Try page 376 & 377 of:
2005ESASP.587..373J Page 377
(it's actually only 5-6 pages long but the page numbers are from the overall journal)
I believe it was Insat IIB.
Andy
Thanks. I looked at this yesterday, and couldn't make much sense out of it because the graphs were too small to read. So I went about the task of looking at a couple sats, looking at the apogee/perigee info for both, for what it was back in the 1990s, and currently. I looked at GOES-6, which is oscillating around 105, and I looked at Westar 2, which is in continual drift. For both sats, the average between apogee and perigee are the same now as they were back in the 90s, however for both sats, the orbits have become more eliptical.
For W2 for example, the apogee in 1998 was 35926, and is now 35931, ie out further, however the perigee in 1998 was 35886, and it's now 35881, ie closer to earth. The average of the two is 35906 both then and now. The only very minor differences I can see other than this is the period seems to have very slightly increased from 1442.20 to 1442.22, and that the inclination of both sats have increased over the decade.
I was about to then do some more runs on these two sats, to see how the varying longitude affects the above parameters, however THEN, I looked once more at the URL you posted. THIS time I hit the PRINT button, and found that when you do that, it spits out a PDF file, which you can enlarge, and see the graphs clearly. DUH.
Anyway, looking at the data there, it seems apparent, that my comparing data from only about 12 years apart probably wasn't of much value, since the graphs there basically show a periodicity whereby the difference between apogee and perigee increase, then decrease, then increase, ie the lowest perigee always associated with the highest apogee, and the average being pretty constant (as I observed). Basically, it doesn't look like any meaningful trends can be observed over a period of only a decade or two, since the natural fluxuations within that period are greater than any long term trend.
So I guess the answer to my question is basically unknown.
I did manage to find (on an old hard drive) a nice geosync sat list from 1998, that I can compare to current similar lists available on the SpaceTrack web page.