SBIRS—Two Heads are Better than One

click on the image for a larger version

The currently active US early warning satellites, SBIRS and DSP: each satellite is removed one at a time to highlight its individual coverage. Orbits have been determined by the amateur satellite observer network, see-sat, with particular thanks to Mike McCants, Greg Roberts, Peter Wakelin, and Scott Campbell

Now a ‘days everyone seems to know that you need at least two DSP or SBIRS satellites to track a missile in stereo. (Actually, that’s not exactly true. If only the US and Russia had continued with their RAMOS project, we could have jointly developed the sensors necessary to determine the 3D position of a missile with only a single satellite. That, however, is a different blog altogether.) But that is far from the only reason why we need at least two satellites observing every point on the Earth’s surface. Perhaps an even more important reason for having multiple observations of a missile’s heat signature is to eliminate false alarms.

Most, but not all, of the background from reflected sunlight is eliminated by looking at the Earth in only a very narrow band tuned to the wavelengths absorbed by water. Yesterday, we considered in detail the image taken by SBIRS HEO 2. One of the features of that photograph was a thunderhead that in all likelihood extended high into the atmosphere, past where most of its reflected light would have been absorbed by the surrounding water vapor. Today, we are going to look at a different source of background, sunlight reflected off of low altitude clouds but with a geometry where the sun, clouds, and satellite nearly line up. This results in what is known as specular reflection as opposed to the more common diffuse reflection. The later reflects much less light into the sensor and is, therefore, easier to eliminate as background.

The SBIRS HEO-1 checkout photograph (taken on 14 November 2006 of the launch of the DMSP F17 satellite from the Vandenberg AFB) provides a good example of how bright low altitude clouds can get:

click on the image for a larger version

The Delta IV rocket lifted off from Vandenberg on a generally southern heading (I estimate an azimuth of about 190 degrees, based on the 98 degree inclination of the orbit) which, in the image as presented above, makes it appear to go downward. There is a clear decrease in the track’s luminosity just below the upper part of the track (which is its start) that is a combination of the trough region, as discussed in my post on signal and background, and the 28 second gap between main engine cut off (MECO) and the ignition of the second stage.

The thing I want to discuss about this photograph, however, is the bright background near the top of the image. Again, this background has been “artificially” increased by combining the images taken over several hundred seconds while the signal has been smeared across a number of different pixels as the rocket moves across the scene. However, the clouds near the Earth’s limb are considerably brighter than the clouds appeared in the image taken on 11 June 2008 of the Delta II even though that image was taken at local noon. Here, the sun, Earth, and satellite have almost exactly lined up, producing the enhancement associated with specular reflection. If this had not been over the United States, and if the satellite had been searching for launches as opposed to waiting for an expected launch, it is possible other meteorological—in combination with the alignment—could have produced a false alarm and perhaps triggered a nuclear war. One possibility might be for a storm front to be moving obliquely across limb of the Earth and different thunderheads to be illuminated in turn. Of course, we are talking about a system that continuously watches the Earth for years at a time and is bound to see all sorts of different and unexpected phenomena.

That is exactly what almost happened in 1983 when specular reflection caused many people in Russia’s strategic forces to think the US had launched an attack of half a dozen or so missiles. Fortunately, Col Petrov, the officer in charge of monitoring the newly launched system, decided such a small attack could not possibly be used to start a nuclear war. He was court-martialed for his troubles but at least we didn’t all die.

It is an interesting question just how much this danger is reduced by the SBIRS very high revisit rates; I would guess that an image is taken at least once a second and added into these composites we see here. After all, with more points on the “trajectory” it becomes more difficult for a natural phenomena to fake a missile launch. However, it is best not to rely on that too much. Instead, the US can usually view the same launch from two or more satellites; in this case SBIRS HEO 1, DSP F14 and DSP F17 with a possible contribution from DSP F16 if it used its above-the-earth-limb sensor; a special sensor that is meant to view rocket launches at the edge of the Earth and which appear silhouetted against the black background of space. This is apparently the only why Russia views missile launches but they can still get into trouble from reflections off of clouds and hence maintain early warning satellites in both geostationary and Molynia orbits looking at the central US missile fields from two very different directions. There are still worries, however, that Russia is not maintaining a complete Molynia constellation of early warning satellites and, some fear, could accidently start a nuclear war triggered by some rare weather phenomena or other benign event.

If the US currently has enough early warning satellites for this overlap, some analysts fear that it might not in a few years as DSP ages. But that is another blog.