The GPS Observer: The National Spatial Reference System of 2007
Was This Readjustment Necessary?The answer is indubitably YES. Too many local (state and region) readjustments, perhaps as many as 30, prevented surveys from crossing state and region lines. The only “national” network was the Continuously Operating Reference Station (CORS) system, and these stations cannot be occupied by field crews.
The problem began after the first NAD 83 readjustment: NAD 83 (1986). This readjustment, for all practical purposes, contained conventional observations established by triangulation and high-accuracy traverses. With the exception of Alaska and perhaps a few isolated areas, GPS observations were not in the readjustment. These were the early days of GPS. But NGS, hoping to release the readjustment in 1983, had to wait three years for Congress to grant funds.2
The readjustment was eventually released in 1986. It was a new datum that used the earth-centered Geodetic Reference System of 1980 (GRS 80) ellipsoid instead of the old Clarke 1866 spheroid.
By 1986, GPS had made a name for itself by being at least an order of magnitude more accurate than observations in the 1986 readjustment. With the influx of GPS data, the recently completed NAD 83 (1986) readjustment was quickly becoming obsolete. Dual-frequency GPS receivers such as the TI4100 were being manufactured by Texas Instruments. These were the days of the Block I constellation only; the P-codes on L1 and L2 were not encrypted. Using relative positioning, it was possible to resolve a vector to a precision of better than 1 part in 10 million. The U.S. Federal Geodetic Control Committee published a document titled “Geometric Geodetic Accuracy Standards and Specifications for Using GPS Relative Positioning Techniques”3 that included, among other things, accuracy standards for GPS control surveys. Figure 1 is Table 1 from that document. Notice that in addition to the old accuracy standards under Order C, there is Order AA, Order A and Order B.
The decision made by NGS to handle the problem with NAD 83 (1986) was, in my opinion, brilliant. It would have been impossible to ask for funds for a new national readjustment. They approached each state with an offer to establish a HARN on a cost-sharing basis. The HARN would be a network of stations with a minimum accuracy of Order B. With dual-frequency receivers, the vector between stations could be significantly longer than those established with single-frequency instruments.
Many states quickly accepted this offer, while others took a wait-and-see position. On page 67, Figure 2 shows the HARN status in 1993 and Figure 3 shows the final status in 1997. In Figure 3, you can see why I used the expression “states and regions.” Many states were combined in one readjustment. After the completion of each HARN, NGS readjusted all horizontal control in the state or region to the HARN. This created, as stated earlier, about 30 local networks. Each station, in every state, had a horizontal accuracy number or letter.
Many states and regions were reobserved after the HARNs were established, with new stations added. That’s why you will see stations that have had three different sets of coordinates.
In the mid-1990s, the network of CORS was established. Through a cooperative agreement, many state agencies and private firms added CORS to this network. I’ve been told that this network is adjusted daily. The coordinates are in the International Terrestrial Reference Frame (ITRF) and are transformed to NAD 83 (CORS). This, in reality, has created another coordinate system.
The New ReadjustmentBecause of the high accuracy and consistency of the national CORS network, NGS made the decision to do the following:
1. Eliminate all non-GPS stations from the readjustment.
2. Hold the CORS fixed.
At first, one might question how they can perform a readjustment by holding more than seven parameters fixed. People like me have been harping “minimally constrained adjustments” ever since GPS came onto the market. In this case, NGS developed a new math model that does just that.
The readjustment was completed in February 2007. The notation for all stations, except the CORS, was originally NAD 83 (NSRS 2007); NSRS is the acronym for National Spatial Reference System. Since that time, the NSRS has been dropped and NAD 83 (2007) is the notation.
Data SheetsThere was a data sheet generated for every control point in every readjustment. In earlier readjustments, each station had a horizontal accuracy, noted HORZ ORDER. The data sheets for the new readjustment eliminated the order and replaced it with “Accuracies in the directions North, East, Ellip.” Most of us are familiar with the notation North, East, Up. They couldn’t use Up because that’s the gravity vector pointed in the opposite direction. Ellip is the vector perpendicular to the ellipsoid pointed in the opposite direction. Figure 4 below and Figure 5 on page 68 show the difference on the data sheets.
