Frequently Asked Questions for
USGG2012/GEOID12A/USDOV2012/DEFLEC12A
USGG2012/GEOID12A/USDOV2012/DEFLEC12A
Last updated July 16, 2011
Q: Why does NGS produce the new geoid models?
A: Survey monuments do not necessarily physically change but their location coordinates may. Location coordinates can change for several reasons. They may change due to natural causes, such as tectonic plate shift as in areas of the western United States or Alaska, or sediment deposition and consolidation as in southern Louisiana, or glacial rebound as in the Great Lakes region. They also may change due to physical changes by man, such as extraction of minerals, as in areas around Houston Texas, or extraction of water as in southern Arizona.
While land masses (which contain physical survey monuments) may change, it is the task of NGS to report location coordinates relative to consistent physical models.
The goal of a geoid model is to be consistent with actual physical characteristics as determined from current observations. Older geoid models, such as GEOID99 and even GEOID03, were built using observation data from the time of their formation, but are considered no longer consistent with the physical earth.
Hence, we periodically produce new geoid models to reflect the changes in both types of heights (orthometric heights are also periodically updated) and to include new observations.
Q: What is the format for the data files?
A: The NGS .bin file format for any sub-grid file is identical: A 44 byte header followed by "nla" rows of data, each row being "nlo" elements long, each element being a 4 byte floating point number. The format chosen is known in FORTRAN lingo as "direct access binary". The exact ordering of the bytes is mapped below:
Bytes Data Variable Variable
Type Name Description
First Record:
1- 8 real*8 glamn Southermost Latitude of grid (decimal degrees)
9-16 real*8 glomn Westernmost Longitude of grid (decimal degrees)
17-24 real*8 dla Latitude spacing of grid (decimal degrees)
25-32 real*8 dlo Longitude spacing of grid (decimal degrees)
33-36 int*4 nla Number of rows of grid
37-40 int*4 nlo Number of columns of grid
41-44 int*4 ikind Set to "1", meaning the gridded data is "real*4"
Subsequent Records:
1- 4 real*4 data(1,1) Gridded value at element 1,1 (Southwest corner)
. . .
The rest of the file continues as 4 byte real values, filling in first the
south row (data(1,nlo) being the last variable in the south row), and then
proceeding northward.
The total number of bytes in a "*.bin" file is:
44 + 4*nla*nlo
Current models have dla and dlo equal to one arc-minute (0.01666666 degrees).
The data following the header record, as I said above, row-major.
Hence, each data record is a list of values "nlo" long (equating to the longitudinal intervals along the row of latitude)
and the record length = "nlo" * 4 bytes
Then the next northern row of data will be listed. This continues until the northernmost row, which is at the bottom of the file.
Hence, the first value in the data is the SW corner of the grid (westernmost point of southernmost row),
while the last point is in the NE corner (easternmost value on the northernmost row).
The binary data format is direct access. This means that the usual head/footer word that FORTRAN appends has been stripped off (just extra baggage). This is more akin to what C or other languages would read.
Inspect the XNTG_V12.ZIP files, available from various the GEOID12A and USGG2012 download pages. This program is designed to read a binary format input file and write an ASCII format output file.
Q: Which models do I use for which regions?
A: The names "USGG2012" and "GEOID12A" have been reserved for each region that is now covered including the conterminous United States (CONUS), Alaska, Hawaii, Guam & the Commonwealth of the Northern Marianas Islands (CNMI), American Samoa, and Puerto Rico and the U.S. Virgin Islands (PRVI). The gravimetric geoid models (USGG2012) all are based on a GRS-80 ellipsoid shell in the IGS08 reference frame and a geopotential surface (W0) of 62,636,856.00 m**2/s**2. These models collectively define the same geopotential surface (geoid) determined from the underlying reference global earth gravity model (EGM2008). Hence, comparisons of these heights between different regions provide consistent values.
However, most people are interested in the hybrid geoid height models (GEOID12A). A hybrid geoid model can transform between a NAD 83 ellipsoid height and the relevant vertical datum for each region. For CONUS and Alaska, a hybrid geoid model can transform from a NAD 83 ellipsoid height to a NAVD 88 orthometric height. For Puerto Rico, the model can yield a PRVD02 elevation. Each region has its own vertical datum with the exception Hawaii. Heights in Hawaii refer to NAD 83 but also to the same geoid surface (W0) defined by the gravimetric geoid. The particular "flavor" of NAD 83 is also used for each region (i.e., NAD 83 (PA11) for Hawaii and American Samoa).
