Determination of EOP from combination of SLR and VLBI data at the observational level
Proceedings of the Journées 2003 "Systèmes de Référence Spatio-Temporels", A. Finkelstein & N. Capitaine (eds.), Institute of Applied Astronomy, St. Petersburg, 182-188 (2004)
About the paper Full textAbstract
Time series of Earth orientation parameters (EOP) are commonly obtained independently from the processing of high accuracy modern observations such as VLBI, SLR, LLR, and GPS. This paper is devoted to an attempt of determination of EOP series from the joint analysis of SLR and VLBI measurements at the observation level. We used laser ranges to geodetic satellites LAGEOS, LAGEOS 2, and Etalon 1&2. All range measurements are taken from the Crustal Dynamics Data Informational System (CDDIS) and European Data Center (EDC). VLBI observations of distant quasars are obtained from the NEOS-A campaign. Processing of these measurements is performed in two steps. On the fi rst stage the short arc technique with the arc length of 7 days is applied to all SLR measurements to adjust orbital parameters along with coefficients to the radiation pressure reflectance model and along track acceleration terms. All these parameters are considered to be non-stochastic. For VLBI measurements zenith component of troposphere delay and its gradients in horizontal and vertical directions are adjusted as stochastic signals on each day of observation. Both coordinates of quasars and site coordinates are considered to be accurately known and are not improved. It is very important that both SLR and VLBI observations are processed by the same program package, using the same astronomical constants and models for diff erent kinds of measurements. On the second stage SLR and VLBI observations are mixed to determine corrections to variables mentioned above along with all fi ve Earth rotation parameters. Kalman filtering procedure is used to solve the system of conditional equations. Combining SLR and VLBI measurements on the short one day arc makes it possible to get standard deviations of parameters 1.5 times smaller to compare with that obtained by means of each technique separately. Applying Kalman filtering method to the longer observational time span of 7 days allows us to derive EOP variations with subdiurnal periods.