The Early Vertical Datums of the U. S. Geological Survey

A 6.519Mb PDF of this article as it appeared in the magazine—complete with images—is available by clicking HERE

During the 1896 field season, the surveying crews of the U. S. Geological Survey (USGS) engaged in higher quality and more precise topographic mapping and spirit leveling than had been previously attempted. This change was largely prompted by an act issued by the Fifty-Fourth Congress on June 11, 1896. For USGS, the most significant result of the act was that permanent bench marks on an approximate sea-level based datum were to be placed in each area of the new surveys. These bench marks were to be marked on the ground with at least two monuments established per township, except in forested and mountainous areas where at least one per township was required. Although USGS had already performed topographic mapping in many regions of the country for 30-minute quad sheets, permanent bench marks had not been placed in those areas.

The framers of the act knew it would be impossible for USGS to accurately record the monuments to their exact height above sea level. Attempting to do so would have required running thousands of miles of precise levels from points already known to be accurate. This would not have been financially feasible and it would have taken years to accomplish, greatly delaying the mapping program.

Therefore, a plan was designed to permit the acceptance of one fixed monument within each region of a particular topographic mapping area to be used as a central datum only for that area. The elevation of each initial bench mark, regardless of its relationship to the true sea-level datum, was usually derived from an outside source such as a railroad company, a river commission, or a local city datum. The elevations of these starting bench marks were of uncertain precision, but the assumption was made that they were very close to sea-level datum. All subsequent bench marks placed from the initial bench mark were then directly related to that monument, but they could not be interchanged with bench marks established from other datums. It was anticipated that through the course of the work, connections would eventually be made with precise monuments established by an agency such as the U. S. Coast & Geodetic Survey (USC&GS). Thus, adjustments in elevation could later be made to the USGS monuments without having to delay the mapping program.

Initially, the plan did not implement a specific code or numbering system for the bench marks. Only the approximate elevation was hand-stamped on the monument to the nearest foot. This stamped number individually identified each bench mark within the datum except in cases where two or more bench marks within the same datum had the same elevation. This system quickly proved insufficient since datums originating from different sources began overlapping into common areas which would make it impossible for surveyors and engineers to determine from which datum each bench mark originated.

In an improved system during the second year, each datum was assigned a specific code name which was stamped onto the bench mark in addition to the elevation. In theory, if a surveyor leveling between any two bench marks discovered that they did not close, the differing code names on the monuments was supposed to prompt the surveyor to contact USGS to inquire as to the problem.

The exact number of early USGS vertical datums is unknown since a concise list has not been found, but published records indicate there were hundreds of different datums located across the United States. The coding for the datums was apparently left to the individual party chief of each crew without much communication between crews working in other parts of the country. The same code names, therefore, existed in different states which were based upon entirely different datums. For instance, the letter "A" was a code used for different datums originating near Anniston, AL; Alexandria, MO; Asheville, NC; Albany, NY; Athens, OH; and Astoria, OR. The letter "B" was a code used for different datums near Benicia, CA; Bannock County, ID; Bangor, ME; Brockport, NY; Blaine, WA; and Baraboo, WI. The "LA" code was used for datums originating at Lafayette, LA, and Los Angeles, CA. Some datums were designed by a date instead of a letter code. The "1906" datum in Colorado had a different origin than the "1906" datums established in Illinois, Kentucky and Oregon. Most codes, however, had some indication of the city name where the datum originated such as "CHYN" for Cheyenne, WY; "GAINV" for Gainesville, TX; "MSLA" for Missoula, MT; "NOGLS" for Nogales, AZ; "VAN HN" for Van Horn, TX; or "YNKTN" for Yankton, SD.

There were variations to the codes within the same datum such as "D", "DENV", and "DENVER" in Colorado; "MIL" and "MILWAUKEE" in Wisconsin; "MLT" and "MALTA" in Montana; and "WP", "WILLETS", and "WILLETS POINT" in New York. These variations were likely the result of individual field crews abbreviating the codes within the same datum and not being consistent with the stamping.

The coding is equally confusing when code names such as "GRAFTON", "GRAFTON 1901", "GRAFTON 1902", and "GRAFTON 1903" were stamped upon bench marks located in the same mapping area. This is due to additional data being obtained each year that was closer to true sea level and a distinction needed to be made in the code as new bench marks were established with slightly adjusted elevations within that datum.

The "ADJ 1903" datum code is prolific in over twenty different states because USC&GS performed a major datum adjustment during that year. To show that new bench marks reflected that shift, USGS began stamping their new monuments with this code. On mountain peaks where leveling was not easily accomplished, the bench marks were marked with "VA", which means the elevation was obtained by "Vertical Angle" from another location.

Although some datums covered large areas, many were isolated and confined to just one particular mapping area. North Dakota had two individual datums. The "JMTN" datum at Jamestown established only 25 bench marks in that area, while the "W" datum at Wahpeton established only 20 bench marks. The permanent bench marks were shown on the 30-minute quad sheets as a "X" with the letters "BM" next to the elevation all in black-colored ink. The temporary bench marks were shown with a brown-colored "X" and elevation, but without the letters "BM".

