Thermosphere Density Calibration in the Orbit Determination of Ers-2 and Envisat

The limited accuracy of empirical thermospheric density models is a major contributor to force model errors in orbit determination and orbit prediction for satellites such as ERS and Envisat. Although orbit accuracy has steadily improved due to advances in gravity field modeling and tracking systems, strong evidence of errors in density models have remained apparent in the orbit determination results, especially during high solar-activity periods. Using newly developed software, empirical density modeling accuracy can be enhanced. Two-Line Element (TLE) data of a large number of objects with perigee altitudes between 200-500 km are used as input observations, which are converted to daily density values, averaged along their trajectories. These data are in turn used to adjust a set of daily correction parameters. In this study, the DORIS, SLR and radar altimeter data of ERS-2 and Envisat are used to evaluate the performance of density model calibration. A reduction from 23.5% to 17.1% RMS error was found by analyzing estimated density scale factors after adjusting the NRLMSISE-00 model using 10 parameters per day. These scale factors allow the reduction of long-period density model errors using tracking data. The remaining reduction in density model error over shorter periods, when using a calibrated model, causes only a modest improvement in the accuracy of ERS-2 precise orbits.