FAQ on Multiyear Reprocessing and Comparisons Between Current and Previous NAD 83 Realizations
The National Geodetic Survey (NGS) operates the Continuously Operating Reference Station (CORS) Network that provides Global Navigation Satellite System (GNSS) data in RINEX format, and associated positions and velocities (coordinates) for each site. Currently NGS provides these coordinates in ITRF00 epoch 1997.00, and in NAD 83(CORS96,MARP00,PACP00) epoch 2002.00. NGS has completed a full reanalysis of all data from CORS and a set of global sites with the goal to compute a fully consistent set of coordinates, GPS satellite orbits and Earth Orientation Parameters (EOP). This Multi-Year CORS (MYCS1) effort is the first of a series reprocessing projects that will occur periodically in the coming years.
| FAQ 1: | Why Reprocess? |
| FAQ 2: | What Datum Tags, Reference Epoch dates and Antenna Calibration values were and are used? |
| FAQ 3: | How did we reprocess the data? |
| FAQ 4: | What do the residuals from stacking weekly solutions (time-series plots) look like for a particular CORS? |
| FAQ 5: | What reference frames are defined after reprocessing? (includes plate-wide velocity plots) |
| FAQ 6: | What has changed between NAD 83(2011,MA11,PA11) epoch 2010.00 and NAD 83(CORS96,MACP00,PACP00) epoch 2002.00? (includes plate-wide and regional plots) |
| FAQ 7: | What coordinates will be assigned to new CORS or CORS that have less than 2.5 years of data? |
| FAQ 8: | How does the change in antenna calibration values from relative to absolute impact users? |
| FAQ 9: | What about the passive network that is currently aligned to NAD 83(2007) epoch 2007.00 (also known as NSRS2007)? |
| FAQ 10: | How will NGS products and services be affected when the new coordinates are released? |
FAQ 1: Why Reprocess?
1) Mixed coordinates from an earlier solution (1994-2002) that used only 3-8 ITRF reference frame sites to align to the global frame
2) Mixed horizontal velocities: modeled from HTDP vs. computed
3) Mixed vertical velocities, which for many of the NAD 83(COR96,MARP00,PACP00) epoch 2002.00 stations are assigned a value of 0 mm/yr
4) NGS's current global frame is ITRF00 epoch 1997.00; projecting positions 13 years to the present is unrealistic
5) NGS's current plate-fixed frame is NAD 83(CORS96,MARP00,PACP00) epoch 2002.00; projecting positions 8 years to the present is also questionable
6) Implementing significant changes to existing processing algorithms and models including compliance with IERS conventions (i.e. software changes)
7) Switching from using relative antenna calibration to absolute antenna calibrations
8) The current state-of-the-art full (~230 stations) global reference frame: IGS08 epoch 2005.00
Given these major inconsistencies and changes, NGS elected to reprocess all its CORS data to provide a single consistent set of coordinates for all sites computed using the best available methods.
FAQ 2: The following Datum Tags, Reference Epoch dates and Antenna Calibration values are used throughout this page.
Table 1 Datum Tags, reference epochs and antenna calibration values
| Frame Type | Frame Name | Ref. Epoch | Antenna PCV Type* | Data Duration | |||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Global | ITRF2000 | 1997.00 | RelAnt | 1994.0-2002.0 | |||||||||||||||||||||||||||||||||||
| Global | ITRF00 (NGS's soln) | 1997.00 | RelNGSAnt | 1994.0-present | |||||||||||||||||||||||||||||||||||
| Plate-fixed-North American | NAD 83(CORS96) | 2002.00 | RelNGSAnt | 1994.0-present
| Plate-fixed-Pacific | NAD 83(PACP00) | 2002.00 | RelNGSAnt | 1994.0-present
| Plate-fixed-Marianas | NAD 83(MARP00) | 2002.00 | RelNGSAnt | 1994.0-present
| Global | IGS08 | 2005.00 | AbsIGS08Ant | 1997.0-2009.5
| Global | IGS08 (NGS's soln) | 2005.00 | AbsIGS08Ant | 1994.0-2010.5 (ongoing)
| Plate-fixed-North American | NAD 83(2011) | 2010.00 | AbsIGS08Ant | 1994.0-2010.5 (ongoing)
| Plate-fixed-Pacific | NAD 83(PA11) | 2010.00 | AbsIGS08Ant | 1994.0-2010.5 (ongoing)
| Plate-fixed-Marianas | NAD 83(MA11) | 2010.00 | AbsIGS08Ant | 1994.0-2010.5 (ongoing)
| |
* Rel=relative antenna calibrations; Abs=absolute antenna calibrations
Links to IGS on implementation of IGS08 and IGS08.atx
Upcoming switch to IGS08/igs08.atx
Upcoming switch to IGS08/igs08.atx - Details on igs08.atx
Advice for coping with the change to IGS08+igs08.atx
igs08.atx including GLONASS satellite antenna corrections
FAQ 3: How did we reprocess the data?
