Blog

This is the story of two surveyors who met after retirement to accomplish two geodetic leveling projects in Maine and New Hampshire. Independent of each other, we had vacationed in those states, skiing and hiking. The idea of doing some geodetic leveling in that area appealed to us.

We first re-leveled parts of a 1942 Coast & Geodetic Survey (C&GS) line between North Conway, New Hampshire, and Gilead, Maine, to honor the surveyors of World War II. During summer trips, we looked for the benchmarks along the line, NGS archive L9971, and figured out which ones were missing. We set new marks as needed. Leveling took place between September 9 and 22, 2013, archived by NGS as L28096, Second Order Class II.

We used K&E Lovar yard rods — nice light wood rods with rod level, such as the U.S. Geological Survey (USGS) is believed to have used leveling up Mount Washington in 1925 — and a Zeiss Ni-2 level with 1:33 stadia interval (to convert yards to feet distance). Leveling up and over Hurricane Mountain Road NE of North Conway was a precursor to leveling up Mount Washington in 2014. We used a handheld GPS receiver to obtain the coordinates for marks that did not already have good accuracy. We obtained Online Positioning User Service (OPUS) coordinates for the reference marks on top of the mountain.

After completing the 2013 releveling of the 1942 C&GS line, we wondered about releveling the 1925 USGS line up Mount Washington. We approached the general manager of the Mount Washington Auto Road, Howie “Weems” Wemyss, explained what we proposed to do, and asked for his buy-in. He approved the project enthusiastically, allowing us full access to the road for the project. On May 12, we began the level run from first-order benchmarks in Gorham, New Hampshire, and finished on June 8.

Until this 2014 project, the elevations of Mount Washington were referenced to the National Vertical Datum of 1929 (NGVD 29) and were only available on paper from USGS. The professional purposes of the project were to connect Mount Washington to the North American Vertical Datum of 1988 (NAVD 88) and make the benchmark data available online from the National Geodic Survey (NGS). Two personal purposes were to continue enjoying the natural area and meet the challenges of the topography and weather.

Deciding where to place each level setup on the steep terrain involved a compromise between time spent trying to read at the very top of the rod and a quicker setup decision that usually sacrificed 1.0-1.7 ft of the rod. The maximum vertical gain on the 10-ft rod was typically about 7.5 ft per setup, sighting high on the lower rod while sighting on the upper rod above the yard equivalent of 0.5 m — thereby reducing refraction errors per recommended procedures. Sight lengths on the Auto Road were as short as 10 ft, averaging 30 ft. Along the less steep terrain of Route 16 between Gorham and the Auto Road, leveling between marks set by NHDOT, sight lengths ranged around 150 ft. Temporary benchmarks on boulders were selected between the MAC marks on the Auto Road to keep the number of setups in a section below 30. That way, if the forward and backward runnings of a section misclosed, it was not a hopeless number of setups to rerun and isolate the error. The grade between MAC marks was 14% in two cases, otherwise around 10%. To eliminate 1-yd reading errors, 1-ft markings were painted on the side of the rod and read to x.1 ft before the 3-yd readings (top, middle and bottom wires), then the foot and yard readings were compared. The typical elevation difference between a MAC mark and a TBM or between TBMs was 150 ft. The typical time per setup was 4.5 minutes, depending on how far apart the turning points were.

One bit of unwanted excitement occurred while the road crew was doing the annual rebuild of the 1-mile gravel section above Mile Post 5. The road workers pulled out the flagging at MAC 102, which is a bit hidden by vegetation. Evidently, it conflicted with the orange flagging they were also using. Leveling continued right past the disk for five setups until Bob noticed green paint on a boulder used as a turning point in the prior run. We had to reverse back to the mark. Fortunately, the section closed even with the extra setups.

There are some beautiful small waterfalls along the road that are easier to admire when you are walking by, not trying to keep your vehicle on the road. One day, we spotted a momma bear and two cubs. One man stopped to reminisce about his time working with Brad Washburn while Brad was doing extensive surveying and mapping on the mountain.

The USGS benchmarks are designated MAC 97 to MAC 104, “MAC” being the initials of the USGS party chief followed by a sequential number as the party progressed along the level line. There are lower numbered ones away from the Auto Road, and there used to be some higher numbered ones, which have since been destroyed.

The project data was accepted, adjusted and published by NGS as archive L28128. The elevation of the highest and most visited summit mark, Mount Wash Reset (DP4904), by the cairn and the brown park sign, dropped 0.77 ft, referenced to NAVD 88 compared to its elevation referenced to NGVD 29. The NGVD 29 elevation published by USGS (solely in feet), based on the 1925 leveling, was 6,288.176 ft (divided by 3.2808333 ft/m to obtain 1,916.634 m). These values were rounded off to 6,288 ft and 1,917 m, respectively, for display on the summit sign.

The NAVD 88 elevation for Mount Wash Reset, published by NGS based on this 2014 leveling, is 6,287.41 ft or 1,916.406 m. These values would round down to the next lower whole unit, changing the elevations that could be displayed on the park sign and making the mountain’s elevation lower. Since re-determining the mountain’s elevation was not a priority of this project, and a different elevation than the 6,288 ft that had been used for years would mean changing all those mugs, T-shirts and other tourist paraphernalia, we are not suggesting changing the elevation with which the public is familiar.

All the differences of elevation observed between the various USGS benchmarks agreed within Second Order Class II limits with the previous 1925 work. At that level of precision (section closures less than or equal to 8 mm multiplied by the square root of the distance in km), the 2014 leveling did not reveal or detect vertical motion on the mountain, after an interval of 89 years. The project could always be re-run with first-order equipment and procedures, cutting the section closure error budget in half to 4 mm multiplied by the square root of the distance. Various marks could be observed in a regional high-precision geoid modeling campaign.