Navy's new NAVGEM sharpens its global weather forecasting
- By Patrick Marshall
- Jun 27, 2013
Some historians argue that storms did more than the British fleet to sink the Spanish Armada in 1588.
Fully aware that accurate weather forecasts can be critical in military operations, the U.S. Navy has just adopted a new global weather forecasting model developed by the Naval Research Laboratory. The Naval Global Environmental Model (NAVGEM) went operational in March after three years of development.
"It is a more complete physics package," said Simon Chang, superintendent of NRL's Marine Meteorology Division. "Over the course of time we learn how to do things better. We understand the atmospheric processes better. We do the algorithms better. And we understand the sensor data better."
The algorithms developed by the team of 15 meteorologists and programmers at the NRL not only reflect a better understanding of weather, they are more efficient than those employed by the previous model. "It will allow us to run NAVGEM at a higher resolution without requiring a large increase of computing resources," said Chang.
The bottom line is that NAVGEM can resolve weather events at a resolution of 35 kilometers, while the previous model could only go down to 50 kilometers. The end result is better and more detailed forecasts. "It is capable of generating more accurate models of weather processes," said Chang. "You can resolve the intensity of cyclones better."
According to Chang, the NAVGEM system is 99.9 percent automated. Real-time data feeds are received from the Global Telecommunications System, an international data collection and communication system operated by the World Meteorological Organization. The data is then run through an algorithm to make it acceptable to NAVGEM. "Then we run the model four times a day," said Chang. "So our supercomputers are going all the time."
The public can access much of the current data and predictions of NAVGEM on a page that updates twice a day.
Although the NAVGEM model is a major improvement, according to Chang and others it is not the best. That honor goes to a system that was developed by the European Centre for Medium-Range Weather Forecasts. "They have a much bigger computer," explained Chang.
Patrick Marshall is a freelance technology writer for GCN.