Three main types of bench marks were chosen for the early USGS monuments. The first was a flat, circular bronze or aluminum tablet, 3½ inches in diameter and one-fourth of an inch thick. These were placed on walls of public buildings, bridge abutments, large boulders, or on other masonry structures. The second was a slotted copper pin known as a "bolt" measuring one-inch in diameter and four inches long. The bolt was used in areas of rock where a drill hole was first made, then a brass wedge was placed into the bottom of the hole, and the slotted bolt was driven down onto it. The third type, and most prolific, was a hollow iron post, 4 feet 6 inches in length, 3½ inches in outside diameter with a 3¾-inch diameter bronze cap riveted to the top. This post had a split bottom that expanded outward to prevent subsidence and also to deter malicious theft by pulling it out. Th
e post was coated with asphalt and only the top one foot was exposed above the ground. For temporary points, a 7/8-inch diameter stamped copper washer with a 1-inch copper nail or a common iron nail was used. In certain areas of the Pacific Northwest, aluminum tags were often placed on trees to mark the location of the temporary bench marks.

Soon after the establishment of the bench marks, USGS began publishing the locations with the elevations relative to the local datums upon which they were established. These descriptions were initially published in the Appendix of USGS 18th Annual Report for the fiscal year 1896-97. The first USGS publication specifically listing only horizontal and vertical monuments, was Bulletin No. 175, published in 1900, Results of Triangulation & Spirit Leveling in Indian Territory (today the State of Oklahoma). The following year, Bulletin No. 185, Results of Spirit Leveling ­ Fiscal Year 1900-01, was published and made available. This publication listed the known elevations for bench marks established in 26 different states. Then, beginning in 1906, USGS began publishing bulletins for each individual state beginning with the state of New York. Eventually, nearly every state had a specific bulletin, or one in combination with an adjacent state, that listed the location, the type of monument, and the elevation of each bench mark established by USGS dating back to the late 1890’s. By 1918, 76 different bulletins had been published listing the early USGS bench marks.

It was a tremendous accomplishment for USGS to map various areas of the country while using local datums with approximate sea-level elevations. As the mapping progressed, it was soon discovered that the program had some serious flaws. A notation in Bulletin No. 441 for the State of Georgia stated the leveling in certain quadrangles performed in 1896, before the code names were assigned, had bench marks stamped with an elevation of 33 feet too high in one area and 15 feet too high in another area because of spurious information supplied by a railroad company. An area in North Carolina, also using a railroad supplied elevation in 1896, was later determined to have bench marks stamped 10 feet too high. Bench marks near Stevensville, Montana, were stamped 49 feet too low on the "NP" datum because of an incorrect elevation supplied by the Northern Pacific Railroad. Bench marks based upon the "NASHVILLE" datum in Tennessee were later determined to be 3 feet too high in one location and 5 feet too high in another location. Similar situations occurred all across the United States to varying degrees.

A confusing situation occurred with the "SHER" datum at Sheridan, Wyoming. The elevations for this datum were based upon a bronze tablet placed in the side of the city hall building at Sheridan and stamped "SHER 3738". The height of this bench mark was derived from a bench mark of the Burlington & Missouri River Railroad (B&MR) on a bridge near Sheridan. The height of the railroad’s bench mark was reportedly "corrected" by subtracting 12 feet on account of a difference discovered from a check between the B&MR Railroad and the Northern Pacific Railway’s junction at Huntley, Montana. The Northern Pacific had carried their elevations west from St. Paul, Minnesota. The locations, the types of bench marks, and the elevations for the Sheridan Datum were initially published in the USGS Annual Report of 1897-98. Each bench mark was stamped to the nearest foot based upon the 12-foot elevation shift. It was later determined, however, that the B&MR Railroad’s elevations were initially correct without making the shift. By the time the quadrangle sheets were published for that area in 1911, the elevations listed next to the bench marks on the quad sheets differed from what was stamped on the bench mark caps by 12 feet. Later, when Bulletin No. 558 was published in 1914 for the State of Wyoming, a note indicated the field work had been dependent upon an assumed railroad datum and that most of the bench marks were stamped 10 feet too low. Then, with preliminary corrections, later bench marks in that area were stamped 3 feet too high. It became a confusing situation where the published notes in the bulletins, the labeling on the quad sheets, and the stamping on the monuments did not agree. Today, the initial bench mark which began the datum at Sheridan has been determined by the National Geodetic Survey to be 12.110 feet higher (NAVD88) than originally thought by USGS.

The early use of isolated datums by USGS mostly ended by 1910, but continued as late as 1920 in some areas. By the 1930’s, USGS began their level lines upon bench marks having precise sea-level elevations on the Sea Level Datum of 1929 (later renamed NGVD29) and ending them upon others. The bench mark monuments established during that era were mostly domed bronze or aluminum disks set in concrete and were published in the Level Line sheets.

Many of the original USGS bench marks established over a century ago still exist and most have never had accurate sea-level based elevations established upon them unless they were included in newer leveling. Surveyors have wrongly assumed that these early bench marks are on the NGVD 1929 Datum, while not correctly understanding that they were established much earlier on local assumed datums with unknown certainty to true sea-level elevations.

Jerry Penry is a licensed land surveyor in Nebraska and South Dakota. He is a frequent contributor to the magazine.

A 6.519Mb PDF of this article as it appeared in the magazine—complete with images—is available by clicking HERE

About the Author

Jerry Penry, PS

Jerry Penry has been surveying for 34 years, is licensed in Nebraska and South Dakota, and has been employed with Lancaster County Engineering for 21 years. He is also serving his second term on the Board for the Professional Surveyors Association of Nebraska.