Description

ITRF2020 is the new realization of the International Terrestrial Reference System. Following the procedure already used for previous ITRF solutions, the ITRF2020 uses as input data time series of station positions and Earth Orientation Parameters (EOPs) provided by the Technique Centers of the four space geodetic techniques (VLBI, SLR, GNSS and DORIS), as well as local ties at colocation sites. Based on completely reprocessed solutions of the four techniques, the ITRF2020 is expected to be an improved solution compared to ITF2014. A number of innovations were introduced in the ITRF2020 processing, including:

• The time series of the four techniques were stacked all together, adding local ties and equating station velocities and seasonal signals at colocation sites;
• Annual and semi-annual terms were estimated for stations of the 4 techniques with sufficient time spans;
• Post-Seismic Deformation (PSD) models for stations subject to major earthquakes were determined by fitting GNSS/IGS data. The PSD models were then applied to the 3 other technique time series at earthquake colocation sites.

Read more...

Input data

Space geodesy solutions from 4 space geodesy techniques. The used time series of solutions of space geodesy are summarized in the following table, indicating for each one, the time span and the type of constraints.

Computation strategy

The strategy adopted for the ITRF2020 generation consists in rigorously stacking the time series of solutions from the four space geodetic techniques all together, adding local ties and equating station velocities and seasonal signals at colocation sites. To that end, intermediate steps listed below were operated in order to mainly identify and reject outliers and properly assign and handle discontinuities and Post-Seismic Deformation (PSD) models for stations subject to major earthquakes.

The main intermediate steps operated prior to the overall ITRF2020 combination are:

• Apply minimum constraints equally to all loosely constrained solutions: this is the case of SLR weekly solutions;
• Apply No-Net-Translation and No-Net-Rotation constraints to IVS solutions provided in the form of normal equations;
• Use as they are minimally constrained solutions: this is the case of IGS daily and IDS weekly solutions;
• PSD parametric models were determined using GNSS/IGS data, and then applied to all technique time series at earthquake colocation sites, prior to the next step;
• Form per-technique long-term solutions (TRF + EOP), by rigorously stacking the time series, solving for station positions, velocities, EOPs and 7 transformation parameters for each weekly (daily in case of GNSS and session-wise for VLBI) solution with respect to the per-technique cumulative solution. During the iterative stacking process:
  • Annual and semi-annual signals were estimated for stations with sufficient time-spans;
  • For the particular case of IGS time series, in addition of the annual and semi-annual signals, periodic signals at the first 8 GPS draconitic harmonics were estimated and then removed from the IGS time series.
  • PSD model corrections were applied to earthquake sites, prior to the construction of the normal equation of the stacking of the time series.

Frame Definition

Origin

The origin of the ITRF2020 long-term frame is defined in such a way that there are zero translation parameters at epoch 2015.0 and zero translation rates between the ITRF2020 and the ILRS SLR long-term frame over the time-span 1993.0-2021.0, using the concept of internal constraints.

The origin of the ITRF2020 seasonal signals expressed in the CM (as sensed by SLR) frame is defined in such a way that there is no seasonal translation between the ITRF2020 seasonal signals and the input SLR solutions over the time-span 1993.0-2021.0, using the concept of internal constraints.

The origin of the ITRF2020 seasonal signals expressed in the CF frame is defined in such a way that the weighted sum of the seasonal signals of a selected subset of stations (approximating an integral over the Earth’s surface) is zero.

Scale

The scale of the ITRF2020 long-term frame is determined using internal constraints, in such a way that there are zero scale and scale rate between ITRF2020 and the scale and scale rate averages of VLBI selected sessions up to 2013.75 and SLR weekly solutions covering the time-span 1997.7 – 2021.0.

The scale of the ITRF2020 seasonal signals, expressed in both the CM and CF frames, is determined using internal constraints, in such a way that the average of the seasonal scale variations between the selected SLR solutions and ITRF2020 is zero.

Orientation

The orientation of the ITRF2020 long-term frame is defined in such a way that there are zero rotation parameters at epoch 2015.0 and zero rotation rates between the ITRF2020 and ITRF2014. These two conditions are applied over a core network (see section transformation parameters between ITRF2020 and ITRF2014).

The orientation of the ITRF2020 seasonal signals, expressed in both the CM and CF frames, is defined in such a way that there is no net seasonal rotation of that same core network.