather Service's computer model as a result of the Great Lakes for a 36 hour period ending 10 AM EST December 20, 1996. Image courtesy of Dr. Greg Mann, National Weather Service Detroit/Pontiac. Lower left image shows simulated radar reflectivity of lake effect snow bands from winter of 2014. Image courtesy of NWS Marquette.

Besides weather radars, satellites, and actual measurements of the atmosphere, meteorologists use computer models to forecast the weather. An atmospheric computer model is a complex program of mathematical equations written by scientists to simulate air motions taking place over the earth. Based on these air motions, we can determine temperature and pressure changes and what kind of weather may take place. The National Weather Service's ability to develop and run more complex weather computer models has allowed for more accurate forecasts of wind, pressure and temperature changes on a smaller scale.

These computer models run by performing millions of calculations at specific locations called grid points. The number of grid points a model can have depends on the processing power of the computer running it. The more grid points that are in a model over a given location, the more accurate it can be in forecasting wind, pressure, and temperature. 

If you live near the Great Lakes, you know that the lakes add heat and moisture to the air during the fall and winter, resulting in lake effect rain and snow. The added heat and moisture can also produce deeper low pressure systems and stronger winds. Computer models in 1975 were not advanced enough to include the effects of heating and moistening from the lakes. It was like the lakes weren't even there! Today's computer models give meteorologists better information upon which to base their forecasts. A recent development in weather forecast modeling includes the ability to simulate future radar reflectivity based on current surface observations, near time radar output, and higher resolution modeled forecast parameters such as temperatures, humidity and winds.  

A significant development in modeling on the Great Lakes occurred in 2013-2014 when the Wavewatch3 model was introduced in forecasting wave heights. Wavewatch3 uses bathymetry and sea surface temperatures of the Great Lakes along with winds and temperatures from the National Digital Forecast Database (NDFD) produced by the NWS to determine stability and resulting wave heights. Horizontal resolution of the model is 2.5 km. The model is run multiple times daily. The image below is of a wave height forecast on Lake Superior using the Wavewatch3 model.

Anyone can look at what the atmospheric computer models are showing meteorologists. Just visit the following weather models web site from the National Centers of Environmental Prediction (NCEP).