Beware of using OIG's STS 107 2-line elements for predictions more than a day or two. The decay terms are too low by about a factor of four. I verified this by comparing predictions of two OIG elsets that are two days apart: 1 27647U 03003A 03020.58333333 .00054970 66098-5 11482-3 0 188 2 27647 39.0142 202.9932 0011031 25.0418 332.9930 15.98413253 628 1 27647U 03003A 03022.58333333 .00062060 69655-5 11960-3 0 283 2 27647 39.0164 189.6316 0010349 44.8390 322.4470 15.99247182 943 At the epoch of the later elset, the earlier one was 31 s late. Adjusting the earlier elset's rate of decay to eliminate the error yields: 1 27647U 03003A 03020.58333334 .00230000 00000-0 46437-3 0 183 2 27647 39.0142 202.9932 0011031 25.0418 332.9930 15.98413253 628 On predictions over a period of 10 days, the unadjusted elset's predictions would be about 20 minutes late. For predictions more than a day or two in advance, I recommend using the projected elements on this NASA Human Spaceflight web page: http://spaceflight.nasa.gov/realdata/sightings/SSapplications/Post/JavaSSOP/orbit/SHUTTLE/SVPOST.html There are also problems with the decay terms on that page; however, the mean motion of individual elsets appear to be correct; therefore, use the projected elset nearest in epoch to the time of the prediction. How is that only the decay terms are incorrect? In the case of OIG, most shuttle elements appear to originate in the shuttle program - probably produced from TDRS Doppler tracking, which should be accurate. The problem with the decay rates almost certainly is caused by inaccurate conversion to the 2-line format. US Space Command's elements are generated as 2-line elements, so they do not suffer from this problem. The projected elements on NASA's Human Spaceflight web page are produced in Mission Control, using very accurate orbit models, taking into account planned manoeuvres. However, as with the Doppler-generated elsets sent to OIG, the projected elements must be converted to 2-line format, and the decay terms are not handled properly. Simple formulae exist enabling approximate conversion of elements from one model to another; however, for high accuracy, it is necessary to perform a differential correction of the approximate elements to fit an ephemeris generated by the source orbital model. Ted Molczan ----------------------------------------------------------------- Unsubscribe from SeeSat-L by sending a message with 'unsubscribe' in the SUBJECT to SeeSat-L-request@lists.satellite.eu.org http://www.satellite.eu.org/seesat/seesatindex.html
This archive was generated by hypermail 2b29 : Wed Jan 22 2003 - 20:35:05 EST