Figure 4 shows the data sheet for station RDG ARP 2 1963, a station in the Pennsylvania HARN. It shows the horizontal coordinates as being from the readjustment NAD 83 (1992). The horizontal order, HORZ ORDER, is B. Figure 5 shows the data sheet for the same station after the 2007 readjustment; comparing coordinates to Figure 4 shows how they have changed. The adjustment is NAD 83 (2007), but instead of HORZ ORDER, it gives “Accuracy Estimates (at 95% confidence level in centimeters).” As stated earlier, the values are North, East, Ellip.
The NAD 83 (2007) is a new readjustment but not a new datum. All CORS are still NAD 83 (CORS). Although I haven’t seen it in print on a data sheet, NGS posted this statement to its Web site on Nov. 15, 2007:
For survey control stations determined ‘NO CHECK’ by the national readjustment, the published NAD83 coordinate line has been designated ‘NO CHECK’ (replacing ‘ADJUSTED’) and the ELLIP HEIGHT line has been designated ‘NO CHECK’ (replacing ‘GPS OBS’).
The Products and Service Committee of NGS voted to rescind the 1998 NGS policy which stated that a transformation would be provided between older versions of NAD 83 and NAD 83(NSRS 2007). Initial tests indicated that the shifts were too small, the accuracy of the shifts too large, and sparseness of the data too great to produce a useful transformation.4
This readjustment came up for discussion at an NGS-sponsored training session I attended on digital leveling. An NGS employee stated that not all CORS were held fixed in the readjustment. I contacted my NGS guru, Dave Doyle (also known as the NGS deputy director). Dave said, “Technically, all were held fixed. However, there were some stations where the stability was questioned, and these were not included in the readjustment.” If any readers have heard Dave speak at a seminar, you know he is a great communicator. At one seminar, he told a story that relates to the stability problem.
Dave’s wife, Nancy, works at NGS in the division that adjusts the CORS network. One day she came to Dave and said, “Our last adjustment shows a major shift in one of the Coast Guard CORS in Florida.” (I won’t mention where in Florida.) The next day she came back to Dave and told him the shift was still there. One of the CORS team members then placed a telephone call to the Coast Guard station and reached one of the officers. He asked if they had moved the GPS antenna on the day the shift was noticed. The officer said no, but later in the conversation said, “We had a major storm pass through this area about that time. Let me go up on the roof [where the antenna was located] to see if the antenna was damaged.” The officer called back later and said the storm broke one of the guy wires, and the antenna had swung to the side of its original position. This shift in position was detected at the NGS office in Silver Spring, Md., more than 1,000 miles away!
The official name of the NAD 83 re-adjustment is "NAD 83(NSRS2007)." In order to minimize impacts on computer programs, which rely on the existing format of NGS datasheets, the "datum tag" on datasheets was shortened to "NAD 83(2007)" with an explanatory statement that this datum tag is a shorthand notation for "NAD 83(NSRS2007)." The use of "NAD 83(2007)" is solely a formatting issue on datasheets and should not be used as an official name for the re-adjustment of NAD 83.
References1 These articles include “The GPS Observer: Readjustment of the National Spatial Reference System” (November 2004) and Mark Meade’s “From the Ground Up: National readjustment of NAD 83” (January 2007).
2 There is an excellent publication available that covers every detail associated with this readjustment: NOAA Professional Paper NOS 2, “North American Datum of 1983,” edited by Charles R. Schwarz, National Geodetic Survey, December 1989.
3 “Geometric Geodetic Accuracy Standards and Specifications for Using GPS Relative Positioning Techniques,” Federal Geodetic Control Committee, Rear Adm. Wesley V. Hull, Chairman, Version 5.0: May 11,1988. Reprinted with corrections August 1, 1989. Available online at www.ngs.noaa.gov/FGCS/tech_pub/GeomGeod.pdf.
4 Both of these announcements, along with other information from NGS about the readjustment, are accessible at www.ngs.noaa.gov/NationalReadjustment