NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Errors in USNO celestial data
From: Paul Hirose
Date: 2010 May 02, 22:47 -0700
From: Paul Hirose
Date: 2010 May 02, 22:47 -0700
If a program expects UT1, you can input UTC, a much more accessible time scale, and be off by no more than .8 seconds of time or 12 seconds of arc in Earth's rotational position. (That discrepancy, UT1-UTC, is called "delta UT". Values are listed in the IERS bulletins.) Incorrect delta T is another issue. Say the correct value is 66 seconds but the program uses 70. In that case, to get the coordinates of the Moon at 12:00 UT1 for instance, input 11:59:56 UT1. The program will add an extra 4 seconds, thereby arriving at the correct TT. However, your position in space relative to the celestial coordinate reference system will be too far west, since you gave the computer a too-early UT1. This doesn't matter for right ascension and declination, but affects topocentric coordinates such as azimuth and altitude. The solution is to multiply the -4 s UT1 correction by 1.002738 to convert to sidereal seconds, multiply by 15 to convert time to arc, and move the observer's longitude *west* by the resulting angle, i.e., a little over a minute east in this case. If utmost accuracy isn't needed, you can simply enter the correct UT1, which puts the observer in the correct place relative to the celestial reference frame. The body's place in that frame will be slightly off, the error amounting to .5 arc second per second of time for the Moon, much less for other bodies. In a few days I'll release a free almanac program for Windows. It lets you enter any value for delta UT or delta T (depending on which time scale is selected). For convenience, there's an internal representation of the delta T table in The Astronomical Almanac. "Auto delta T" mode linearly interpolates the table for any date from 1620 to 2010. The program is practically complete. The remaining problems are nitpicky little things, like inconsistent upper and lower case spelling on control labels. Accuracy is not a problem. It duplicates the precise stellar and planetary reduction examples in The Astronomical Almanac. This accuracy is combined with a comprehensive set of heliocentric, geocentric, and topocentric coordinate reference systems and format options. An ICAO standard atmosphere model provides pressure and temperature for refraction, or you can enter the values by hand. Time may be specified in the Gregorian or Julian calendar, in the UTC, UT1, TT, Greenwich apparent time, local apparent time, and local mean time scales. Output is entirely numeric, and for only one point in time. There's no provision to generate tables. My main task is to finish the Web page describing the program. That will be the only documentation (what do you expect for free?), so I have to make an effort to cover all the features and pitfalls. This is actually a small and simple program, so a full description needs only one page. I should be done in a couple days. --