Contact me!

apratt@
bestbits.org

 

APRS in the American West

In May of 2006, I did a study of APRS data to find out where the coverage is (and where the holes are) in the American West and California. This page has the results of that study.

Results

Click on an image for a larger version


The American West

California and Nevada

Northern California close-up

Southern California close-up

Analysis

As you can see from the maps, the coverage is concentrated in population centers and along Interstate and US highways. In California, the large gaps and areas of spotty coverage can be described as follows, running clockwise from "noon" (the top center, where I-5 enters Oregon):

  • The entire northeast: Modoc, Lassen, Shasta, Plumas, and Sierra County, except for I-5, state highway 36, and parts of state highway 299 and 89.
  • The Sierra Nevada, running south from Lake Tahoe to Bakersfield and all the way to the Nevada border (including Death Valley), except for a corridor along US395.
  • The Mojave Desert, east of a line from Barstow to Yuma, AZ, with exceptions for Palm Springs and the I-15, I-10, and I-8 corridors.
  • The entire Central Coast, west of I-5 to the Pacific, from Ventura north to Monterey, except for the US-101 corridor. The un-covered area includes the Pacific Coast Highway from Hearst Castle to Carmel.
  • The entire northwest: Sonoma, Mendocino, Trinity, Humboldt, Del Norte, and Siskiyou Counties, except for areas near US Highway 101, Interstate 5, and Fort Bragg.

Some of these areas have spotty or intermittent coverage, not a complete void. And in some areas without roads, it's likely that no APRS user has even attempted to get a packet out, so those areas would not appear as "covered" on maps like this. But anywhere you see yellow showing through the maps above, you probably shouldn't expect APRS to reach the Internet.

The study

For this study, I used the APRS archives at aprsworld.net. The data I got was from January 1, 2006 through March 31, 2006.

The data at the APRS archive consists of only one report per hour from each station heard, so this is a statistical sampling of all APRS data. It is not 100% complete. If only a small number of mobile APRS users moved through a given area, it's possible it will falsely appear not to have coverage because the archives didn't keep any of those position reports from those stations.

The data I used can not support a study of paths, hop counts, or recommended routes. Also, I could not tell if any reports were from stations that were feeding APRS-IS directly (not over the air).

Using text-based scripting tools (sed, sort, uniq) I filtered the data to include only the American West (30 to 49.9 degrees north, 100 to 124.9 degrees west), and reduced resolution to one tenth of one degree of latitude and longitude. Those are the points you see on the maps: each point represents an area a tenth of a degree on a side.

The resulting comma-separated lat/long data file is 19kb compressed. It's manageably small, with 8475 data points. To get the map images you see here, I used StreetAtlas (version 6) with different icon sizes depending on the map scale.

This study is about APRS packets reaching the Internet via APRS-IS. The study says nothing about APRS for local over-the-air tracking and position reporting, for search-and-rescue or special events where no Internet access is required or desired.


Go to the BestBits home page.

 


This page was last edited April 26, 2008.