Using Modified Allan Variance for Time Series Analysis
Allan Variance (AVAR) was introduced more than 40 years ago as an estimator of the stability of frequency standards. Now it is also used for investigations of time series in astronomy and geodesy. However, there are several issues with this method that need special consideration. First, unlike frequency measurements, astronomical and geodetic time series usually consist of data points with unequal uncertainties. Thus one needs to apply data weighting during statistical analysis. Second, some sets of scalar time series naturally form multidimensional vector series. For example, Cartesian station coordinates form the 3D station position vector. The original AVAR definition does not allow one to process unevenly weighted and/or multidimensional data. To overcome these deficiencies, AVAR modifications were proposed in Malkin (2008. On the accuracy assessment of celestial reference frame realizations. J Geodesy 82: 325–329). In this paper, we give some examples of processing geodetic and astrometric time series using the classical and the modified AVAR approaches, and compare the results.
KeywordsAllan variance Time series analysis
This work has made use of data and products provided by the International VLBI Service for Geodesy and Astrometry (IVS, Schlueter and Behrend 2007) and European Permanent GPS Network (EPN, Bruyninx and Roosbeek 2006). The author is grateful to two anonymous reviewers for careful reading of the manuscript and helpful comments and suggestions.
- Bregni S, Primerano L (2005) Using the modified allan variance for accurate estimation of the hurst parameter of long-range dependent traffic. In: arXiv: cs/0510006, MilanoGoogle Scholar
- Bruyninx C, Roosbeek F (2006) The EUREF permanent network: recent achievements. EUREF Publication No. 16, Mitteilungen des Bundesamtes für Kartographie und Geodäsie, Band, vol 40, pp 105–112Google Scholar
- Feissel M, Gontier A-M, Eubanks TM (2000) Spatial variability of compact extragalactic radiosources. Astron Astr 359:1201–1204Google Scholar
- Feissel-Vernier M, de Viron O, Le Bail K (2007) Stability of VLBI, SLR, DORIS, and GPS positioning. Earth Planets Space 59:475–497Google Scholar
- Gordon D, Ma C, Gipson J, Petrov L, MacMillan D (2008) On selection of “defining” sources for ICRF2. In: Finkelstein A, Behrend D (eds) Proceedings of fifth IVS general meeting, pp 261–264Google Scholar
- Le Bail K, Feissel-Vernier M (2003) Time series statistics of the DORIS and GPS colocated observations. Geophysical Reseaech Abstracts, EGS-AGU-EUG Joint Assembly, vol 5, Nice, 6–11 April 2003, p 04078Google Scholar
- Ma C, Arias EF, Bianko G et al. (2009) The second realization of the international celestial reference frame by very long baseline interferometry. In: Fey A, Gordon D, Jacobs CS (eds) Presented on behalf of the IERS/IVS working group, IERS technical note no. 35, Frankfurt am Main, Verlag des Bundesamts für Kartographie und GeodäsieGoogle Scholar
- Malkin Z (2008b) On construction of ICRF-2. In: Finkelstein A, Behrend D (eds) Proceedings of fifth IVS general meeting, pp 256–260Google Scholar
- Malkin Z (2009) Some results of analysis of source position time series. IVS Memorandum 2009-001v01. http://ivscc.gsfc.nasa.gov/publications/memos/