Radio Line of Sight Plot Server

What is this?

The form on the following page will allow you to make a request to a radio line of sight plot server. By return mail, you will receive a GIF picture of the elevation of each point on the earth between the starting point and ending point that you specify. This elevation plot is pre-warped to account for the fact that microwave radio signals don't really propagate in a straight line. They curve toward the earth so that the apparent radius of the earth is 4/3 the actual radius.

If a line drawn from the starting point to the end point intersects the earth profile on the chart, then you probably don't have a line of sight path between the two points. If the line doesn't intersect the profile, you may have a line of sight path. (The plots don't take into account trees, buildings, local obstructions, or geographic features that are too small to be resolved on the digital elevation maps.)

What are the limitations here?

The first and most concrete limitation is that these plots are for amateur use only. No commercial, industrial, governmental, or other use is permitted. The author of these tools, the service providers involved, the manufacturers and producers of the hardware, software, and databases may not be held liable for any damages or disappointments resulting from the use of these plots. We all disclaim all responsibility for any damages or disappointments resulting from use of these plots. If you use it for commercial purposes, you'll be tracked down and haunted for the rest of your life by the ghost of the thousands of freeware producers that have shed this mortal coil.

Other than that, things seem to work reasonably well. The limitations include

• The Digital Elevation Maps provide a grid of most of the continental US at 3 arc-second intervals. In New England, this translates to an interval of one data point every 90 or 100 meters. The altitude resolution is about 3 meters.
• This service is not new. But, I'm still working out bugs.
• There is a limit to the length of a path that the server will calculate. The current limit is about 4000 Km. I can't vouch for the accuracy of plots at that length, but you and I already know that there are no line of sight paths that long for anybody but pilots and mountain climbers.
• I'm doing this as a part time service. I'll try to see to it that the server sends back responses once a day, but vacations, power outages, equipment changes and such may make service a little more erratic.
• The responses are mailed to the requester. You must have a valid mail address and a mailer capable of handling GIF files that are mailed to you. If you don't have that stuff, I can't help you. I can't help you with mail problems, mail handler problems, ISP problems, or much of anything.

What do I need before I fill out the form?

• You need to know the grid square or latitude and longitude of the start and end of your path.
• You need to know your return mail address.
• You need to know how to fill out a web form.
• You need to read the sections above about appropriate use limitations of this service.
• If you know how high the antenna at each end will be above local ground that will be helpful, otherwise I'll assume that they are 2m above ground level.

If you specify a location as a grid square, the server will find the highest point in the grid square before calculating your plot. This makes a big difference for grids that include mountain tops. I'm assuming that if you're in a grid with a mountain, you'll drag your stuff up the mountain. If you need greater control, supply a lat/lon pair for one or both end points.

Where do the plots come from?

The plots are produced on my AMD Athlon PC running Linux. Each night, Murphy permitting, the PC collects requests made in the previous period and scans a very large database of digital elevation maps provided by the USGS. Each profile takes about one second to produce and is then mailed (in GIF format) to the mail address that you provide.

Why isn't this in real time?

The data set required to produce the plots is huge. The maps covering the continental U.S. occupy a little less than 600MB of disk space. My internet provider charges for space, so storing the maps on the web server isn't practical. More importantly, the time to calculate a plot is dominated by the generation of a great-circle route from the starting point to the end point. This calculation can dominate the cost of producing a plot on processors that don't provide good floating point performance. ISPs aren't set up to provide this kind of compute horsepower.