One factor which must be remembered is the apparent effect of the geomagnetic storm on January 10. Very many satellites show an increase in drag over the following few days, well above any expected from the normal progress of a decay. The perturbation is probably beyond anything which might be accommodated by any smoothly varying curve. It is also apparent in the calculated ndot2s for USA 129 r, as in the following list: In message <199701191626.RAA09306@angel.algonet.se>, Bjoern Gimle <b_gimle@algonet.se> writes >... > >Elsets posted ndot2 ndotdot6 MM e : > >96365.52702280 .01097930 00000-0 15.12918679 0533247 >96365.52701960 .01098464 00000-0 15.12923644 0537085 >96366.58430690 .01220474 00000-0 15.15596164 0542171 >97 01.50791800 .01244224 00000-0 15.17919842 0531505 >97 02.56192110 .01020686 00000-0 15.20047638 0528127 >97 04.53436490 .00959924 00000-0 15.23839338 0506368 >97 06.56741040 .00954858 00000-0 15.27717527 0498408 >97 07.54926320 .00964146 00000-0 15.29637502 0492390 >97 08.52975180 .01097522 00000-0 15.31910590 0476076 >97 09.50886640 .01078668 00000-0 15.33990437 0479853 >97 11.78648514 .01625000 00000-0 15.40570033 0435000 >97 11.85243630 .01446399 00000-0 15.40455303 0442942 >97 13.47262092 .01510000 00000-0 15.45505325 0435000 >97 13.53835740 .01510046 00000-0 15.45676063 0419000 >97 15.79859908 .01435000 00000-0 15.52064688 0417000 >97 15.86402340 .01433330 00000-0 15.52210943 0394003 >97 16.82909254 .01518000 00000-0 15.55387590 0361000 >97 17.15133102 .02490000 00000-0 15.56077282 0367600 >97 17.15133102 .01330000 00000-0 15.56077282 0367600 Notice how ndot2 shows a sharp increase in the first elset after the storm (97010.25-97010.50). It then declines, though the ndot2 of 0.0249 on the 17th appears anomalous. I have taken a closer look at recent OIG-published elsets for #24702 = Bion 11 SL-4 r, comparing the published ndot2s with ndot2s predicted by SatEvo for the corresponding MM (mean motion). The SatEvo evolution derives from the published elset for 97009.036, prior to the storm. The following table lists just the elset epoch (rounded) and the ratio of the published to the predicted ndot2. When the ratio is more than 1.0, we may take the atmospheric density to be greater than expected by that amount. Where there were duplicate elsets for the same epoch, they are listed in elset-number, so the latter is (presumably) the more accurate. Epoch ndot2 ratio 97009.036 1.00 (fixed) 9.411 0.99 9.473 0.98 10.160 1.07 10.160 1.05 10.285 1.03 (at the beginning of the geomagnetic storm) 10.597 1.21 10.659 1.86 11.159 1.62 11.595 1.55 12.156 1.44 12.593 1.36 13.216 1.27 13.465 1.17 13.776 1.16 13.901 1.15 14.212 1.11 14.586 0.96 14.897 1.09 15.146 1.01 15.146 1.02 15.892 0.99 16.265 1.03 16.452 1.00 16.762 1.04 17.196 1.01 17.819 0.95 18.502 0.93 18.874 0.94 19.246 0.99 19.495 0.94 The effect of the storm is apparent until the 13th. During this period, the perigee of #24702 fell from 211 km to 199 km, while the apogee fell from 332 km to 283 km. Admittedly, the perigee is rather higher than that of USA 129 r (around 150 km), but it is hard to find lower-perigee objects which have plenty of published elsets covering the period and which are not also much closer to decay. A check of many other objects shows a similar pattern, though with evidence that the peak ndot2 is reached later for higher perigees. At 400 km, for example, a peak of 2.0 times "normal" seems to have occurred on the 14th. One caveat, however, is that there may be a "time constant" in the orbital derivation method which may be longer for more slowly decaying (higher-perigee) objects. Perhaps the atmospheric density increases in phase at all altitudes, but it takes longer for USSPACECOM to register the effect when the decay rate is lower. Alan -- Alan Pickup | COSPAR site 2707: 55d53m48.7s N 3d11m51.2s W 156m asl Edinburgh | Home: alan@wingar.demon.co.uk +44 (0)131 477 9144 Scotland | Royal Observatory: A.Pickup@roe.ac.uk +44 (0)131 668 8224