Across the three launches, an apparently inconsistent pattern of launch windows and launch times has surfaced. They point to the possibility of the X-37B being imaged soon after launch by a reconnaissance satellite.

Inspection is reminiscent of the post-Columbia space shuttle era when cameras were used by STS crews to check the tiled areas for damage once in orbit. There is a difference between STS and OTV missions in that OTV is launched on top, rather than the side of, a rocket stack. For the critical part of the flight through the lower atmosphere, it is inside a protective shroud so is not exposed to the air flow and should be safe from the sorts of events that used to hit the shuttle.

Vibration is, however, a factor and in the event something should be damaged, there is no crew aboard to wield a camera. Overall condition of the vehicle's exterior is about the only thing than can't be measured by a sensor and sent back as telemetry, hence the need for a remote look.

First - why is there a need to develop a habit of inspecting an orbiting spacecraft anyway? The answer lies in Project Management. With the risk that the thermal insulation could be damaged during launch and lead to the vehicle being lost on re-entry then, with mission likely to be 1-2 years in duration, it's a long time to wait before finding out.

By inspecting it right at the start, if something points to the vehicle not making it back to Earth then planning for building its successor can start earlier rather than later. It may lead to the mission being managed in a different way if it is known that the craft will never be retrieved. There's also avoidance of the embarrassment factor. It can be ditched in the deep Ocean rather than unexpectedly showering Californian coastal waters with re-entry debris!

Information can be gleaned from the pattern of change in planned lift off time from day to day when launches have to be delayed. For the majority of launches, the time either stays the same or drifts by a uniform daily amount. Drift can be driven by on-orbit solar lighting angles, or the need to align the orbit position with those of satellites in the same constellation already in space.

In the case of OTV, potential launch times have jumped around from day to day suggesting that something more than the usual influences are at work. These are some of the times published for OTV launch windows over the years:

The change in the structure of the Mission 2 launch window between 2011 Mar 4 and Mar 5 suggests that all of the longer (2+ hours) windows hide two short windows within them. Using the way the Mar 5 window splits up allows an estimate of what the other windows look like in reality. Note - the Mission 1 windows run across midnight into the following day.

The 'window within a window' format was again confirmed during the ULA Launch webcast covering the final minutes of the Mission 3 countdown. Three launch windows were given for each of Dec 11 and Dec 12:

The pattern seems to be that a time for a launch opportunity on any particular day is fixed and the time for the next day is 30-33 minutes later. After three such steps, the time jumps back by about 90-96 minutes.

Taken together, all the launch window times suggest that there is an underlying cycle time of 96-98 minutes, and that is what drove thinking towards a link with an already-orbiting satellite, or satellites. A candidate that fits the bill is the KH-11 constellation of imaging satellites as they have orbital periods of the right sort of value.

The intervals between OTV windows on a particular day are to ensure that that the OTV and the KH-11 arrive their orbit crossing point simultaneously. The daily drift is to allow for the fact that the KH-11 complete an whole number of circuits in about 24.5 hours.

The KH-11 factor also dictates the period in day when a launch must occur because the X-37B has to be suitably illuminated for a photograph to be taken. The published launch windows point to a possible slot being less than two hours centred around 22:30 UTC.

The possibility of KH-11 involvement was raised at the time of Mission 2 in two postings to the Seesat-L discussion group: http://satobs.org/seesat/Mar-2011/0076.html and http://satobs.org/seesat/Mar-2011/0084.html

A related note in connection with Mission 3 also appears on Seesat-L:
http://www.satobs.org/seesat/Dec-2012/0079.html

Launch and orbit notes for the three missions are extracted from the annual launch diaries: