>One or two on this list have discussed how to find the location >for the brighest flare of a particular Iridium pass, and have found >that, in fact, just a few miles can make a large difference in magnitude= =2E >The idea is to enter different lat/longs into the programs and = >check the results. This is, of course, rather long-winded. >I just wonder if the problem could be tackled by the existing authors The point where the sun light reflected by an Iridium mirror hits the Earth's surface can indeed be calculated with a modification of the algorithm kindly published by Randy John on his web page. Essentially, the topocentric satellite position vector must be replaced by the vector Sun -> satellite which is very easy to calculate. The reflected vector (calculated in exactly the same way as the topopocentric satellite position vector) may or may not intersect the Earth's globe. This boils down to the mathematical problem of finding the intersection between a line in space and a sphere. People knowledgeable in 3D analytical geometry can probably solve it by hand, I myself have done it with MATHEMATICA about a year ago for a different purpose (finding the transit of a satellite in front on the Moo= n or the Sun). The final step is to replace the sphere by the Earth's geoid. I have developed a little algorithm for that which works by iteration. The program, still under development, works quite well. For the time being it gives the latitude/longitude as a function of time of the point where the reflection angle would be zero. = I would not attempt to calculate the visiual magnitude, at least for the time being. The relationship between reflection angle and magnitude is not well enough known. The data points on curves published so far have= a far too large scatter. If someone is interested, I can give more details. For the sake of all the readers on this list who are not intersted in this rather special= subject, I would suggest to contact me privately. Bruno Tilgner Saint-Cloud, France Bruno_Tilgner@compuserve.com