Analysis Standing Committee (ASC) POS+EOP Pilot Project
INTRODUCTION
On March 22 and 23, 2001, the ILRS Analysis Standing Committee had its fourth workshop in Nice, France. The results of the pilot project "positioning + earth orientation" were evaluated there, and the next step on the way to (an) official data product(s) were discussed and defined.
Among the tasks of the ILRS ASC are the coordination of SLR/LLR analyses, the quality control of such results and the development of (an) official ILRS product(s). To this aim, the ASC started activities on these matters in Florence (September 1999).
In particular, pilot projects were defined on various aspects of the SLR and LLR analyses.
The overall goal of these projects is to be able to deliver a combined analysis product with maximum quality to our customers (such as IERS, and the scientific community).
The projects have proven very useful for identifying a variety of analysis problems and has very much stimulated our steps on the road to (an) official ILRS data analysis product(s). At this moment, many activities focus on the pilot project "positioning + earth orientation". The most important new aspects coming out of the meeting in Nice are:
(1) the extension of the set of satellites to be used in the analysis (we have introduced an expansion to also use SLR observations on Etalon-1 and -2), and
(2) the extension of the parameters that are to be solved for (an expansion to estimate the time derivatives of the EOPs, in addition to the standard parameters "coordinates" and "EOPs").
It is important to be aware of the background for this expansion. Currently, the SLR products for earth orientation play a very marginal role in combination solutions, and it is recogniozed that we should do our best to improve the quality of our products, if we want to remain serious partners in this "EOP business". The pilot project "pos+eop" will be a very important tool to assess this fundamental capability and strengthen the products. The main themes of the activities (network/EOP computation and network/EOP comparison/combination) are still present.
Details on the two elements of the project "pos+eop" can be found below.
PILOT PROJECT POS+EOP
In its current phase, the pilot project "pos+eop" is aimed at the computation of a time-series of 4-weekly coordinates solutions for the global network of SLR stations and of daily EOPs, and the comparison and/or combination of the solutions provided by individual ILRS Analysis Centers and Associate Analysis Centers.
The project may provide a tool to study additional parameters like station velocities and geocenter motions, and may evolve into an (official) ILRS contribution to the IERS pilot project on time-series of network solutions and EOPs.
NETWORK/EOPs
As for the analysis part, the characteristics of the dataset and the constraints on the solution are the following:
- Two different data periods:
Period A: Sunday January 3, 1999 (inclusive) until Saturday January 1, 2000 (inclusive). SLR observations on LAGEOS-1 and LAGEOS-2.
Period B: Sunday April 1, 2001 (inclusive) until ... SLR observations on LAGEOS-1 and LAGEOS-2 (mandatory), and on Etalon-1 and Etalon-2 (expansion).
- Parameter solutions:
Period A: Station coordinates at the midpoint of each 4-week data interval, plus 1-day EOP (x/y-pole and UT1-UTC).
Period B: Here we can have 4 different solutions, depending on which satellites are used (LAGEOS-1 and LAGEOS-2, or the LAGEOS satellites and Etalon-1 and Etalon-2), and which parameters are solved for (EOPs only, i.e. x/y-pole and UT1-UTC, or EOPs plus EOPdots, i.e. also the time- derivatives of the parameters mentioned here, including LOD):
---------------------------------------------------
option satellites parameters
---------------------------------------------------
B1 LAGEOS-1/2EOP
B2 LAGEOS-1/2 + Etalon-1/2EOP
B3 LAGEOS-1/2EOP + EOPdot
B4 LAGEOS-1/2 + Etalon-1/2EOP + EOPdot
---------------------------------------------------
Analysis centers are encouraged to contribute as many of these solutions (A, B1-B4) as possible. For each possibility, a combined satellite product is requested (i.e. no solution for LAGEOS-1 or LAGEOS-2 individually). Please be aware that we like to improve the quality of our products as much as possible, something which can be done only if we are able to compare results obtained for the various options listed above. All options are important, but the best results should be expected (and identifiable as such) for option B4.
- The observations are to be processed in batches of exactly 4 weeks.
As for period A, they start on:
Sunday January 3, 1999
Sunday January 31, 1999
Sunday February 28, 1999
Sunday March 28, 1999
Sunday April 25, 1999
Sunday May 23, 1999
Sunday June 20, 1999
Sunday July 18, 1999
Sunday August 15, 1999
Sunday September 12, 1999
Sunday October 10, 1999
Sunday November 7, 1999
Sunday December 5, 1999
As for period B, they start on:
Sunday April 1, 2001
Sunday April 29, 2001
Sunday May 27, 2001
Sunday June 24, 2001
Sunday July 22, 2001
Sunday August 19, 2001
.
.
.
- For each of the 4-week batches: a priori station positions taken from ITRF2000, at the midpoint of the 4-week data period (so, the epoch for the first 4-week batch is January 17, 00:00 hrs). All station coordinates must be solved for.
- Within each of these batches, the station velocities have to be taken from ITRF2000 and kept fixed at their values. In case ITRF2000 velocities are not available, use NUVEL-1A-NNR values (and keep these fixed too).
