Regarding a satellite with a nadir facing surface, Bjoern Gimle wrote: > I haven't checked the references, nor the reasoning in the mails that > followed Russell's, but to me it seems it wouldn't flare: > Looking from the satellite, the Sun would have to have a zenith angle > greater than 90, but still above the true horizon. > The reflection from the nadir-normal surface will proceed in the same > azimuth, with the same zenith angle, i.e. it will pass above the > horizon, as did the incoming sun rays, and not hit Earth. Technically, Bjoern is correct. If the surface in question is truly = pointed at nadir, then you can never see a perfect specular reflection of the sun from this surface. All you have to do is consider the symmetry of the situation -- put yourself on the satellite. For there to be a specular reflection, the sun must be above the horizon. But since the reflecting surface is pointed at nadir, the reflection direction is = *also* above the horizon -- by the same amount as the sun. In other words, the reflection just misses the earth. Now there are some additional factors that still make a glint possible: 1. the earth isn't exactly spherical, 2. the satellite surface may not be pointed exactly at nadir, but most importantly, 3. satellite surface imperfections will scatter light in a cone centered on the specular direction axis. (i.e. you can probably be at least 3 degrees off the specular axis and still see some obvious brightening.) For a satellite in low-earth orbit, being 3-degrees off the specular axis still doesn't buy you much. For example, if the satellite altitude is 400 km, and the sun is setting (or rising) as seen by the satellite, then the satellite altitude angle corresponding to the 3-degree off-axis direction is only 11.5 degrees. So at your ground site, if the satellite in question was at the same azimuth as the sun, the sun was about 31 degrees below your horizon, and the satellite was no more than 11.5 degrees above your horizon, you could expect to see a glint. The lower the satellite altitude angle, the brighter the glint. But things improve dramatically for satellites at geosynch. Assuming an altitude of 35,900 km, the nadir-angle of the earth limb (from the satellite's perspective) is only about 8.7 degrees. Thus, if the sun were rising or setting at the satellite, the specular direction would be on the opposite side of the world from the sun. If you then relax that angle by 3 degrees (the angle off-specular), then you're looking for the ground altitude angle that corresponds to a satellite nadir angle of 5.7 degrees. That angle is about 49 degrees! In other words, when a geosynch satellite is approaching eclipse, if it is at nearly the same azimuth as the sun (as seen from the ground), it will glint if it is less than 50 degrees above your horizon. The lower the altitude angle, the brighter. For comparison, if your requirement is to be within 1-degree of the specular axis, then the satellite altitude angle needs to be less than 28 degrees above the horizon. Of course, for the geosynch case, the solar elevation angle is critical. You have to be watching the satellite just before eclipse entry, or just after eclipse exit. Every degree that the sun is above the earth limb (as seen from the satellite) is 1 more degree off the specular axis. Thus, two limited geographical areas would be in position to see such a glint each night. (One area just before eclipse entry, another just after eclipse exit). The northern hemisphere would be favored in February, March, October and early November; the southern hemisphere would be favored in April, early May, August and September. The best ground longitudes would be 60 to 80 degrees east or west of the geosynchronous satellite's longitude. Apologies all for the long post! --Rob