Although the NGVD29, until 1973, was known as the Sea Level Datum of 1929, it is not sea level or sea level extended across the continent (geoid) but could be considered an attempt to model a sea level surface across the continental United States. The origin of NGVD29 was based on an adjustment that distorted actual measurements by constraining to local mean sea level at 26 tide stations along the east and west coasts of the United States, Canada and the Gulf of Mexico.

In 1978 the National Geodetic Survey (NGS) began an effort to combine leveling surveys into a single least squares adjustment to provide improved heights for over 700,000 vertical control points. The adjustment was completed in June, 1991 and was designated NAVD88. This new datum has its origin at the primary tidal benchmark at Father Point/Rimouski, Quebec, Canada. This water level station is located on the mouth of the Saint Lawrence River and was selected, among other reasons, to minimize the variation nationally from NGVD29.

The critical thing to consider at this point is the NGV29 and NAVD88 datums are unrelated because they have different origins and definitions of height and also measurements. For this reason there is no direct (constant) relationship between the two systems even though they may be referenced to the same monument. In California, elevations based on NAVD88 will be greater than NGVD29 by two to three feet. A computer program, “Vertcon”, is available at no cost from the NGS website that will approximate a translation from one datum to the other.

For the majority of our work in the City the foregoing may be sufficient to gain an understanding of the vertical datums in use and perhaps keep us out of trouble. However, for those so inclined to read further, let’s explore the past, some more definitions and what the future may bring.

While the previous describes the essential differences between the two datums, there are some further definitions to consider. A geoid can be defined as a theoretical irregular surface that represents mean sea level extending across the face of the earth. A way to visualize the geoid is to imagine the surface of the planet consisting of only water and completely free of currents, tides, temperature variations and all other physical forces, except gravity. It is a equipotential surface, meaning the force of gravity is equal (water does not flow). The surface of the geoid is irregular because the force and direction of gravity is not constant across the planet. Surveyors measure elevations (orthometric heights) relative to the geoid using differential or trigonometric leveling techniques. A surveyor’s leveling instrument is “level” when it is perpendicular to the force of gravity at that setup. The direction of the perpendicular line (or a plumb bob on the end of a string) is not likely to be coincident with a line radiating to the center of the Earth’s mass. This concept will be dealt with in more detail when we look at horizontal datums.

Another important definition, particularly with regard to the use of Global Positioning System (GPS) equipment, is that of the ellipsoid defined by the North American Datum of 1983 (NAD83). This datum reference is an earth-centered, earth-fixed ellipsoid of revolution that best fits the global geoid. Again we’ll look at the NAD83 datum in more detail under horizontal datums. Therefore NAD83 is also a vertical, in addition to a horizontal, datum. The distance between the ellipsoidal surface and the geoid is called the geoid separation. In California the geoid is below the ellipsoid by approximately 30 meters. To clarify, at the ocean shoreline the ellipsoid surface is about 100 feet overhead. GPS measurements are related to the ellipsoid and provide very accurate relative ellipsoidal height differences between points on the earth's surface. Unfortunately GPS measurements do not give any information about orthometric heights such as would be obtained by conventional leveling. One way to obtain orthometric heights with GPS is by employing a “geoid model” that seeks to closely approximate the vertical difference between the ellipsoid and the geoid for each geographic location. The geoid model is available from NGS and they are continually improving the accuracy of the model.

