As new weather data becomes available-such as a fresh metar on the hour or a new Nexrad radar composite image-it is transmitted from the ground and automatically replaces the previous data in the airborne system memory. You don't have to be in continuous line-of-sight range of a ground station to have weather information in the cockpit, but you do need to be in range of a station for at least a few minutes to update the radar image and other weather data. Bendix/King's goal is to have virtually complete contiguous U.S. datalink signal coverage at and above 5,000 feet agl in the future, but there are huge gaps in coverage now, and for some time to come.
Garmin gets its datalink weather from the OrbCom constellation of low earth orbit (LEO) satellites. The advantage of a LEO is that the signal is powerful compared to those from a geosynchronous satellite that is more than 20,000 miles above earth. The signals from a LEO can be received with an ordinary whip type of antenna, but to use a geosynchronous satellite you need a high gain antenna, a dish, in other words. And the dish needs to be steerable so that it remains precisely pointed at the satellite as the airplane moves and maneuvers. That's a complex and very expensive task, and such an antenna requires more room than is available on a light airplane, even if money were no issue.
The satellites in the OrbCom constellation orbit in several planes, so there is continuous coverage as individual satellites whiz by overhead. The orbital groups are labeled "A," "B" and so on, and individual satellites are assigned a number, "B4," for example. At least one satellite in the constellation is in view something like 99 percent of the time, but the geometry can be poor as a satellite rises into view over the horizon or sets below the horizon, and the signal is not always optimum even though it is in view. Garmin shows you which satellite is being tracked and when data is being sent or received. I often watch this information to see if the satellites in orbital plane "C" are more reliable than those in plane "F," for example, but I haven't been able to reach any conclusion.
The downside of a LEO is that the data rate is low, making it impractical to broadcast all available weather data all of the time. The solution is to use a request-reply scheme where the pilot requests specific weather information, such as a Nexrad radar view over a specified part of the country. That request is then transmitted up to the LEO. The satellite then relays the request to a ground station, where the weather data package requested is assembled. The information is then retransmitted up to the LEO, which in turn sends it back down to the GDL 49 receiver in the airplane. It typically takes about two and a half minutes from the time you make a request until the information is available for viewing, but it can take much longer if the "connection" between the satellite in view and the receiver is not optimum. I have very occasionally seen a request take as long as 20 minutes to be returned, but it is uncommon for the process to take more than 10 minutes, and less than five minutes is typical.
The fundamental advantages and disadvantages of each technology are obvious: Weather data is available at any altitude with the Garmin satellite system versus large areas of no coverage from the ground-based Bendix/King system when flying at lower altitudes. On the other hand, the much larger bandwidth of the ground-based system allows Bendix/King to broadcast much more information, while the Garmin data rate can't be increased by flying at a higher altitude. Dick Collins explores these differences in his story.