So use the hybrid geoid for each region to transform between NAD 83 and the relevant vertical datum.
| Area | Ellipsoid Reference Frame | Vertical Datum | Latitude | Longitude | ||
| Min | Max | Min | Max | |||
| Conterminous US | NAD83 (2011) | NAVD88 | 24N | 58N | 60W | 130W |
| Alaska | NAD83 (2011) | NAVD88 | 49N | 72N | 126W | 188W |
| Hawaii | NAD83 (PA11) | See Note Below | 18N | 24N | 154W | 161W |
| Guam and Northern Mariana Islands | NAD83 (MA11) | GUVD04/NMVD03 | 11N | 18N | 143E | 146E |
| American Samoa | NAD83 (PA11) | ASVD02 | 17S | 11S | 186E | 192E |
| Puerto Rico / U.S. Vigin Islands | NAD83 (2011) | PRVD02 and VIVD09 | 15N | 21N | 64W | 69W |
NAVD88 - North American Vertical Datum of 1988
ASVD02 - American Samoa Vertical Datum of 2002
GUVD04 - Guam Vertical Datum of 2004
NMVD03 - Northern Marianas Vertical Datum of 2003
PRVD02 - Puerto Rico Vertical Datum of 2002
VIVD09 - Virgin Islands Vertical Datum of 2009 (pending)
Note: The vertical datum surface in Hawaii is identical to that of USGG2012, which is offset by 50-60 cm from some Local Tidal bench mark values.
Q: What is the major difference between GEOID09 and GEOID12A?
A : Both determine the respective vertical datum surface for each region (e.g., NAVD 88 for CONUS). The differences arise from the underlying gravimetric geoids (USGG2009 and USGG2012, respectively) and control data sets (GPSBM2012 and OPUSDBBM12) available at the time of generation.
Both USGG2009 and USGG2012 were based on the EGM2008 reference model. However, GOCO02S signal was incorporated into the reference model in 2012 and the Residual Terrain Model signal from 3 arcseconds through 5 arcminutes was accounted for in 2012 and not in 2009. Hence, USGG2012 has better and more consistent terrain models, which affects the shortest wavelengths of the gravity field. The expectation is that USGG2012 is significantly more accurate than USGG2009. Why is this important?
Because the gravimetric geoid determines most of the actual features in the hybrid geoids. GPS-derived ellipsoid heights on leveled bench marks (GPSBM's) give the separation between NAD 83 and NAVD 88 at discrete points. These differences can be used to develop a conversion surface that effectively warps the gravimetric geoid surface to fit the GPSBM points. The resulting hybrid geoid then fits at the discrete GPSBM points, while still honoring the shorter wavelength features determined by the gravimetric geoid model in between the GPSBM's. Hence, the GPSBM's are also fairly important. GPSBM's continue to change though - their values are not static. There are regions where the ground changes constantly, but most "changes" in height at the BM locations are due to a different realization of the heights. An adjustment is performed using different parameters, and a new value is determined for a height at a given location. The actual position of the height didn't change, just our understanding of where it is located.
Since 2009, NGS performed the National Adjustment of 2011. This significantly affected the ellipsoid heights of data all around the country. Since the GPSBM's are determined from the difference between the ellipsoidal (NAD 83) and orthometric (NAVD 88) heights, a change in one height type changes the control value and thus affects the hybrid geoid.
GEOID09 was fit to the data available in 2009. If you are still using data from 2009, then GEOID09 will fit better. However, NGS has changed the heights in the database for many of these points. Hence, GEOID09 will no longer fit the current values to (in some cases) better than the dm-level. GEOID12A was developed using current values. Hence, GEOID12A should be used now if you want to get results consistent with those showing on the NGS datasheets (our current best estimate of actual heights) as well as OPUS solutions.
This brings up an additional point. While most data were from the NGS Integrated Database (and are listed as GPSBM2012), several hundred supplemental points were obtained from the OPUS Database (OPUSDBBM12) and provideadditional coverage in between the GPSBM2012 control data with decimeter level impacts.
Both GEOID09 and GEOID12A yield estimates of NAVD 88. However, GEOID12A fits better based on our current understanding of the true position of the coordinates. This will change, too. As we develop better coordinates, the values in the database and expressed on the datasheets will change. Hence, GEOID12A will eventually become obsolete and need to be replaced with an update that reflects the future database.
Q: If I know the geoid height model used to compute an orthometric height,
how can I obtain an orthometric height based on a
different geoid height model?
For example, a height is determined
using GEOID12A but I want it based on GEOID09 instead.
A : Geoid height models are determined from the most current heights in the database. Sometimes, it is beneficial to look at the value that would have been determined from an earlier model. To determine the orthometric height based on a new geoid height model, you would take the old value and add the old geoid height value and subtract the new geoid height value:
Ortho height _(new geoid) = Ortho height _(old geoid) + Geoid height _(old geoid) - Geoid height _(new geoid)
Have a question? CONTACT US!
