In the course of a comment thread with Ben of Snowpetrel, I shared this information, perhaps it will be useful to others:
GPS signals are only about 50 watts at the transmitter. What you receive from a bird at 20200 km moving 3.9 km per second is about -160dbw down from that–(0.0000000000000001watt) per channel best case. The more channels the better on the antenna end. And for ship’s units, the wiring must be impeccable and the amplifier at the antenna end in perfect order. Here are the things that can make it GPS less reliable
- Ionosphere and troposphere delays – Accounted for in the software but solar flares and sun spots are not.(activity for both was low in 1995)
- Signal multipath – GPS signal bounces off hard surfaces (tall buildings, large rocks, icebergs) before reaching the receiver. This increases the travel time of the signal, thereby causing errors. (Like your mile jump — jumping is classic multipath)
- Receiver clock errors – slight but real and can be affected by receiver temperature (we make sure ours don’t sit in the sun)
- Orbital (ephemeris) errors – these can be caused by variations in gravity, solar wind, debris impacts on the satellite, etc.
- Satellites visible – need three for a useful solution — disappearing satellites and appearing satellites can result in some situations where even if you have four or more in view, the receiver will treat them as two until the received wattage is acceptable
- Satellite geometry/shading – GPS satellite navigation solutions are a cocked hat just like a sextant, if the birds are too close the solution will have a significant x, y, or z error, but it doesn’t last long and WAAS can help if it is available.
- Antenna blockage. The best place for GPS antennas is as low as possible — this can almost guarantee at some point they will be blocked from some direction — wet laundry can do it — we proved it.
- Intentional degradation of the satellite signal – Selective Availability (SA) was used until 2000 and then disabled and eventually removed. Specific targeted jamming in areas of hostility has replaced it.
As to a sextant. I carry one, but not for celestial. I’m not expecting GPS to be out of order long enough off-soundings for DR to be insufficient. I also keep the tables to calculate by hand because I know my celestial calculator will go TU when lightning is close by when I have it out of its Faraday box.
I carry the sextant for piloting. Mostly for horizontal & vertical angles for distances off etc. There are still many places where the land masses are not in their proper positions on paper charts. So in this case, I have the sextant to account for the fact the GPS will know exactly where you are on a theoretical ellipsoid, but some honored explorer from centuries past was a bit off on where they marked the rock.
I carry one “ships GPS” integrated with radar/chartplotter. I have three handhelds (not to mention what’s in the phones and laptop). Phones, laptop and one GPS live in a shielded box.
We also always keep the track function enabled and on about a 500 meter FOV as this helps us to notice possible anomalies more quickly.
I keep daily track of GPS status through an email message service from the USCG. it is available at:
One final thought, early general civil use GPS units were not all that hot when it came to antenna design and integration. For the units to serve best, there needs to be an antenna element assignable to each satellite signal. The early units had 1 then 3 then 5 then 9 then 12 elements. By the time they got to 12 things were pretty well sorted out. Back in the 1 to 5 element days a unit as (intrinsically) reliable as what we have now could have cost around $2000. I know because in the Persian Gulf War I had to buy them for our troops because we didn’t have enough military units on hand. The satellite constellation was also not fully populated and navigational accuracies would only be available for about 15 hours in very complexly shaped and ever-changing regions.