On filtering and compressing lageos laser range data

A method of compressing LAGEOS laser range measurements into normal points is described. First, the raw range measurements are screened, on a pass-by-pass basis, by a filter which utilizes the fact that the range varies with time very nearly like a quadratic curve. Second, the time series of the filtered range data over each pass is divided into small segments, typically of 150 s duration. Third, each separate segment is modelled by a unique low-order Chebyshev polvnomial, typically of order four. Fourth, the polynomial is interpolated at some instant, typically that corresponding to the mean time of the segment, to vield the normal point. Baseline measurements across Australia obtained from full rate data and from data compressed into 150 s normal points, agree to within the measurement error. These differences are mainly due to the effects of data distribution and pass geometry. Computer costs are substantially reduced.