In looking at the history of vertical datums and leveling in the City of Los Angeles we must include yet another datum. The Survey Division began a network of leveled bench marks in 1908. The lines radiated out from City Hall (presumably the City Hall in the 200 block of Broadway) into all sections of the city as of that date. I was not able to discover the datum for this leveling. However we do know that for the Los Angeles City datum prior to July 1, 1925, “zero” was 1.075 feet above the highest observed tide from 1924 to 1945. The “zero” for NGVD 29, mean sea level at San Pedro Harbor, is 5.775 lower than this previous datum. By Ordinance 52222 (July 1, 1925) the City adopted, what would be know as, the NGVD29 which was based on observations by the NGS for the years 1924-1932. I know that it is a little confusing to be adopting a datum 4 years before the apparent creation of the datum. The same thing will occur with horizontal datums. Suffice it to say that the definitions for the datums are developed in advance of the naming and implementation of the datum. The important thing to be aware of is that the elevations shown on plans produced prior to July 1, 1925 will need to have 5.775 added to them to make them equivalent to the NGVD29. Thankfully this is a constant value that can be used to convert elevations between these datums. In 1933 the Survey Division proposed a new network of leveling. It was to consist of two large nets and one loop. One net would cover the Valley, Eagle Rock and Boyle Heights areas with a total length of 111.5 miles a second net would cover the metropolitan, Venice and southern areas with a total length of 82.5 miles. The loop would cover the Shoestring and Harbor district and total 31 miles. 300 new permanent bench marks would be set at a spacing of about 4,000 feet for this network. The level party would consist of one surveyor (party chief), one instrumentman and four chainmen. The cost (salary) for this six-person crew was $40 per day! The total cost for 225 miles of first order leveling, setting of 300 monuments and office calculations was $5,565. I think the foregoing affords a good insight into the Great Depression of the 30’s. From this time until the 1980’s the City has been re-leveled at approximately five-year intervals, the network growing to over 17,000 bench marks. For reasons of economy only 85% of the 1975 adjusted bench marks were reobserved for the 1980 adjustment. The final observation on the NGVD29 was realized in the 1985 adjustment. Since this time, due to budget constraints, the Division’s leveling efforts have been sporadic. Beginning in the late 1990’s levels were run in the Valley on the new NAVD88 and tied into bench marks established by the NGS following the Northridge earthquake. The current procedures call for establishing geographic coordinates on the bench marks, enabling the new bench mark data to be included in the GIS.

Let’s look at benchmark 08-05835, near the intersection of Laurel Cyn and Woodbridge St.

Note that 2000 is the NAVD88, the others are of course NGVD29.

The differences in elevation shown for this bench mark do not necessarily reflect movement of the monument but rather an adjustment needed to constrain this bench mark to a primary line of leveling to which this point is related. For example with the 1980 adjustment, the primary lines were simultaneously adjusted to fit the 1978 Southern California Releveling Program performed and adjusted by the NGS.

Therefore it is important to consider both the datum and the adjustment of the benchmarks from which elevations are generated and shown on preliminary surveys and construction plans. For example caution must be used in relating elevations between sewer plans that have utilized different bench mark adjustments. It may be necessary to perform a new preliminary survey that will tie the two projects to a common adjustment/datum.

What about the future of leveling? Since the 1920’s the City has relied on a relatively dense network of vertical control provided by the NGS to base our leveling upon. Due to technological advances and budget constraints (surprise!) the NGS will be performing very little field work in the future. With Federal grants and NGS guidance an organization called the California Spatial Reference Center (CSRC) was initiated to develop and disseminate a modern, statewide control network for California. The ultimate goal of the CSRC is to establish both vertical and horizontal control networks consisting entirely of the array of permanent Continuously Operating Reference Stations (CORS). These CORS collect GPS data 24/7/365 and the position information is made available over the Internet. The spacing, while quite dense in Southern California, is roughly 10 kilometers. Perhaps in the future elevations will be obtained exclusively with GPS, using a CORS for a bench mark reference (backsite) that is 10 kilometers distant. Technology and the expense of maintaining passive, in-ground monumentation will drive the transition to “space based surveying” (GPS).

I hope this will help to clarify the subject of vertical datums and elevations. I intend to follow this up with a discussion of horizontal datums and coordinates in the near future. An excellent tutorial on GPS by Trimble Corp. can be found here: http://www.trimble.com/gps/index.html. Further information on CORS from the NGS: http://www.ngs.noaa.gov/CORS/cors-data.html.