Earlier this afternoon Perry and I had a lengthy editorial telephone discussion on the subject of Georgia. While we agreed broadly there was one area in which we had intense debate until I finally figured out how we were talking past each other.
The question is, how the hell did US intelligence assets miss the Russian Black Sea fleet movements? How did they miss the massive transport job of the troops and their logistical tail? They did not just materialize in position. It takes time and planning to make such moves. I will leave the detail of that to Perry as he seems to have been thinking about it in great detail.
My take is there is a limited amount of time available on the black satellites. The manpower and resources have been re-targeted on the Middle East. Orbits have been shifted to give maximal coverage in those areas of interest and experienced personnel have moved to ‘where the action is’.
This is not to say Russia is being ignored. It is however a very big place and I am going to guess that the time between scanning particular areas has greatly lengthened. Russian troop movements are mainly rail based and with enough eyeballs and Cold War era periodic coverage one might hope to pick up changes in traffic patterns and notice “something is going on”. But… this requires a certain periodicity in coverage. Changes in static positions like silos and strategic air bases are much easier to pick up even with occasional coverage. Dynamic changes, such as train and road movements are a different story. You have to have a satellite taking pictures at just the right time or often enough to pick up a signal just by chance.
This is what took Perry and I awhile to meet minds on: I have been thinking of this issue as a communications/information theory problem. How often do you have to sample an area to notice a change in the density of train traffic? I would posit it would have to be several times a week at the very least if the spike in traffic was huge and extended; if the spike were smaller and flatter you would need to sample daily or multiple times daily. You would have to do it at night and through clouds as well if you were to get a statistical value high enough to ring alarm bells. It is an issue of sampling rate versus the highest detectable signal frequency, pure and simple.
I doubt they have even been scanning large areas of Russia more than a few times a week (I suspect much less often) except in areas of nuclear strategic interest. They could easily miss large troop movements in a part of Russia which is not of great national interest to the United States.
Let the discussion begin. There is a lot of meat on this bone!