Anyone know where a schedule of the times of orbit of the GPS sattelites might be found? Given such data, forecasting a good time for taking readings at a specific location should be possible. Eliminating multiple readings over several days would be the enjoyable result.
Will predict positions of most satellites bright enough to see with the naked eye (or which flare bright enough to see sometimes).
Feeding in the orbital elements of the GPS satellites or knowing their names should get you predictions of visibility for lat, long & time.
One I have never used seems to have their orbital elements in its database:
formatting link
Quite a lot of work to do it manually so you are probably going to need to write a program taking as input their latest orbital elements and then looking for the times when N or more GPS satellites are visible to you.
Are you taking readings over a wide area? The constellation doesn't vary with location until you move significant distances... unless you are dealing with significant variations in horizons. I don't think horizon effects are going to be included in any databases.
You will get your best readings from sats spread around the sky. So including sats near the horizon would be important to minimize the calculation errors.
Only to some extent. While you want a good spread, you actually want to exclude SVs that are very close to the horizon because they're the ones most vulnerable to multipath and propagation disturbances. Many receivers, including all timing-grade receivers, use an elevation mask to avoid that.
For the same reason, the better antennas tend to use choke rings or other ground topologies to discourage low-angle reception.
Martin - thanks for the info, especially the "spacecraft elements" at in-the-sky.org. From what's there, I need to learn the meaning of the terms shown and the procedure for predicting positions at a given time. Any suggestions?
Hul
Mart> > Anyone know where a schedule of the times of orbit of the GPS sattelites
Gerhard I looked at "heather" and I'll keep a record of same. Currently I'm too much of a rube in sattelite language and information to be able to pick needed data from the mass available there.
Hul
Gerhard Hoffmann snipped-for-privacy@arcor.de wrote:
John I am looking at positions generated each second by one of Ublox's
9 series of GPS receivers. Some readings at a single location show (monday at
10, tues at 5..) good repeatability and some are noticebly off. The idea is to see the satellite positions at the time of known readings, gain a feel for good/bad satellite positions and use that to shedule further readings.
Hul
John Miles, KE5FX snipped-for-privacy@gmail.com wrote:
I discard everything other the GPRMC since I'm tracking mobile units but GPGSV gives the information on the satellites in view. Not useful for prediction but it might be of interest.
The bible for that sort of thing is Jean Meeus's book Astronomical Algorithms but it is nearly out of print and copies in the USA sell for insane prices. These guys still have reasonably priced stock.
formatting link
(chapters 29 through 33 - assumes moderately advanced maths)
Bad news is it doesn't deal with terrestrial satellites only solar system objects but the principles are the same it is just that there are drag and oblateness corrections for orbiting the Earth in addition.
And also a much more significant correction for the observer being sat on the surface of the Earth rather than at its centre. Objects in near Earth orbit are subject to a lot of parallax from the ground.
Offhand I don't know of any code that does it very accurately (like the GPS code does) but then you may not need great accuracy provided the satellite positions are indicative of how many you have in the sky at any one time. Never used any of these but worth a look:
I'm using the NMEA sentences from the reciever. That allows shifting from, in my case, Ublox to Skytrack receivers, with minimal effort. The "terminal" I've setup collects satelitte names & info which ... hmmm... that includes the position of the satellites which should be useful, could even generate the ability to predict positions from a heap of past data. Should be something simpler though. I'll take a look qsl.net.
Hul
rbowman snipped-for-privacy@m> > John I am looking at positions generated each second by one of Ublox's
Martin - thanks for the suggestion. I'll give Meeus' book a try. Accuracy of the orbital need not be great for this application, as you mention. I looked at Meeus' book once when seeking a procedure for calculating the sun's position and was supprised at the lack of accurracy. Someone's keeping their methods close to the vest.
Hul
Mart> > Martin - thanks for the info, especially the "spacecraft elements" at in-the-sky.org. From
The full methods for the solar system are published as VSOP82 onwards. Current one is VSOP2013. It is a semianalytic fit to numerical integration and incredibly accurate.
Observations of a pulsar that got really close to Jupiter found a problem with the FORTRAN continuation cards (>10) generated by the symbolic algebra program. For a moment it looked like relativity was not predicting the right time delay for light passing close to Jupiter. But it turned out that Jupiter wasn't where the theory said it was.
Fixed from VSOP87 onwards. That is good enough for most purposes...
Almost all planetarium programs use this VSOP method now.
I don't think it handles satellite ephemera, but for anyone interested in Astronomy this package might prove useful:
formatting link
A few lines of code can give you telescope pointing directions (alt/az) from any point on earth to any object given its celestial coordinates, for example.
I'm sure the author would be appreciative of extension code to handle satellites.
Unless you implement your own receiver, that isn't going to work reliably. The receiver can pick and choose which satellites it includes or excludes in its calculations and there is no guarantee that it will pick the same ones at two different times, even if the same ones happen to be visible. For the same reason, you can't use two ordinary consumer GPS receivers to do differential GPS by just subtracting the known error in the position solution of one station from the other station's position solution - that won't work at all if they are using different satellites.
About 15 years ago I recall finding a software GPS receiver somewhere on the internet, and I'm sure there have been more or better ones developed since then. Perhaps that would suit your needs, as you could tweak its behaviour, get it to print out the satellite positions, etc.
Martin the VSOP methods sound like what I was after, thanks for mentioning it. The Predict program at qsl.net appears useful but the source is barred by an indemnity clause in their "terms of service".
Hul
Mart> > Martin - thanks for the suggestion. I'll give Meeus' book a try. Accuracy of
Martin the readings I've recently taken are formed by waiting about 2 minutes after the reciever has "captured" a useable number of sattelites and then averaging all readings for 8 minutes. This has shown uniquely accurate (repeatible) readings one day and some distinctly at variance on another. The objective now is to identify the good days and also the bad days in order to avoid the latter. Averaging more than the most basic method mentioned above would be counter to current intent, at least at this point.
Hul
Mart> > John I am looking at positions generated each second by one of Ublox's
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.