- The solutions for the EOPs and EOPdots are requested for 1-day intervals, which start on January 3, 1999, 00:00 hrs, and (of course) end exactly 24 hours later (and so on). The EOP and EOPdot solutions are to refer to the midpoint of these 1-day intervals (i.e. January 3, 12:00 hrs, and so on).
- No requirements on analysis models or technique; use the IERS 1996 Conventions as a starting point. You are bly encouraged to use the official center-of-mass value for LAGEOS-1 and LAGEOS-2: 251 mm.
- Coordinates solutions refer to SLR monuments whereever possible (i.e. subtract the eccentricity vector if available). As for Zimmerwald and San Fernando, refer to the optical centers of the new telescopes. As for Simosato: also refer to the optical center of the telescope. The eccentricity vector which is present in the CDDIS files (2.10 meter up) should be ignored.
- DOMES numbers to be used as ids for coordinate solutions.
- Results are to be made available in a SINEX file. The ILRS implementation of SINEX is at:
sinex_file_description.html. and the official SINEX format is located at:
ftp://alpha.fesg.tu-muenchen.de/iers/sinex/format/
- Parameters must be cartesian coordinates [m] for positions, and x/y-pole coordinates [mas], UT1-UTC [ms], x/y-pole derivatives [mas/d] and LOD [ms] for the EOPs and EOPdots, respectively. In addition, you may solve for range and troposphere biases [m] and time biases [ms] (if you do so, include the solutions in the SINEX files).
- Results are to be accompanied by 1-sigma uncertainties and a covariance or correlation matrix (see also the SINEX requirements); the (a priori and a posteriori) uncertainties of the parameters have to be scaled by the variance factors as specified in the (ILRS) SINEX description.
- The solutions must be loosely constrained: use for the good (poor) stations an a priori standard deviation of at least 1 (10) m, and for the EOPs the equivalence of at least 1 m. Or use scaled versions of these values if you use measurement weights which are not in the order of cm. Do not fix anything (other than velocities)!
- Observations:
Period A: The LAGEOS-1 and LAGEOS-2 data are already available in the ILRS NP format by anonymous ftp from cddis.gsfc.nasa.gov, as: ftp://cddis.gsfc.nasa.gov/pub/slr/products/ilrsac/data, and in this directory the files qldata.990103-000101.lageos1 and qldata.990103-000101.lageos2.
Period B: These data will become available as time goes. The data will be stored in the same directory as mentioned above, with the following naming convention (the 1st period is taken as an example here): qldata.010401-010428.lageos1, qldata.010401-010428.lageos2, qldata.010401-010428.etalon1, and qldata.010401-010428.etalon2. The datasets for the other periods will have similar names, but with different ids for the begin and end dates of course. The datsets for period B will be availabe within 1 week after the end of a particular 4-week period. The files include the observations as they have been submitted to CDDIS: no screening has been applied. Every analyst is free to treat the data in the way he/she thinks best to achieve the most accurate coordinates and EOP/EOPdot solutions possible.
The procedure for submitting is similar to the one for previous submissions: give your SINEX file a name like "INST.YYMMDD.solN.vM.snx", where INST is to be replaced by an abbreviation of your institute (like "bkg", or "nasa", or "crl"), YYMMDD is to be replaced by the beginning of the 4-week interval (including possibly leading zeroes), N is the solution id (use "A" for the LAGEOS-1/2 solution based on the 1999 data, "B1" for the LAGEOS-1/2 solution for coordinates and EOP based on the 2001 data, and "B2", "B3" and "B4" for the solutions as described above; this will help to identify the contents without having to open each and every file), and M is the version number (start with 1).
The solutions are to be submitted by ftp to cddis.gsfc.nasa.gov. Notify Carey Noll (carey.noll@nasa.gov) and send a "cc" to me (ron.noomen@deos.tudelft.nl). Carey Noll will transfer "A" and "B" solution(s) to the archive directories.
COMPARISON/COMBINATION As for the comparison/combination, please note the following: - It may be on station coordinates and/or EOP or EOPdot solutions. - Use the solutions that will be available at the archive directory. - In order to be able to make sure that we will have comparable results on the table during our next meeting, the comparison/combination centers must at least assess the following things: * internal quality of the time-series of solutions, as provided by each individual analysis center. * internal quality of the individual solutions, w.r.t. the combination product. * external quality of both the individual solutions and the combination by comparing it with "independent" solutions (C04, Bulletins A and B, IGS, IVS, ... for EOPs; ITRF2000, IGS, ... for station positions). Comparisons may be done on a 1-dimensional basis (i.e. baselines, geodesics) or on a 3-dimensional basis (coordinates). Since the ultimate goal of this project is a full 3-dimensional product, participants are encouraged to work towards such a product. In particular the qualification of the solutions B1 through B4 (see above) is a very important element of this next phase of the project. This activity will hopefully lead to a situation where the ILRS will have a number of official comparison/combination centers, one directly responsible for the official product and the other(s) performing a benchmarking role. Very much like the situation that has been in use at IGS so successfully for a number of years now.