RINEX observations from the set of global sites were reduced to obtain weekly sets of station positions, satellite orbits, and Earth Orientation Parameters, using the latest NGS PAGES software. These weekly "global" solutions were submitted to the IGS for their IG1/repro1 data reprocessing campaign). Then, CORS data were tied to the framework of the global solution on a weekly basis at the normal equation level resulting in weekly SINEX files containing the positions and full variance-covariance information for CORS plus global stations. Regularized positions and secular velocities were obtained by stacking the SINEX files with the CATREF software from Institut Géographique National (IGN), Paris, France. These new coordinates form the basis for NGS's realization of IGS08, and thus, NAD 83(2011).
For a summary of the technical aspects of our RINEX data reduction strategy, view our "IGS Analysis Strategy Summary".
For additional information on the details of the stacking, follow this link. This link has not been updated in some time and some information may be outdated.
All the results presented on this webpage use data spanning 1 January 1994 to 16 April 2011 and are from our stacked solution using computed coordiates. No results for sites with modeled velocities are shown on this page.
FAQ 4: What do the residuals from stacking weekly solutions look like for a CORS?
From the list below, select a CORS site (or two) that you are interested in and then click "Plot Both".
What you see in the plots is the weekly residual position as a green dot and its associated error bar as a blue line. Vertical magenta lines indicate discontinuities that may represent an equipment change, an earthquake, or an unknown event.
| Select Site 1 | Select Site 2 |
|---|---|
Figure 1 CORS site Time series plots
FAQ 5: What reference frames are defined after reprocessing
The solutions obtained from the processing are first aligned to the global reference frame IGS08 epoch 2005.00
Figure 2 Site velocities with respect to IGS08 epoch 2005.00. Horizontal velocities are shown as black vectors. Vertical velocities are shown as colored surface.
The large positive vertical velocities towards Hudson Bay reflect ongoing visco-elastic response of the earth from unloading of continental glaciers during the last glaciation
The IGS08 epoch 2005.00 solutions are then transformed to a plate-fixed NSRS consistent frame namely NAD 83(2011) epoch 2010.00.
Figure 3 Site velocities with respect to NAD 83(2011) epoch 2010.00. Horizontal velocities are shown as black vectors. Vertical velocities are shown as colored surface.
The horizontal velocities across central to eastern CONUS show a pattern indicating that not all the motion of North America (see Figure 2) has been fully removed. The vertical velocities are similar to those in Figure 2.
FAQ 6: What has changed between NAD 83(2011,MA11,PA11) epoch 2010.00 and NAD 83(CORS96,MACP00,PACP00) epoch 2002.00?
There has been no datum change, that is, the origin, scale and orientation of NAD 83(2011,MA11,PA11) are identical to those of NAD 83(CORS96,MACP00,PACP00). The changes in coordinates are driven by multiple factors including the switch to absolute antenna calibrations, new/revised processing algorithms, improved discontinuity identification, longer data sets, and our improved definition of the global reference frame. However for the end user the biggest change is caused by the new reference epoch, which has changed by 8 years from 2002.00 to 2010.00.
To show how small the change in reference frames is, we compare the coordinates of all CORS sites with more than 2.5 years of data at the same epoch of 2002.00 in both frames:
Figure 4 Difference in positions of NAD 83(2011) epoch 2002.00 minus NAD 83(CORS96) epoch 2002.00 Horizontal (left), Vertical (right).
Average differences of NAD 83(2011) epoch 2002.00 minus NAD 83(CORS96) epoch 2002.00:
Mean values of: East -0.14±1.04 cm; North 0.19±0.94 cm; Vertical 0.80±1.89 cm
Median values of: East -0.10 cm; North 0.20 cm; Vertical -0.70 cm
What we see is that the changes for most stations are very small and generally random in the horizontal (left)(remember that the bulk of the sites have such small differences that they are hard to see on the map).
In the vertical (right) the changes are a little larger, but more systematic, this is largely attributed to the fact that in our re-processed solution NAD 83(2011), vertical velocities were computed and not assigned a 0 cm/yr velocity as occurred for many NAD 83(CORS96).
For a more detailed comparison please use the table below that lets you see a blowup for each state. Please note that for some sites the labels may overlap if sites are very close e.g. For USCG sites, which are 40m apart, this will also result in multiple vectors starting from the same point.
| Select Map | |
|---|---|
Figure 5 Detailed maps by state or region of differences in horizontal (left) and vertical (right) positions of NAD 83(2011) epoch 2002.00 minus NAD 83(CORS96) epoch 2002.00
For many users, when they compare the positions they might compare the positions at the two different epochs (2002.00 vs 2010.0) and then the changes will be larger. Remember if a site has a velocity in NAD 83 of only 2 mm/yr over 8 years this will result in a change of 1.6 cm!
