NAD 27 DATUMS and COORDINATES

There are three things about the NAD 27 datum and coordinates that are often overlooked, but are important to recognize:

* As stated in Part I, to within the accuracy the observations and methods available at the time make possible, NAD 27 coordinates are Earth Centered-Earth Fixed (ECEF) coordinates.

* The points whose coordinates are directly computed from the observations are not the surface reference stations, but, rather, the projection of these reference stations onto the surface of the reference ellipsoid of the datum.

* The coordinates obtained for the projection points are three dimensional (3D) coordinates.

With the NAD 27 datum, as was the case with datums previous to it, defining the datum was achieved by defining the reference ellipsoid through specifying the following parameters:

* The semimajor axis and the flatening, ie. the size and shape, of the reference ellipsoid.

* The 3D ellipsoidal coordinates (latitude, longitude, and ellipsoid height) of one of the reference stations, designated as the datum origin point.

* The geodetic azimuth from the datum origin point to a nearby station

Beginning from the origin point, the coordinates of all other reference stations are computed from the observations.

To satisfy to the extent possible the four required conditions indicated in Part I of these Posts the condition imposed in the selection of the above named parameters was that they be chosen so as to minimize the average difference between geodetic and astronomic latitude and longitude at those reference stations across the network where astronomic coordinates had been observed. Another use of the astronomic observations was to compute corrected geodetic azimuths at stations where astronomic longitudes and azimuths had been observed. In both cases astronomic observations of optical stars played an important role in defining NAD 27 and computing NAD 27 coordinates.

In the 1920's, when all computations were performed by hand, less than rigorous approaches were taken to minimize computational effort as much as possible. For example, the coordinates used for the NAD 27 datum origin point, Meades Ranch, were taken to be those used in a previous datum computation, since it was found that they could be used with the new International ellipsoid and give satisfactory results and this would greatly reduce computations.

As to the other two overlooked points, consider how NAD 27 reference station coordinates were actually computed. First, all observations were "reduced" to the surface of the reference ellipsoid. That is, corrections were applied to the observations made at the reference stations on the Earth's surface to obtain estimates of what these observations would have been if they had been made at the projections of the surface points onto the surface of the NAD 27 reference ellipsoid. These "reduced" observations were then used to compute 3D coordinates of the projection points of all other reference stations while holding fixed the latitude and longitude of the origin point. Since the projection points lay on the surface of the reference ellipsoid, their ellipsoid heights were, by definition, zero and only the latitudes and longitudes of the projection points actually needed to be computed from the observations to have their 3D positions. The reductions to the ellipsoid were performed in such a way that the latitude and longitude of a projection point on the reference ellipsoid and the latitude and longitude of the associated reference station on the Earth's surface would nominally be the same. Thus the computations produced, not only 3D coordinates of the projection points but also the latitudes and longitudes of the reference stations themselves.

There are several points to be made about the NAD 27 reference station coordinates:

* Enforcing the astronomic constraints mentioned above

It was a surprise that this one did not spark more comments. I waited a few days to give others a chance before I poked at it.

It has always been a struggle to get GPS work to come together on NAD27. After some experiments at Fort Sill, Oklahoma a bunch of years ago I decided that it made more sense to process the data on NAD83(1993) and then convert the coordinates to NAD27.

The damage done by the coordinate conversion was not worse than the damage done by constraining a survey to several NAD27 first order monuments.

These days I try to focus on converting background maps to NAD83 rather than messing up good GPS work to make it match a crummy coordinate system. Seems to me that the damage done to the maps by the conversion is a smaller problem than the damage done to the GPS work by that same conversion.

Smart move Phil. I have recommended to surveyors for 17 years that they move away from NAD 27. I have seen so many try and perform constrained adjustments of GPS to NAD 27, and with very rare exceptions they all fail to maintain the integrity of the original data. By using a recognized transformation such as NADCON the ultimate user of the data at least maintains integrity when they hopefully transform back to NAD 83 at some point in the future.

It doesn't make sense to take good data and warp it to fit into the NAD-27 adjustment. (GPS holding NAD-27 stations)

Use the map transformations if you need NAD-27 (although I can't think of any requests for NAD-27 in a few years)..

NAD-27 was excellent for it's time, but it's time has passed (actually many years ago).

I still marvel at the agreement that Phil's 1996 survey has with our 2000 survey.

tm

Based on Mr. G's comments below, I would assume that the change from 83 to ITRF will not be as painful.

I would agree that using 27 coordinates for constraining a survey is a bad idea. This is one of the areas where a surveyor new to GPS and datums can stub his/her toe very easily. Maybe all receivers should carry a warning " The Surveyor General has determined that using NAD27 coordinates can be harmful to your health".

Mark

That's good.

"NAD-27 can be harmful to your position"..

tm