Hi Dave, Yes, I am using the ultra rapid data, and have taken the other issues you mention into account (not necessarily correctly...but I try). I am not familiar with NOVAS, but just looked it up on the usno pages. I will look at it more carefully, since it is the XYZ coordinates that I worry about. Thanks for the suggestion.
Regards, Dale ________________________________ From: David MacMahon [dav...@astro.berkeley.edu] Sent: Tuesday, June 19, 2012 1:31 PM To: Gary, Dale Cc: Andrew Martens; Jesús García LLedó; casper list Subject: Re: [casper] Some questions Hi, Dale, On Jun 19, 2012, at 7:14 AM, Gary, Dale wrote: > Does anyone know an accurate way to find the location of GPS satellites? I > wrote software to calculate this, nominally to an accuracy of a few meters > for the satellites, but I never succeed in getting steady phases after > correction for the calculated position, indicating that my calculations are > not correct to better than about 30 arcminutes. My calculations are based on > the so-called "SP3" files kept at ftp://igscb.jpl.nasa.gov/. If anyone knows > of suitable software for getting local pointing coordinates of GPS satellites > to arcsec accuracy, please let me know. Are you using the ultra-rapid SP3 files? They claim pretty impressive accuracy for the predicted half of these files (~5 cm): http://www.igs.org/components/prods.html I have used NOVAS to get local XYZ positions. From there it seems deceptively easy to get local az/el, but actually there are a few considerations to watch out for (Earth orientation parameters like ut1-utc, ellipsoid model, atmospheric refraction, ionospheric delay, etc.). I had a routine that compared the SP3 GPS positions with those generated from TLEs, but I don't think I ever went to local az/el. Hope this helps, Dave