To illustrate this we compare the coordinates of all CORS sites with more than 2.5 years of data at the two reference epochs of 2010.00 and 2002.00.
Figure 6 Difference in positions of NAD 83(2011) epoch 2010.00 minus NAD 83(CORS96) epoch 2002.00 Horizontal (left), Vertical (right).
Average difference of NAD 83(2011) epoch 2010.00 minus NAD 83(CORS96) epoch 2002.00:
Mean East 0.05±5.25 cm; North 2.12±6.08 cm; Up -0.66±2.24cm
Median values of: East -0.12 cm; North 0.00 cm; Vertical -0.80 cm
What we see in Figure 6 is that the changes for most stations are larger and some systematic regional patterns unlike when we compared positions at the same epoch. The main cause of these differences reflects errors in our plate fixed (NAD 83) velocity estimates. The errors however are greater than they seem since we are mixing true plate fixed sites in the Central and Eastern US with sites in the western US that are in the plate boundary zone between the Pacific and North American Plate.
Because we are comparing positions at two different epochs the differences in position are dominated by non-zero NAD 83(2011) velocities. In the eastern CONUS we see a small amount of motion related to the non-zero velocities. Horizontally, in westernmost CONUS, we instead see the dominant effects of the broad deformation zone from the North American Plate to the Pacific plate in westernmost California, which is moving to the northwest. In Washington and Oregon, we see the effects of the subducting Juan de Fuca plate beneath the North American Plate. The subduction causes deformation to the northeast. Vertically (right), it appears that most positions have increased (hence arrows are negative--2010.00-2002.00) reflecting the incorrect assumption used in many NAD 83(CORS96) epoch 2002.00 sites that the vertical velocities were 0 cm/yr.
For a more detailed comparison, please use the table below that lets you see a blowup for each state. Please note that for some sites the labels may overlap if sites are very close e.g. For USCG sites, which are 40 m apart, this will also result in multiple vectors starting from the same point.
| Select Map | |
|---|---|
Figure 7 Detailed maps by state or region of difference in horizontal (left) and vertical (right) positions of NAD 83(2011) epoch 2010.00 minus NAD 83(CORS96) epoch 2002.00
FAQ 7: What coordinates will be assigned to new CORS or CORS that have less than 2.5 years of data?
For new CORS we will publish positions in both IGS08 epoch 2005.00 and NAD 83(2011) epoch 2010.00, and assign modeled velocities from HTDP to them. After ~2.5 years we will have sufficient data to compute robust velocities and we will change the published velocities from modeled to computed values.
FAQ 8: How does the change in antenna calibration values from relative to absolute impact users?
The new coordinates IGS08 epoch 2005.00 and NAD 83(2011) epoch 2010.00 were established using IGS08 absolute antenna phase center patterns. If a user processes data using relative antenna phase center values instead of absolute antenna phase center values they may get positions that differ by up to a few centimeters as compared to processing using absolute antenna phase center values. Users must therefore change the antenna phase center values they use when processing data. NGS has established a beta website with IGS08 consistent absolute antenna phase center values that users can download to test in conjunction with the coordinates listed on this page. For more detailed information on absolute antenna calibrations, especially the FAQ go here.
FAQ 9: What about the passive network that is currently aligned to NAD 83(2007) epoch 2007.00 (also known as NSRS2007)?
NGS is in the process of adjusting most, if not all, of the passive vector control in the NGSIDB to make it consistent with MYCS1 solution. The results of this adjustment (National Adjustment of 2011) will have the same reference frame tag and epoch date as the MYCS1, namely NAD 83(2011) epoch 2010.00. The timeline for completing this project has not yet been finalized, but every effort is being made to get this done as quickly as possible. Please consult the FAQ created by the passive adjustment team.
FAQ 10: How will NGS products and services be affected when the new coordinates are released?
We have made every effort to minimize the impact of this coordinate change to all NGS products and services in terms of formatting changes. At this time we will change the name of the frame, but the spacing in all our products and services should remain the same.
CORS Coordinate Sheet/PV-File
New name - {ssss}_08.coord.txt where {ssss} is the 4-character id of a CORS site.
Old name - {ssss} where {ssss} is the 4-character id of a CORS site.
CORS Data Sheets
CORS data sheets with new coordinates will not be available till after the passive adjustment is completed.
UFCORS
UFCORS will provide in its compressed archive both CORS coordinate files as described above. As of 6 September 2011 this new file has not yet been implemented, but will very shortly.
Online Storage/Anonymous FTP servers
No change in the structure of the online storage has been made. Some file names have changed and a new directory for the new coordinates has been added. For details see the README.txt.
OPUS
OPUS will allow users to choose whether they want to upload their solution and get the results in old or new coordinates. This option will remain in place until the passive adjustment is completed.
Do you have a suggestion/comment/question about this page please email:
Suggestion/Comment/Question about the MYCS1
